public- ct 5
lolo1339
WorldCancerReport.pdf
Lyon, 2014
World Cancer Report 2014 Edited by BERNARD W. STEWART and CHRISTOPHER P. WILD
Published by the International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon Cedex 08, France
©International Agency for Research on Cancer, 2014
Distributed by WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland
(tel: +41 22 791 3264; fax: +41 22 791 4857; email: [email protected]).
Publications of the World Health Organization enjoy copyright protection in accordance with the provisions of Protocol 2 of the Universal Copyright Convention. All rights reserved.
The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any
country, territory, city, or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned.
Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.
The authors alone are responsible for the views expressed in this publication.
The International Agency for Research on Cancer welcomes requests for permission to reproduce or translate its publications, in part or in full. Requests for permission to reproduce or translate IARC publications – whether for sale
or for noncommercial distribution – should be addressed to the IARC Communications Group, at: [email protected].
Cover images, from top to bottom: A woman in The Gambia prepares cereal, using scales to measure portions (Credit: C.P. Wild/IARC). Next-generation gene sequencing equipment in use at the Joint Genome Institute in Walnut Creek, California, USA (Credit: Joint Genome Institute, U.S. Department of Energy Genomic Science Program). Manhattan
plot of pooled genome-wide association study (GWAS) results (Credit: P. Brennan/IARC). A group of researchers from the Epigenetics Group in the Section of Mechanisms of Carcinogenesis at IARC (Credit: R. Dray/IARC).
Representative example of high-grade serous carcinoma, one of the main types of ovarian carcinoma (Credit: J. Prat, Autonomous University of Barcelona). Background image: The Blue Marble: Next Generation is a mosaic of satellite
data taken mostly from a NASA sensor called the Moderate Resolution Imaging Spectroradiometer (MODIS) that flies aboard NASA’s Terra and Aqua satellites (Credit: NASA/Goddard Space Flight Center/Reto Stöckli).
IARC Library Cataloguing in Publication Data
World cancer report / edited by Bernard W. Stewart and Christopher P. Wild
1. Neoplasms – etiology 2. Neoplasms – prevention and control 3. Neoplasms – genetics 4. Neoplasms by site 5. World Health I. Stewart, B.W. II. Wild, C.P.
ISBN 978-92-832-0443-5 (NLM Classification: W1)
IARC acknowledges generous assistance from the American Association for Cancer Research with the promotion of World Cancer Report 2014.
World Cancer Report 2014
Editors Bernard W. Stewart
Christopher P. Wild
Associate Editors Freddie Bray
David Forman
Hiroko Ohgaki
Kurt Straif
Andreas Ullrich
Managing Editor Nicolas Gaudin
English Editor Karen Müller
Project Manager Sylvia Lesage
Production Assistant Solène Quennehen
Illustrations Roland Dray
Layout www.messaggio.eu.com
Printing Naturaprint, France
IARC acknowledges generous assistance from the American Association for Cancer Research with the promotion of World Cancer Report 2014.
World Cancer Report 2014
Jean-Pierre Abastado (Singapore)
Cary Adams (Switzerland)
Isaac F. Adewole (Nigeria)
Hideyuki Akaza (Japan)
Naomi E. Allen (United Kingdom)
Marcella Alsan (USA)
Nada Al Alwan (Iraq)
Mahul B. Amin (USA)
Benjamin O. Anderson (USA)
Ahti Anttila (Finland)
Daniel A. Arber (USA)
Bruce K. Armstrong (Australia)
Héctor Arreola-Ornelas (Mexico)
Silvina Arrossi (Argentina)
Rifat Atun (United Kingdom and USA)
Robert A. Baan (France)
Yung-Jue Bang (Republic of Korea)
Emmanuel Barillot (France)
Jill Barnholtz-Sloan (USA)
Laura E. Beane Freeman (USA)
Rachid Bekkali (Morocco)
Agnès Binagwaho (Rwanda)
Elizabeth H. Blackburn (USA)
Evan Blecher (USA)
Ron Borland (Australia)
Fred T. Bosman (Switzerland)
Peter Bouwman (Netherlands)
Guledal Boztas (Turkey)
Elisabeth Brambilla (France)
Freddie Bray (France)
Paul Brennan (France)
Louise A. Brinton (USA)
Nathalie Broutet (Switzerland)
Michael P. Brown (Australia)
Heather Bryant (Canada)
Nikki Burdett (Australia)
Robert C. Burton (Australia)
Agnès Buzyn (France)
Kenneth P. Cantor (USA)
Federico Canzian (Germany)
Fátima Carneiro (Portugal)
Webster K. Cavenee (USA)
Eduardo L. Cazap (Argentina and Switzerland)
Frank J. Chaloupka (USA)
Stephen J. Chanock (USA)
Bob Chapman (USA)
Simon Chapman (Australia)
Chien-Jen Chen (Taiwan, China)
Il Ju Choi (Republic of Korea)
Rafael Moreira Claro (Brazil)
Hans Clevers (Netherlands)
Vincent Cogliano (USA)
Aaron J. Cohen (USA)
Carlo M. Croce (USA)
Min Dai (China)
Sarah C. Darby (United Kingdom)
Peter B. Dean (France)
Lynette Denny (South Africa)
Ethel-Michele de Villiers (Germany)
John E. Dick (Canada)
Joakim Dillner (Sweden)
Susan M. Domchek (USA)
Tandin Dorji (Bhutan)
Roland Dray (France)
Majid Ezzati (United Kingdom)
Lee Fairclough (Canada)
Jacques Ferlay (France)
David Forman (France)
Silvia Franceschi (France)
A. Lindsay Frazier (USA)
Christine M. Friedenreich (Canada)
Søren Friis (Denmark)
Tamara S. Galloway (United Kingdom)
Maurice Gatera (Rwanda)
Wentzel C.A. Gelderblom (South Africa)
Margaret A. Goodell (USA)
Sharon Lynn Grant (France)
Mel Greaves (United Kingdom)
Adèle C. Green (Australia and United Kingdom)
Murat Gultekin (Turkey)
Christine Guo Lian (USA)
Prakash C. Gupta (India)
Ezgi Hacikamiloglu (Turkey)
Andrew J. Hall (France)
Stanley R. Hamilton (USA)
Marianne Hammer (Norway)
Curtis C. Harris (USA)
Takanori Hattori (Japan)
Zdenko Herceg (France)
Hector Hernandez Vargas (France)
Rolando Herrero (France)
David Hill (Australia)
Martin Holcmann (Austria)
James F. Holland (USA)
Ralph H. Hruban (USA)
Thomas J. Hudson (Canada)
Peter A. Humphrey (USA)
Elaine S. Jaffe (USA)
Prabhat Jha (Canada)
Jos Jonkers (Netherlands)
Margaret R. Karagas (USA)
Michael Karin (USA)
Contributors
Namory Keita (Guinea)
Ausrele Kesminiene (France)
Michael C. Kew (South Africa)
Tim Key (United Kingdom)
Thiravud Khuhaprema (Thailand)
Paul Kleihues (Switzerland)
Günter Klöppel (Germany)
Felicia Marie Knaul (USA and Mexico)
Manolis Kogevinas (Spain)
Nobuo Koinuma (Japan)
Barnett S. Kramer (USA)
Guido Kroemer (France)
James R. Krycer (Australia)
Sunil R. Lakhani (Australia)
René Lambert (France)
Johanna W. Lampe (USA)
Robert R. Langley (USA)
Lawrence H. Lash (USA)
Mathieu Laversanne (France)
Eric Lavigne (Canada)
Eduardo Lazcano Ponce (Mexico)
Maria E. Leon (France)
Alex C. Liber (USA)
Jonathan Liberman (Australia)
Dongxin Lin (China)
Dana Loomis (France)
Alan D. Lopez (Australia)
Joannie Lortet-Tieulent (France)
Judith Mackay (Hong Kong Special Administrative Region, China)
Roger Magnusson (Australia)
Reza Malekzadeh (Islamic Republic of Iran)
Andrew Marx (USA)
Beela Sarah Mathew (India)
James D. McKay (France)
David Melzer (United Kingdom)
Oscar Méndez (Mexico)
Kathryn R. Middleton (USA)
Martin C. Mihm Jr (USA)
Anthony B. Miller (Canada)
J. David Miller (Canada)
Saverio Minucci (Italy)
Holger Moch (Switzerland)
Elizabeth A. Montgomery (USA)
Rebecca Morton Doherty (Switzerland)
Raul Hernando Murillo Moreno (Colombia)
Liam J. Murray (United Kingdom)
Richard Muwonge (France)
Robert Newton (United Kingdom)
Fidele Ngabo (Rwanda)
Mark J. Nieuwenhuijsen (Spain)
Jane R. Nilson (USA)
Omar Nimri (Jordan)
Folakemi T. Odedina (USA)
G. Johan Offerhaus (Netherlands)
Hiroko Ohgaki (France)
Magali Olivier (France)
Jørgen H. Olsen (Denmark)
Hongchao Pan (United Kingdom)
Pier Paolo Pandolfi (USA)
Gianpaolo Papaccio (Italy)
Chris Paraskeva (United Kingdom)
Werner Paulus (Germany)
José Rogelio Pérez Padilla (Mexico)
Richard Peto (United Kingdom)
Paul Pharoah (United Kingdom)
Gérald E. Piérard (Belgium and France)
Christopher J. Portier (USA)
Jaime Prat (Spain)
Rachel Purcell (France)
Pekka Puska (Finland)
You-Lin Qiao (China)
Thangarajan Rajkumar (India)
Kunnambath Ramadas (India)
G.K. Rath (India)
Cecily S. Ray (India)
Jürgen Rehm (Canada and Germany)
Luis Felipe Ribeiro Pinto (Brazil)
Elio Riboli (United Kingdom)
Leanne Riley (Switzerland)
Ronald T. Riley (USA)
Eve Roman (United Kingdom)
Isabelle Romieu (France)
Hana Ross (USA)
Lesley Rushton (United Kingdom)
Anne Lise Ryel (Norway)
Jonathan M. Samet (USA)
Massoud Samiei (Austria)
Rengaswamy Sankaranarayanan (France)
Rodolfo Saracci (France)
Guido Sauter (Germany)
Catherine Sauvaget (France)
Fabio Savarese (Austria)
Augustin Scalbert (France)
Ghislaine Scelo (France)
Mark Schiffman (USA)
Stuart J. Schnitt (USA)
Joachim Schüz (France)
Janice Seinfeld (Peru)
Surendra S. Shastri (India)
Kevin Shield (Canada)
Maria Sibilia (Austria)
Jack Siemiatycki (Canada)
Ronald Simon (Germany)
Pramil N. Singh (USA)
Rashmi Sinha (USA)
Nadia Slimani (France)
Avrum Spira (USA)
Gustavo Stefanoff (Brazil)
Eva Steliarova-Foucher (France)
Bernard W. Stewart (Australia)
Kurt Straif (France)
Simon B. Sutcliffe (Canada)
Rosemary Sutton (Australia)
Steven H. Swerdlow (USA)
Neil D. Theise (USA)
David B. Thomas (USA)
Lester D.R. Thompson (USA)
Michael J. Thun (USA)
Mark R. Thursz (United Kingdom)
Massimo Tommasino (France)
William D. Travis (USA)
Edward L. Trimble (USA)
Giorgio Trinchieri (USA)
Ugyen Tshomo (Bhutan)
Murat Tuncer (Turkey)
Andreas Ullrich (Switzerland)
Toshikazu Ushijima (Japan)
Carlos Vallejos (Peru)
Piet van den Brandt (Netherlands)
James W. Vardiman (USA)
Jim Vaught (USA)
Cesar G. Victora (Brazil)
Paolo Vineis (United Kingdom)
Lawrence von Karsa (France)
Melanie Wakefield (Australia)
Guiqi Wang (China)
Frank Weber (Germany)
Elisabete Weiderpass (Norway and Sweden)
Zena Werb (USA)
Theresa L. Whiteside (USA)
Christopher P. Wild (France)
Walter C. Willett (USA)
Dillwyn Williams (United Kingdom)
Rena R. Wing (USA)
Deborah M. Winn (USA)
Martin Wiseman (United Kingdom)
Scott Wittet (USA)
Magdalena B. Wozniak (France)
Hai Yan (USA)
Teruhiko Yoshida (Japan)
Jiri Zavadil (France)
Harald zur Hausen (Germany)
For a complete list of contributors and their affiliations, see pages 596–605.
Foreword ix Preface x Introduction xii
Perspectives Elizabeth H. Blackburn Reconciling stress and cancer: insights from telomeres 77
Cesar G. Victora Early-life exposures, birth cohorts, and noncommunicable diseases (with special reference to cancer) 177
Harald zur Hausen Prenatal infections with subsequent immune tolerance could explain the epidemiology of common childhood cancers 261
Mel Greaves An evolutionary foundation for cancer control 337
Walter C. Willett Diet, nutrition, and cancer: where next for public health? 432
Barnett S. Kramer Controversies in cancer screening and their resolution: a view from the United States “battleground” 522
Richard Peto The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality 586
1 Cancer worldwide 1.1 The global and regional burden of cancer 16
1.2 Transitions in human development and the global cancer burden 54
1.3 Childhood cancer 69
2 Cancer etiology 2.1 The global tobacco epidemic 82
2.2 Tobacco smoking and smokeless tobacco use 88 Genetic susceptibility to tobacco-related cancers
2.3 Alcohol consumption 96
2.4 Infections 105 Inflammation and cancer Do infections play a role in breast cancer?
2.5 Reproductive and hormonal factors 115
2.6 Diet, obesity, and physical activity 124 Challenges of measuring diet in cancer epidemiological studies and new perspectives Single-nucleotide polymorphisms relevant to meat consumption and cancer risk Diet and the gut microbiome
2.7 Occupation 134 Evaluating carcinogens: dioxins and dioxin-like substances Preventing occupational cancer: successes and failures
2.8 Radiation: ionizing, ultraviolet, and electromagnetic 143 An IARC announcement that made waves
2.9 Pollution of air, water, and soil 151 Bisphenol A and cancer Pesticides and cancer
2.10 Pharmaceutical drugs 161 Breast implants and cancer
2.11 Naturally occurring chemical carcinogens 169
3 Cancer biology 3.1 Genomics 184
TP53 mutations and human cancer
3.2 Genome-wide association studies 193 Gene–environment interactions and breast cancer DNA repair polymorphisms and human cancer
3.3 Gene expression 203 Therapeutic implications of dysregulated DNA damage signalling
3.4 Epigenetics 214 The non-coding RNA revolution in medical research Causes and consequences of microRNA dysregulation in cancer
3.5 Metabolic change and metabolomics 222 3.6 Stem cells and cancer stem cells 228
Stem cells in cancer: determinants of clinical outcome?
3.7 Tumour microenvironment 236 3.8 Signal transduction and targeted therapy 244 3.9 Immunology and immunotherapy 253
Role of the innate and acquired immune systems in mammary development and breast cancer Premortem autophagy and endoplasmic reticulum stress as immunogenic signals in cancer therapy
Contents
4 Cancer prevention 4.1 Changing behaviours – tobacco control 268
Australia’s plain packaging of tobacco products Tobacco and China
4.2 Changing behaviours – physical activity and weight control 281 Energy restriction, age, and cancer risk EPIC as a model study vehicle
4.3 Designing and evaluating population-wide campaigns 290
4.4 Prevention strategies common to noncommunicable diseases 298 The global economics of chronic and noncommunicable diseases
4.5 Legislative and regulatory initiatives 305 Environmental pollution: old and new Taxing sugar-sweetened beverages: the Brazilian case
4.6 Vaccination 314 Early implementation and monitoring of HPV vaccination in Bhutan Introduction and monitoring of a national HPV vaccination programme in Rwanda
4.7 Screening – principles 322 4.8 Screening – implementation 330
5 Cancer by organ site A guide to the epidemiology data in World Cancer Report 348 5.1 Lung cancer 350
Earliest molecular evidence of tobacco-induced injury in the airway Biomarkers of air pollution
5.2 Breast cancer 362 Biological mechanisms mediating reduced breast cancer risk through physical activity Treatment determined by BRCA1/2 mutation
5.3 Oesophageal cancer 374 Risk of malignant progression in Barrett oesophagus patients
5.4 Stomach cancer 383 Gastric cancer prevention
5.5 Colorectal cancer 392 Lgr5 stem cells in self-renewal and intestinal cancer Tumour-elicited inflammation and malignant progression in colorectal cancer
5.6 Liver cancer 403 Treating chronic hepatitis with antiviral drugs to prevent liver cancer
5.7 Pancreatic cancer 413 5.8 Head and neck cancers 422 5.9 Kidney cancer 436 5.10 Bladder cancer 444
5.11 Cancers of the male reproductive organs 453 5.12 Cancers of the female reproductive organs 465 5.13 Haematopoietic and lymphoid malignancies 482
Molecular detection of minimal residual disease in childhood leukaemia De novo DNA methyltransferases in normal and malignant haematopoiesis
5.14 Skin cancer 495 Evaluating sunscreen as a measure to reduce the risk of melanoma
5.15 Thyroid cancer 503 5.16 Tumours of the nervous system 511
Glioma genomics and its implications in neuro-oncology
6 Cancer control 6.1 National cancer control plans 529 6.2 Current global national cancer control
capacity 538 Cervical cancer prevention in six African countries Traditional household cooking: risks and prevention
6.3 Health systems strengthening for cancer control 547 Towards understanding the economics of cancer: priority areas for research
6.4 Research infrastructure – biobanks, cohorts, registries, and data linkage 554
6.5 Advocacy for cancer control 562 6.6 Law in cancer control 569 6.7 The global economic burden of cancer 576
Cancer control experiences Cancer control in Africa: options for a vulnerable continent 528 Cancer control in Canada: challenges and strategies in a high-income country 537 Cancer control in China: preventive policies and accessible health care 546 Cancer control in France: towards patient-centred precision medicine 553 Cancer control in India: cancer care through a four-tier system 561 Cancer control in Jordan: goals for low- and middle-income countries 568 Cancer control in Morocco: action in harmony with the socioeconomic and cultural context 575 Cancer control in Peru: “El Plan Esperanza” – the Hope Plan 584 Cancer control in Turkey: an encouraging national cancer control plan for the future 585
Contributors 596 Disclosures of interests 606 Sources 608 Subject index 619
World Cancer Report ix
The world has opened its eyes to the threat posed by cancer and other noncommunicable diseases (NCDs). Realization is growing, in global political circles and in civil society, that these diseases constitute a major obstacle to human development and well-being.
The United Nations General Assembly High-Level Meeting on the Prevention and Control of Noncommunicable Diseases, held in September 2011, marked a turning point in the awareness of political leaders and the international community of the need for urgent action to avert a worldwide crisis.
The new figures and projections of the global cancer burden presented in this edition of World Cancer Report starkly highlight the problem: the incidence of cancer has increased from 12.7 million in 2008 to 14.1 million in 2012, and this trend is projected to continue, with the number of new cases expected to rise a further 75%. This will bring the number of cancer cases close to 25 million over the next two decades. The greatest impact will unquestionably be in low- and middle-income countries, many of which are ill-equipped to cope with this escalation in the number of people with cancer.
Many developing countries find themselves in the grip of cancers from two vastly different worlds. Those associ- ated with the world of poverty, including infection-related cancers, are still common, while those associated with the world of plenty are increasingly prevalent, owing to the adoption of industrialized lifestyles, with increasing use of tobacco, consumption of alcohol and highly processed foods, and lack of physical activity.
This rising burden of cancer and other NCDs places enormous strains on the health-care systems of developing countries. Coupled with ageing populations and the spiralling costs of cancer treatment, increasing demands are placed on the health-care budgets of even the wealthiest nations. As a result, prevention is central to reducing or reversing the rise in cancer burden. The central role of prevention was acknowledged in the Political Declaration adopted at the United Nations meeting, which described it as the cornerstone of the global response.
The United Nations Political Declaration gave WHO a clear mandate to coordinate the global response to this threat along with some important time-bound responsibilities captured within the Global Action Plan for the Prevention and Control of NCDs 2013–2020. IARC’s contribution has been and will continue to be instrumen- tal in this process. Independent, robust scientific evidence is the foundation of the formulation of sound public health policies. The high-quality research produced by IARC is essential for the development of evidence-based guidelines and policy by WHO, and for the adoption of regulatory decisions by national institutions to protect the health of their populations.
This new edition of World Cancer Report represents a timely update on the state of knowledge on cancer statistics, causes, and mechanisms, and on how this knowledge can be applied for the implementation of effective, resource-appropriate strategies for cancer prevention and early detection. I am confident that World Cancer Report 2014, like the previous editions, will constitute a key reference tool that will find extensive use among scientists, public health workers, and governments in supporting the implementation of national and regional plans for cancer prevention and control.
Dr Margaret Chan
Director-General World Health Organization
Foreword
x
Cancer is costly. First and foremost there is the human cost, comprising the uncertainty and suffering that a diagnosis of cancer brings in its wake. Behind each statistic of a new cancer case is an individual face, accom- panied by the faces of family and friends drawn into this singular event. The harrowing experience of a cancer diagnosis is a truly universal one, played out in every community worldwide, every day.
In contrast, as a cancer patient moves beyond diagnosis, individual experiences diversify across the world. In fact, the future of a cancer patient depends in large part on where the person lives. In less economically developed countries, cancer is typically diagnosed at more advanced stages of disease, while access to effec- tive treatment is limited or unavailable, as is palliative care. Even within more economically developed coun- tries, there are disparities in access to care among different communities. The experiences of individual cancer patients all too frequently reflect the worst of global inequalities.
Cancer also has a societal cost; enormous human potential is lost, and treating and caring for an increasing number of cancer patients has an escalating economic impact. This too is a universal experience, but again the details differ greatly between countries. World Cancer Report 2014 reveals a cancer burden that is pro- jected to increase by about 70% worldwide in just two decades, but it is in the lowest-income countries with the least-developed cancer services that the impact will be greatest. Already, the early onset of some com- mon cancers (e.g. cervix, liver) and generally poorer survival in low- and middle-income countries mean that the burden of years of life lost to cancer in these countries is similar to that in higher-income countries. Given population growth, ageing, and the spread of risk factors, such as tobacco use, the situation will worsen in the next decades, posing a major challenge to health systems in low- and middle-income countries, so that this divide between the experiences of individual cancer patients will only broaden. Taken in isolation, this is a dark prediction.
It is time to take up the challenges posed by the markedly increasing number of cancer cases globally. The particularly heavy burden projected to fall on low- and middle-income countries makes it implausible to treat our way out of cancer; even the highest-income countries will struggle to cope with the spiralling costs of treat- ment and care. Therefore, elucidating the causes and devising effective prevention strategies are essential components of cancer control, as is the gathering of accurate data on cancer occurrence from population- based cancer registries. These approaches will complement the benefits in improved access to affordable and effective cancer treatment.
In parallel to work carried out on causes and prevention, remarkable progress has been made in understand- ing the molecular and cellular events that transform a normal functioning cell into part of a malignant growth that can kill its host. These exciting advances in basic science have ramifications that are evident throughout this edition of World Cancer Report, notably in classifying cancers, in providing new avenues for clues about their causes, in highlighting opportunities for early detection and prevention, and in laying a foundation for the development of new, targeted treatments in the clinic. As never before, there is an opportunity to bring together interdisciplinary cancer expertise so that the advances of basic science are translated into both improved treatment and more widespread prevention and early detection. This integrated and complementary approach reflects not only the duty of care to today’s cancer patients but also the duty of care to the next generation, to free as many people as possible from the threat of this disease.
Preface
World Cancer Report xi
In the highlands of Kenya among the Kalenji people, who incidentally have provided some of the greatest middle- and long-distance runners in the world, there is a saying: “We should put out the fire while it is still small.” It is this saying that greets patients arriving at the Tenwek Mission Hospital in the Western Highlands (see photo). We might adopt it as an idiom for cancer prevention. Since the middle of the last century, enormous progress has been made in identifying the causes of cancer, so that more than 50% of cases could be prevented based on current knowledge. These successes in identifying cancer causes must be com- plemented by an evaluation of the most effective interventions and an understanding of how best to support their implementation into specific health-care settings. Collectively, this knowledge provides huge potential for reducing the cancer burden; one can only imagine the interest that would follow an announcement of the availability of new cancer treat- ments able to cure 50% of all patients. Therefore, prevention must be writ large in cancer control plans if we are to defy the dark prediction of the statistics.
The International Agency for Research on Cancer (IARC) will play its part as it works for cancer prevention and control, with a particular commit-
ment to low- and middle-income countries. IARC has a primary role in promoting international collaboration, and this role, acting as a catalyst for research, is increasingly proving to be vital because national questions can only be answered by international studies. Furthermore, the interdisciplinary approach taken by IARC is bearing fruit as knowledge about mechanisms of carcinogenesis casts new light on the causes and prevention of the disease.
The personal impact of cancer should never be far from the minds of all whose careers lead them to join in efforts to reduce the burden of suffering due to cancer. At the same time, cancer professionals from all disci- plines need reliable knowledge on which to act, and the general public has the same need in order to make informed decisions. It is in this context that World Cancer Report 2014 provides its up-to-date description of the occurrence, causes, underlying mechanisms, and prevention of cancer. My hope is that it will be a catalyst for collectively meeting the challenges of cancer in a way that benefits people in an inclusive way worldwide.
Dr Christopher P. Wild
Director International Agency for Research on Cancer
xii
Research underpins the development and implementation of all measures calculated to reduce the cancer bur- den. In addressing cancer research developed over the five-year period since the previous edition of World Cancer Report was published, all the contributors to this volume faced the challenge of identifying the most pertinent developments. At the same time, an interval of five years represents a suitable time frame to reappraise the descriptive epidemiology of cancer occurrence worldwide – with this edition of World Cancer Report incor- porating GLOBOCAN 2012 data as they became available in the weeks immediately before publication – and to assess cancer occurrence in relation to the broad and rapidly expanding knowledge available for cancer control.
Cancer etiology and biology Specification of cancer incidence and mortality data with varying degrees of confidence for virtually all countries is fundamental to cancer control. These data – particularly as they relate to specific tumour types – not only establish the burden of disease as it may impinge upon public health and clinical services planning; they also indicate, in many instances, causative relationships, the impact of socioeconomic differences, and priorities that may be accorded to particular cancer control options.
In many cases, differences in cancer incidence between countries reflect the decades-old perception that cancer is a disease of affluence. But this perception is inadequate, having been displaced initially by the level of tobacco- induced lung cancer in China and some other Asian countries, and more recently as cancer rates are expected to grow due to the impact of rapid increases in the prevalence of obesity, which is not confined to high-income countries. Cancers associated with chronic infections remain a particularly important challenge in low- and mid- dle-income countries. These differing and evolving profiles of risk factors are occurring against the background of marked demographic changes, characterized in many countries by an ageing and growing population, which will see the greatest proportional increases in cancer burden falling on some of the economically poorest regions of the world. World Cancer Report illustrates that cancer is truly a global problem.
The impacts of tobacco, obesity, and infections are just part of a broad spectrum of other agents and risk factors that contribute to cancer development and that, together, influence the striking geographical heterogeneity in incidence rates. Certain of these risk factors are non-modifiable, for example race, familial genetic background, and reproductive and hormonal history. Exposure to carcinogens may result from what are often characterized as lifestyle choices, which include alcohol consumption and behaviour in relation to avoidable sun exposure. Finally, people may be exposed to carcinogens in circumstances over which they have little or no control, which is the case in relation to occupational exposures, the effects of pollution (e.g. of ambient air or water), and ex- posures resulting from the use of particular foods, drugs, or consumer products. Priorities accorded to avoiding the impact of various causative agents may be influenced by attributable risk: the proportion of total cancers for which a particular agent or circumstance played a causal role in the development. Such quantitative determina- tions may vary markedly depending on which community or country is under consideration. The overarching principle, however, is that people should not be knowingly exposed to circumstances likely to increase their risk of developing cancer.
Analytical epidemiological studies, often incorporating molecular and biological measurements made possible through the availability of, for example, archived blood and/or tissue samples, increasingly identify key biologi- cal processes whose relevance is initially indicated through experimental studies. The past five years has wit- nessed the identity of many cancer pathways being derived from whole-genome sequencing for multiple cases of each major tumour type, and from analogous comprehensive data of the “–omics” sciences, which encompass genomics, transcriptomics, proteomics, metabolomics, and the like. These data have been generated through international collaboration. Genomic and similar data provide singular insight into the nature of cancer cell devel- opment within the context of normal tissue. These data offer, for example, the prospect of improved detection of early-stage disease, but also more refined molecular classification of malignancy with relevance to descriptive and etiological epidemiology. They also reveal perturbed signalling and other alterations in cancer cells, which,
Introduction
World Cancer Report xiii
by definition, establish at least a basis for what is termed targeted therapy. Such opportunities are already being translated into clinical practice. The elucidation of biological changes that characterize cancer cells has been paralleled by observations at the cellular level to the effect that malignant tumours are inadequately understood as simply a mass of cancer cells: malignant tumours are also made up of fibrous, inflammatory, vascular, and im- munological cell populations. Any one or more of these populations may be, at particular times, critical to tumour development and hence may offer an approach to prevention or therapy.
Reducing cancer incidence and mortality Community awareness of the burden of cancer is inevitably focused on the development of improved therapies to the benefit of cancer patients, and media reports of “breakthroughs” are often the vehicle for reporting novel developments. The challenge posed by agent-specific resistance largely accounts for the reality of persistent disease despite incremental progress. This challenge highlights the certain benefits accruing from the adoption of measures to prevent cancer – benefits that accrue, however, in the longer term. For example, the impact of reduced cigarette consumption at the population level on lung cancer incidence came with a lag of some two to three decades. While smoking cessation is the most effective cancer preventive measure involving reduced exposure to proven carcinogens, cancer prevention involves a broad spectrum of initiatives, extending from vac- cination – to ameliorate or prevent entirely the impact of relevant infections – through to screening tests aimed at detection, and consequently treatment, of early-stage disease.
National planning and international collaboration have emerged as critical to effective cancer control. Along with GLOBOCAN 2012 and the availability of genomic and related data, this edition of World Cancer Report highlights the WHO Framework Convention on Tobacco Control, the first international treaty negotiated under the auspices of WHO. As a response to the major recognized cause of cancer, tobacco control measures are almost univer- sally applicable: what is effective in one country has been established as likely to be effective in most, if not all, countries. Although priorities must be accorded on a national basis, options for cancer control are increasingly available to countries and communities based on what has succeeded elsewhere. The inherent worth and benefit of collaboration across national boundaries is established for cancer control. Without doubt, national govern- ments are seeking this internationally established evidence base, developed free from vested interests, for im- plementation at the local level.
The scope and design of World Cancer Report Comprehensive texts on clinical oncology, often including sections on epidemiology, cancer biology, and public health matters, are published regularly and typically extend to thousands of pages. Likewise, annual reviews pro- vide both researchers and clinicians with comprehensive coverage of recent publications in particular disciplines or concerning specified types of cancer or advances in therapies. World Cancer Report is readily distinguished from both these types of publication with reference to its scope and design. Concerning scope, emphasis on the global burden of cancer, and the environmental, lifestyle, and biological factors that might account for that bur- den, elevates the means of cancer prevention and their implementation to singular prominence. World Cancer Report does not address clinical care and the determination of optimal therapies, notwithstanding the exciting promise of these areas. Concerning design, World Cancer Report seeks to provide authoritative assessments through several different presentational approaches, while maintaining a publication of manageable length.
We, as editors, are grateful to the contributors to World Cancer Report, who, without exception, are aware of literally hundreds of publications that could be reasonably cited in the respective chapters were it not for length constraints that the present publication imposes. Because of these limitations, and the possibility that inevitable generalizations may preclude an appreciation of complexity, World Cancer Report features “boxes” in which a further group of investigators outline how certain precise issues are being elucidated with relevance to particular chapters.
xiv
The structure of World Cancer Report 2014 is essentially in line with that adopted for earlier editions. However, World Cancer Report 2014 is distinguished from its predecessors by the inclusion of a series of “Perspectives”. Several prominent investigators have been invited to provide a personal viewpoint without the boundaries implicit in the headings of particular chapters. The perspectives offered are both distinctive and challenging, and serve to indicate the variety of issues immediately relevant to cancer control that either remain as challenges for further research or have yet to achieve their full potential by comprehensive implementation.
Cancer continues as a scourge of humankind. Increasingly, cancer is a particular burden on the populations of low- and middle-income countries. Cancer control may be achieved in large part through the insight gained from research, through detailed knowledge of how individuals and communities are affected, and through implementa- tion of policies whose efficacy is often proven by the experience of other countries or groups of countries. The inclusion in this volume of several examples of national cancer control planning further demonstrates both the specific and general experiences from which lessons can be drawn in translating research-derived evidence into practice. World Cancer Report 2014 therefore captures the dynamic state of both cancer research and cancer control worldwide with respect to what has been achieved, and what remains to be accomplished, to the benefit of the global community.
Bernard W. Stewart and Christopher P. Wild (Editors) Lyon, December 2013
Cancer affects all of humankind, but there are marked differences across local, national, and regional bound- aries, particularly when considering specific tumour types rather than cancer as a whole. Epidemiological data on incidence of cancer and deaths caused by cancer vary enormously in coverage and quality between countries and regions worldwide, ranging from complete coverage by national cancer registries to population-based regis- tries covering a part of the country, hospital-based reg- istries, or no available data at all on cancer occurrence. In the absence of data, inferences must be drawn from surrounding countries to provide the best estimate pos- sible. This edition of World Cancer Report provides data
from GLOBOCAN 2012, the most current appraisal of the distribution of cancer worldwide. The findings show that high-resource countries have the highest incidence of cancer and also provide the best services for detection, diagnosis, and treatment, as may be inferred from mortal- ity and survival data. The highest prevalence proportions of cancer also occur in these populations. The most com- mon cancers include lung, breast, prostate, and colorectal cancers. In countries in epidemiological transition, these cancers are increasingly common but incidence of stom- ach, oesophageal, and liver cancers remains high. Data from low-resource countries show that cervical cancer is still often the most common cancer among women. In low- and middle-resource countries, incidence of particu- lar tumours may be relatively low, but corresponding mor- tality data often reflect late-stage diagnosis as the norm and consequently poor clinical outcomes. Worldwide, dif- ferences in cancer incidence have been recognized for more than half a century as indicating different causes and, by inference, different opportunities for prevention. These lines of investigation have been greatly refined in recent years. Accordingly, cancer epidemiological data as now presented not only establish the burden of cancer but also underpin and very often confirm determinations of causation and opportunities for prevention, as elabo- rated in subsequent sections of this Report.
Cancer worldwide
1
16
Summary • Cancer is a major cause of
morbidity and mortality, with approximately 14 million new cases and 8 million cancer-re- lated deaths in 2012, affecting populations in all countries and all regions. These estimates correspond to age-standard- ized incidence and mortal- ity rates of 182 and 102 per 100 000, respectively.
• Among men, the five most com- mon sites of cancer diagnosed in 2012 were the lung (16.7% of the total), prostate (15.0%), colorectum (10.0%), stomach (8.5%), and liver (7.5%). Among women, the five most common incident sites of cancer were the breast (25.2% of the total), colorectum (9.2%), lung (8.7%), cervix (7.9%), and stomach (4.8%).
• Among men, lung cancer had the highest incidence (34.2 per 100 000) and prostate cancer had the second highest in- cidence (31.1 per 100 000). Among women, breast cancer had a substantially higher inci- dence (43.3 per 100 000) than any other cancer; the next high- est incidence was of colorectal cancer (14.3 per 100 000).
• Prevalence estimates for 2012 indicate that there were 8.7 mil- lion people (older than 15 years) alive who had had a cancer di- agnosed in the previous year, 22.0 million with a diagnosis in the previous 3 years, and 32.6 million with a diagnosis in the previous 5 years.
• The worldwide estimate for the number of cancers diagnosed in childhood (ages 0–14 years) in 2012 is 165 000 (95 000 in boys and 70 000 in girls).
• For all cancers combined (ex- cluding non-melanoma skin cancer) the highest incidence rates are associated with the high-income countries of North America and western Europe (together with Japan, the Re- public of Korea, Australia, and New Zealand).
• More than 60% of the world’s cancer cases occur in Africa, Asia, and Central and South America, and these regions account for about 70% of the cancer deaths.
• The distribution of cancer in world regions indicates marked, and sometimes extreme, differ- ences with respect to particu- lar tumour types. Such data are key to any understanding
1.1 1 WORLDWIDE
The global and regional burden of cancer David Forman Jacques Ferlay
Bernard W. Stewart (reviewer) Christopher P. Wild (reviewer)
of causation, and hence the development of preventive measures.
Robust statistics on cancer occur- rence and outcome are an essential prerequisite for national and region- al programmes of cancer control and for informing the cancer re- search agenda. Since 1984, IARC has regularly published estimates of the global and regional burden of cancer in terms of both incidence and mortality [1]. Since 2001, such estimates have also been made available for all major countries of the world through the GLOBOCAN project [2,3]. These estimates are based on all sources of informa- tion available for any given country
Fig. 1.1.1. Fishermen in Bahia de Kino, Mexico. On a world scale, intermediate cancer incidence rates occur in Central and South America.
Chapter 1.1 • The global and regional burden of cancer 17
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
[4] and, where possible, make use of incidence data published in the IARC series Cancer Incidence in Five Continents [5–7] and mortal- ity data made available by WHO [8]. For more details about the graphics presented in this chapter, see “A guide to the epidemiology data in World Cancer Report ”.
Global burden of cancer Incidence and mortality Results from GLOBOCAN [3] show that in 2012 there were an estimated 14.1 million new cases of cancer di- agnosed worldwide (excluding non- melanoma skin cancer) and 8.2 mil- lion estimated deaths from cancer. These estimates correspond to age-standardized incidence and mortality rates of 182 and 102 per 100 000, respectively. There were slightly more incident cases (53% of the total) and deaths (57%) among men than among women. The es- timated global annual numbers of new cases and cancer deaths for the most important types of cancer are shown in Figs 1.1.4 and 1.1.5, for both sexes combined and for men and women separately.
Among men, the five most com- mon sites of cancer diagnosed in
2012 were the lung (16.7% of the total), prostate (15.0%), colorec- tum (10.0%), stomach (8.5%), and liver (7.5%). These sites also rep- resented the most common causes of cancer death in men; the rela- tive importance of lung cancer in- creased to 23.6% of the total, fol- lowed by liver cancer (11.2%) and stomach cancer (10.1%) as the sec- ond and third most common causes of cancer death.
Among women, the five most common incident sites of cancer were the breast (25.2% of the to- tal), colorectum (9.2%), lung (8.7%), cervix (7.9%), and stomach (4.8%). These sites also represented the most common causes of cancer death in women; the relative impor- tance of breast cancer decreased to 14.7% of the total, followed by lung cancer (13.8%) as the sec- ond most common cause of cancer death. In both sexes combined, the five most common incident sites of cancers were the lung (13.0% of the total), breast (11.9%), colorectum (9.7%), prostate (7.9%), and stom- ach (6.8%); together, cancers of these five sites constitute half of the overall global cancer burden.
Estimated age-standardized in- cidence and mortality rates for the
15 most common cancers world- wide for men and women in 2012 are shown in Fig. 1.1.6. Among men, lung cancer has the highest incidence and mortality rates (34.2 and 30.0 per 100 000, respective- ly). Prostate cancer has the sec- ond highest incidence rate (31.1 per 100 000), not far below that for lung cancer. However, the mortal- ity rate for prostate cancer (7.8 per 100 000) is considerably lower than that for lung cancer. The differential between mortality and incidence for these two cancers reflects the much lower fatality rate (or improved sur- vival) of prostate cancer compared with lung cancer. Stomach, liver, and oesophageal cancers are three of the other major cancers in men that, like lung cancer, have relative- ly poor survival and hence mortality rates that are close to the incidence rates (incidence and mortality of 17.4 and 12.7 per 100 000, respec- tively, for stomach cancer, 15.3 and 14.3 per 100 000 for liver cancer, and 9.0 and 7.7 per 100 000 for oesophageal cancer). The other major cancer in men is colorectal cancer, which has an incidence rate of 20.6 per 100 000 but a substan- tially lower mortality rate of 10.0 per 100 000.
Among women, breast cancer has a considerably higher inci- dence rate (43.3 per 100 000) than any other cancer. The next highest is cancer of the colorectum (14.3 per 100 000), and then cancers of the cervix (14.0), lung (13.6), cor- pus uteri (8.2), and stomach (7.5).
Fig. 1.1.3. Times Square, New York. The four major cancers that affect populations in North America are the same as those in Europe: cancers of the prostate, breast, lung, and colorectum.
Fig. 1.1.2. A view of Paris. High-income countries, exemplified by countries in western Europe and North America, have the highest overall cancer rates in both sexes.
18
Fig. 1.1.4. Estimated world cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 19
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.5. Estimated world cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
20
However, breast cancer has a rela- tively low fatality rate, and thus al- though it has the highest mortality rate of any cancer in women (12.9 per 100 000), this is less than one third of the incidence rate and not much higher than the lung cancer mortality rate (11.1 per 100 000), which is the next highest. As in men, the stomach cancer mortality rate (5.7 per 100 000) is not far below the incidence rate, but the mortality rates for the other major cancers in women – cancers of the colorectum (6.9 per 100 000), cervix (6.8) and, especially, corpus uteri (1.8) – are substantially lower than the corre- sponding incidence rates.
Prevalence A further measure of cancer bur- den is prevalence: the number of people who have been diagnosed with cancer and are still alive at a specific time [9]. Prevalence es- timates for 2012 show that there
were 8.7 million people (older than 15 years) alive who had had a can- cer diagnosed in the previous year, 22.0 million with a diagnosis in the previous 3 years, and 32.6 million with a diagnosis in the previous 5 years [3]. Estimates of 5-year prevalence for the nine most com- mon cancer sites worldwide are shown in Fig. 1.1.7, for both sexes combined and for men and women separately. With 6.3 million survi- vors diagnosed within the previous 5 years, breast cancer is by far the most prevalent cancer, even when results for both sexes are com- bined. The second most prevalent is prostate cancer, with 3.9 million, and then colorectal cancer, with 3.5 million (1.9 million men and 1.6 million women), and lung can- cer, with 1.9 million (1.3 million men and 0.6 million women). Prevalence reflects the integration of incidence and extent of survival, and for lung cancer, because of its very poor
survival, the 5-year prevalence is very close to the annual mortality (1.6 million).
Incidence by age group Cancer has a strong relationship with age, and Fig. 1.1.8 shows the GLOBOCAN estimated inci- dence rates for all cancers com- bined (excluding non-melanoma skin cancer) by age group for men and women. Rates in the youngest age group (0–14 years) are about 10 per 100 000, increasing to 150 per 100 000 by 40–44 years and to more than 500 per 100 000 by 60– 64 years. Only in the youngest age group (0–14 years) are rates similar between the sexes. After childhood, rates are higher in women than in men until about the age of 50 years, when rates in men overtake those in women and then become sub- stantially higher from the age of 60 years. The excess in women before the age of 50 years is due
Fig. 1.1.6. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Chapter 1.1 • The global and regional burden of cancer 21
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.7. Estimated world cancer 5-year prevalence proportions by major sites, in both sexes combined, in men, and in women, 2012.
22
to the relatively earlier age at onset of cervical cancer and, especially, breast cancer compared with other major cancers. Above the age of 60 years, prostate cancer and lung cancer in men become more com- mon. The worldwide estimate for the number of cancers diagnosed in childhood (0–14 years) in 2012 is 165 000 (95 000 in boys and 70 000 in girls), and in teenagers and young adults (15–39 years), the estimate is just more than 1 million (380 000 in men and 670 000 in women). For more information on the descriptive epidemiology of cancer in these age groups, see Chapter 1.3.
Global distribution Incidence World maps of estimated age- standardized incidence rates for all cancers combined (exclud- ing non-melanoma skin cancer) in 2012 are shown in Fig. 1.1.9 for men (Fig. 1.1.9A) and women (Fig. 1.1.9B). In general, and with some exceptions, the highest in- cidence rates are associated with the high-income countries of North America and western Europe (to- gether with Japan, the Republic of Korea, Australia, and New Zealand). This is the case for both sexes, although the absolute level
of the upper rate category is higher in men than in women. Intermediate rates are seen in Central and South America, eastern Europe, and much of South-East Asia (includ- ing China), and the lowest rates are seen in much of Africa and West and South Asia (including India). There are some interesting excep- tions to this pattern. For example, Uruguay and Mongolia fall into the high-incidence category, Uruguay partly because of particularly high rates for lung cancer (and other smoking-related cancers), which are now being brought under con- trol [10], and Mongolia because of the extraordinary high rate of liver cancer resulting from the particular- ly high prevalence of infection with hepatitis B and C viruses [11].
Mortality Similar world maps are shown in Fig. 1.1.10 for age-standardized mortality rates. These also show ex- tensive international variation, but the contrasts are less marked than those for incidence. For example, there is less variation between North and South America and be- tween western and eastern Europe. In both cases, this is a result of the impact of clinical care and generally improved cancer survival in North America and western Europe, lead- ing to lower mortality rates (relative to the incidence rates) than is ob- served in South America and east- ern Europe [12]. In addition, and as described in the discussion of regional patterns of cancer below, cancers associated with a lifestyle typical of industrialized countries, including cancers of the breast, colorectum, and prostate, have a relatively good prognosis, whereas cancers of the liver, stomach, and oesophagus, which are more com- mon in low-income countries, have a significantly poorer prognosis. In general, therefore, relatively more cancer deaths are seen for a given number of incident cases in the less economically developed countries, and cancer mortality rates are not very different from those in more developed countries.
Fig. 1.1.8. Estimated world cancer incidence rates per 100 000 by 5-year age group, for all sites combined (excluding non-melanoma skin cancer), in men and women, 2012.
Chapter 1.1 • The global and regional burden of cancer 23
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.9. Global distribution of estimated age-standardized (World) cancer incidence rates (ASR) per 100 000, for all sites combined (excluding non-melanoma skin cancer), in (A) men and (B) women, 2012.
A
B
24
Fig. 1.1.10. Global distribution of estimated age-standardized (World) cancer mortality rates (ASR) per 100 000, for all sites combined (excluding non-melanoma skin cancer), in (A) men and (B) women, 2012.
A
B
Chapter 1.1 • The global and regional burden of cancer 25
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Regional patterns of cancer A regional breakdown of the global cancer incidence and mortality bur- den by continental region is provid- ed in Fig. 1.1.11. About half of the in- cidence burden occurs in Asia, and almost a half of this, or 22% of the global total, arises in China, and 7% in India. A quarter of the incidence burden occurs in Europe, and the remainder is divided between the Americas and Africa (with 1% in Oceania). The mortality proportion- al distribution shows an increase in the proportion of cancer-related deaths occurring in Asia and Africa, together with a decrease in the pro- portions occurring in the economi- cally more developed regions of Europe and North America.
Further analyses of the regional patterns are shown in Figs 1.1.12 to 1.1.39, in which, for each world region, a proportional breakdown is provided of the estimated 2012 numbers of cancer cases and deaths by sex, together with histo- grams showing the age-standard- ized incidence and mortality rates by cancer site and a chart showing the major 5-year prevalent cancers.
Europe In Europe, the incidence pattern is dominated by prostate and breast cancers, which are the most com- mon sites in men and women, re- spectively, together with lung and colorectal cancers (Fig. 1.1.12). In terms of mortality, lung cancer, because of its poor survival, is the most important cause of cancer death in men (Fig. 1.1.13). Bladder, stomach, and kidney cancers also contribute significantly to the bur- den in men, with incidence rates of more than 10 per 100 000, whereas in women, cancers of the corpus uteri and cervix also have incidence rates of more than 10 per 100 000 (Fig. 1.1.14). Breast, pros- tate, and colorectal cancers are the most prevalent; together, cancers of these three sites represent half of all the 5-year prevalent cases in Europe (Fig. 1.1.15).
North America In North America, patterns of the four major cancers are very similar to those in Europe. Prostate and breast cancer are the most common
incident sites in men and women, respectively, and lung cancer is the most common cause of cancer death in both sexes (Figs 1.1.16, 1.1.17). However, unlike in Europe, lung cancer is also relatively more important than colorectal cancer in women in terms of incidence. Also, alongside bladder and kidney can- cers, malignant melanoma, non- Hodgkin lymphoma, and leukaemia contribute significantly to the can- cer burden in men, with incidence rates of more than 10 per 100 000 (Fig. 1.1.18). This is also evident for malignant melanoma, non-Hodgkin lymphoma, corpus uteri, and thy- roid cancers in women. Compared with the European pattern, stom- ach cancer in men and cervical cancer in women are much less common, at least partially reflecting the greater heterogeneity in rates across Europe for these cancer sites. The prevalence pattern in North America is very similar to that in Europe, with prostate, breast, and colorectal cancers together accounting for half of all the 5-year prevalent cases (Fig. 1.1.19).
Fig. 1.1.11. Estimated world cancer incidence and mortality proportions by major world regions, in both sexes combined, 2012.
26
Oceania In Oceania, one again sees the European and North American pat- terns in relation to prostate and breast cancers as the most com- mon sites in men and women, re- spectively, and lung cancer as the most common cause of cancer death in men (Figs 1.1.20, 1.1.21). However, after colorectal cancer, malignant melanoma is the third most important incident cancer in both sexes, although it ranks relatively low in terms of mortality. Lung, kidney, and bladder cancers as well as non-Hodgkin lymphoma and leukaemia in men and can- cers of the lung, corpus uteri, cer- vix, and thyroid in women are the other cancers with rates of more than 10 per 100 000 (Fig. 1.1.22). Malignant melanoma makes an important contribution to cancer prevalence in Oceania, as the third most prevalent type. Otherwise, the pattern of prevalence is similar to that in Europe and North America (Fig. 1.1.23).
Latin America and the Caribbean For the region Latin America and the Caribbean, in common with Europe and North America, pros- tate and breast cancers are the most common incident sites in men and women, respectively (Fig. 1.1.24). Breast cancer is the most common cause of cancer death in women, whereas prostate and lung cancers make similar con- tributions to cancer mortality in men (Figs 1.1.25, 1.1.26). However, and unlike the situation in the more eco- nomically developed regions, cervi- cal cancer makes a major contribu- tion to the cancer burden in women, as the second most common inci- dent cancer and cause of cancer death in women. The incidence rate for cervical cancer (21.2 per 100 000) is substantially larger than that estimated for Europe or North America (11.4 and 6.6 per 100 000, respectively) (Fig. 1.1.26). Stomach cancer is also important, especially in men, where it is the fourth most common incident cancer (with lung
and colorectal cancers as second and third, respectively). Whereas the incidence rates for stomach cancer (12.8 and 7.1 per 100 000 in men and women, respectively) are of similar magnitude to those in Europe (13.2 and 6.4 per 100 000), these are much higher than those in North America (5.5 and 2.8 per 100 000). Cervical cancer also contributes to the pattern of cancer prevalence in the region, repre- senting 8.6% of all 5-year preva- lent cancers, third in importance after breast and prostate cancers (Fig. 1.1.27).
Middle East and North Africa In the Middle East and North Africa region (Figs 1.1.28, 1.1.29), breast cancer again dominates the pattern of cancer in women, representing 30% of the cancer cases. Colorectal cancer is the next most common cancer in women, but breast cancer is the only type for which the inci- dence or mortality rates (43.0 and 16.2 per 100 000, respectively) ex- ceed 10 per 100 000 (Fig. 1.1.30). Among men, lung cancer is the most important incident cancer and cause of cancer death, with incidence and mortality rates of 27.1 and 24.5 per 100 000, respec- tively (Fig. 1.1.30). Prostate, blad- der, colorectal, and liver cancers are the next most common, all with incidence rates of more than 10 per 100 000. In this region, breast cancer represents a quarter of all 5-year prevalent cancers, and whereas prostate cancer is rela- tively less prevalent compared with other world regions, bladder and thyroid cancers make important contributions to the overall pattern (Fig. 1.1.31).
Sub-Saharan Africa Sub-Saharan Africa provides sev- eral contrasts with other world re- gions. Among women, this is the only region where cervical cancer is equivalent to breast cancer in terms of incidence (each consti- tutes approximately a quarter of the total burden) and is the most
common cause of cancer death in women (23.2% of the total) (Figs 1.1.32, 1.1.33). The incidence and mortality rates for cervical can- cer are 34.8 and 22.5 per 100 000, respectively (Fig. 1.1.34), the high- est of any world region. Among men, prostate and liver cancers are the most common forms of incident cancer and causes of cancer death. Although the leading role of pros- tate cancer in the cancer incidence pattern in men is shared with most other world regions, this region also has mortality rates comparable to incidence rates. The rates of 27.9 and 20.9 per 100 000 for prostate cancer incidence and mortality, re- spectively, stand in marked contrast to those in Europe (64.0 and 11.3 per 100 000, respectively) or North America (97.2 and 9.8 per 100 000, respectively), where incidence is much higher but mortality is much lower. The importance of liver can- cer in this region should be empha- sized; it is the second most com- mon cancer in men and the third most common in women. Cervical and breast cancers in women and prostate and liver cancers in men are the only cancers with sex-spe- cific incidence or mortality rates of more than 10 per 100 000, but the high rates of Kaposi sarcoma in sub-Saharan Africa, especially in men, should also be noted. Kaposi sarcoma is the third most common cancer in men and represents 9.2% of all cancer diagnoses, with an incidence rate of 7.2 per 100 000. This reflects the very high regional level of HIV infection and associ- ated cancer sequelae before the advent of highly active antiretroviral therapy [13]. Cervical cancer and Kaposi sarcoma also make impor- tant contributions to the pattern of 5-year prevalence in the region (Fig. 1.1.35).
East and Central Asia East and Central Asia is by far the most populous world region, con- taining 57% of the global popu- lation (19% in China and 18% in India). Among men, the most com- mon cancers and causes of cancer
Chapter 1.1 • The global and regional burden of cancer 27
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
death are cancers of the lung, stomach, liver, colorectum, and oesophagus (Figs 1.1.36, 1.1.37), sites with incidence rates of more than 10 per 100 000 (Fig. 1.1.38). Although relatively high, the inci- dence and mortality rates for lung cancer in men (35.1 and 31.5 per 100 000, respectively) are still sub- stantively below those in Europe and North America. However, the
corresponding rates for stomach cancer (23.3 and 16.9 per 100 000, respectively) and liver cancer (20.7 and 19.5 per 100 000, respectively) are the highest of any world region. Among women, breast cancer is the most common incident cancer, fol- lowed by cancers of the lung, cer- vix, colorectum, and stomach. Due to the varying fatality rates of these cancers, the mortality patterns are
a little different in the region, and lung cancer is the leading cause of cancer death. As in other regions, breast cancer makes a substantial contribution to 5-year prevalence in East and Central Asia, but in con- trast to elsewhere, cancers of the stomach and lung are also impor- tant contributors to the pattern of prevalence (Fig. 1.1.39).
28
Fig. 1.1.12. Europe. Estimated cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 29
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.13. Europe. Estimated cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
30
Fig. 1.1.14. Europe. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Fig. 1.1.15. Europe. Estimated cancer 5-year prevalence proportions by major sites, in both sexes combined, 2012.
Chapter 1.1 • The global and regional burden of cancer 31
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.16. North America. Estimated cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
32
Fig. 1.1.17. North America. Estimated cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 33
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.18. North America. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Fig. 1.1.19. North America. Estimated cancer 5-year prevalence proportions by major sites, in both sexes combined, 2012.
34
Fig. 1.1.20. Oceania. Estimated cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 35
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.21. Oceania. Estimated cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
36
Fig. 1.1.22. Oceania. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Fig. 1.1.23. Oceania. Estimated cancer 5-year prevalence proportions by major sites, in both sexes combined, 2012.
Chapter 1.1 • The global and regional burden of cancer 37
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.24. Latin America and the Caribbean. Estimated cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
38
Fig. 1.1.25. Latin America and the Caribbean. Estimated cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 39
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.26. Latin America and the Caribbean. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Fig. 1.1.27. Latin America and the Caribbean. Estimated cancer 5-year prevalence proportions by major sites, in both sexes combined, 2012.
40
Fig. 1.1.28. Middle East and North Africa. Estimated cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 41
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.29. Middle East and North Africa. Estimated cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
42
Fig. 1.1.30. Middle East and North Africa. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Fig. 1.1.31. Middle East and North Africa. Estimated cancer 5-year prevalence proportions by major sites, in both sexes combined, 2012.
Chapter 1.1 • The global and regional burden of cancer 43
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.32. Sub-Saharan Africa. Estimated cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
44
Fig. 1.1.33. Sub-Saharan Africa. Estimated cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 45
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.34. Sub-Saharan Africa. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Fig. 1.1.35. Sub-Saharan Africa. Estimated cancer 5-year prevalence proportions by major sites, in both sexes combined, 2012.
46
Fig. 1.1.36. East and Central Asia. Estimated cancer incidence proportions by major sites, in both sexes combined, in men, and in women, 2012.
Chapter 1.1 • The global and regional burden of cancer 47
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Fig. 1.1.37. East and Central Asia. Estimated cancer mortality proportions by major sites, in both sexes combined, in men, and in women, 2012.
48
Fig. 1.1.38. East and Central Asia. Estimated age-standardized (World) cancer incidence and mortality rates (ASR) per 100 000, by major sites, in men and women, 2012.
Fig. 1.1.39. East and Central Asia. Estimated cancer 5-year prevalence proportions by major sites, in both sexes combined, 2012.
Chapter 1.1 • The global and regional burden of cancer 49
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
Country-specific patterns of cancer The consideration of incidence and mortality rates in specific countries selected to be representative of the world regions provides a further de- gree of focus on the international variation in the cancer burden. Figs 1.1.40 and 1.1.41 show age- standardized incidence rates in men and women, respectively, in about 2005 for all cancers combined (ex- cluding non-melanoma skin cancer) for a range of countries representa- tive of different world regions and selected because of the high qual- ity of their cancer incidence data. These data are derived from cancer registry data sets, and all of the reg- istries have met the quality criteria for publication in Cancer Incidence in Five Continents, Volume X [7].
Comparable mortality data, where available from the WHO Mortality Database [8] or elsewhere [14], for
the same countries are shown in Figs 1.1.42 and 1.1.43. Cancer in- cidence and mortality information for the same set of countries is pre- sented in the site-specific chapters later in this Report. It should be not- ed that cause-specific death certifi- cation is available for only a limited sample of the Chinese population and is not routinely undertaken in India and Uganda, and thus rele- vant mortality data are not available for India and Uganda.
Among men, incidence (Fig. 1.1.40) varies by approximately 4-fold, from 100 per 100 000 in Indian registries to more than 400 per 100 000 in the USA Surveillance, Epidemiology and End Results (SEER) Black pop- ulation. Among women (Fig. 1.1.41), incidence varies by approximately 3-fold, from just more than 100 per 100 000 in Indian registries to about 300 per 100 000 in the USA SEER White population and in Denmark.
For both sexes there is, within this group of countries, a gradation in all-cancer incidence rates between the extremes and, especially in men, the higher rates are in eco- nomically more developed popu- lations. Among women, while the association with economic devel- opment is also apparent, the rates in Japanese registries are at the lower end of the spectrum (170 per 100 000), whereas those in Uganda (212 per 100 000) are relatively high. As observed in the world maps, the between-country variation in mortal- ity (Figs 1.1.42, 1.1.43) is less pro- nounced than that for incidence and, in both sexes, is less than 2-fold. Thus, in men (Fig. 1.1.42) the vari- ation extends from 99 per 100 000 in Costa Rica to 203 per 100 000 in Slovakia, whereas in women (Fig. 1.1.43) it extends from 69 per 100 000 in Japan to 121 per 100 000 in Denmark.
Fig. 1.1.40. Age-standardized (World) cancer incidence rates per 100 000 in selected cancer registry populations, for all cancers combined (excluding non-melanoma skin cancer) in men, 2003–2007.
Fig. 1.1.41. Age-standardized (World) cancer incidence rates per 100 000 in selected cancer registry populations, for all cancers combined (excluding non-melanoma skin cancer) in women, 2003–2007.
50
Data from the same populations and the same sources as used in Figs 1.1.40 to 1.1.43 are present- ed in terms of trends over time in Figs 1.1.44 and 1.1.45 (incidence) and Figs 1.1.46 and 1.1.47 (mortali- ty) for all cancers combined (exclud- ing non-melanoma skin cancer). It is difficult to discern any general consistent trends in incidence from country to country, although, with few exceptions (e.g. the decline in incidence among men in Chinese registries), rates have tended to in- crease from the 1980s to the pres- ent. In some cases, for example in Slovakia, Colombia, and Denmark, this has been a fairly constant in- crease over the entire time period, whereas elsewhere, for example in Japan and the USA, an initial period of increase has been followed by a drop in rates. Some of the observed effects can be readily explained. For example, the peak in the ear- ly 1990s among men in the USA
(Black and White) and Australia is explained by the rapid increase in the number of prostate cancer di- agnoses after the introduction of testing for prostate-specific antigen [15]. Changes in the prevalence of smoking and hence, after a suit- able lag period, in the incidence of lung cancer and other smoking- associated cancers, will also be important determinants of all-can- cer incidence and mortality rates [16]. However, other trends require detailed consideration of country- specific patterns by type of cancer.
The mortality trends (Figs 1.1.46, 1.1.47) are more consistent in that, in both sexes, most populations have shown a substantive decline in age-standardized mortality, with the exception of China (although the Chinese data are for a much shorter time series). The mortality decline is evident in countries such as Denmark and Costa Rica, which have, respectively, relatively high
and low absolute levels of mortality. Such mortality decreases against a background of increase in incidence are indications of improved survival as a result of improved therapy and/ or earlier detection (including as a consequence of screening).
Nearly all cancer types show important variations between re- gions, subregions, and countries in terms of both contemporary in- cidence rates and their trends over time. The broad overview present- ed in this chapter inevitably masks some of these geographical asso- ciations. For example, the consid- eration given above, at the regional level, to East and Central Asia does not identify the extraordinarily high rates of thyroid cancer among women in the Republic of Korea, estimated as 89 per 100 000 – the most common cancer among wom- en in the country, and the highest rate of any country in the world. It is currently unclear to what extent this
Fig. 1.1.42. Age-standardized (World) cancer mortality rates per 100 000 by year in selected countries, for all cancers combined (excluding non-melanoma skin cancer) in men, 2003–2007.
Fig. 1.1.43. Age-standardized (World) cancer mortality rates per 100 000 by year in selected countries, for all cancers combined (excluding non-melanoma skin cancer) in women, 2003–2007.
Chapter 1.1 • The global and regional burden of cancer 51
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
epidemic is driven by improved di- agnostic technology [17]. Similarly, the estimated European incidence rate of cervical cancer (11.4 per 100 000) obscures the internal vari- ation, ranging from 29 per 100 000 in Romania to 4 per 100 000 in Switzerland, due primarily to the impact of cytological screening programmes [18]. Likewise, in sub- Saharan Africa, the average inci- dence rate for oesophageal cancer in men is 6 per 100 000, ranging from 25 per 100 000 in Uganda and Malawi to less than 1 per 100 000 in Guinea and Nigeria. The reason for the variation in this case is little understood, which should act as a stimulus for research [19].
Conclusions Monitoring geographical variation in disease prevalence and search- ing for an understanding of the
underlying reasons is frequently an important starting point for cancer epidemiology [20]. The un- derstanding of the causes of any specific cancer is not complete without an explanation of the geo- graphical distribution, and cancer is especially complex, given the diversity in cancer types, with each type often having its own set of risk factors and giving rise to a unique geographical pattern. Such variations have significant implica- tions for tailoring cancer control to region- or country-specific priori- ties [21]. This chapter has focused primarily on the current overall bur- den of cancer globally and in ma- jor world regions and serves as a backdrop to the consideration of specific types in later chapters of this Report. However, with approxi- mately 14 million new cases and 8 million cancer-related deaths per year, cancer must be viewed as a
major cause of morbidity and mor- tality that affects populations in all countries and all regions.
While there are some common- alities, such as breast cancer being the major type of cancer among women in all world regions, there are also important differences, and the variation in the relative impor- tance of cervical cancer across dif- ferent regions is one such example. Some of these differences relate to the level of economic development, for example the higher incidence of cancers associated with infections in less-developed economies [22], and the associations with develop- ment are considered in more detail in Chapter 1.2. The importance of genetic variation as a cause of between-population differences in risk, such as for prostate cancer, is also recognized.
At a more general level, under- standing the pattern of cancer in a
Fig. 1.1.44. Age-standardized (World) cancer incidence rates per 100 000 by year in selected populations, for all cancers combined (excluding non-melanoma skin cancer) in men, circa 1975–2012.
Fig. 1.1.45. Age-standardized (World) cancer incidence rates per 100 000 by year in selected populations, for all cancers combined (excluding non-melanoma skin cancer) in women, circa 1975–2012.
52
Fig. 1.1.46. Age-standardized (World) cancer mortality rates per 100 000 by year in selected countries, for all cancers combined (excluding non-melanoma skin cancer) in men, circa 1975–2012.
Fig. 1.1.47. Age–standardized (World) cancer mortality rates per 100 000 by year in selected countries, for all cancers combined (excluding non-melanoma skin cancer) in women, circa 1975–2012.
population is an essential prereq- uisite for planning any programme for cancer control. It is impossi- ble to develop such plans without knowing which types of cancer are common and whether cancers are changing in incidence. Likewise, the impact of any control mea- sures cannot be evaluated without
monitoring the effect on cancer in- cidence or mortality in the years, or sometimes decades, after an inter- vention. To generate the relevant incidence data, as used extensively in this chapter, the work of cancer registries throughout the world is vi- tal [23] and needs to be supported as part of any cancer control plans,
especially in less-developed re- gions, where cancer registries are often absent or under-resourced [24].
The authors thank Joannie Lortet-Tieulent and Mathieu Laversanne for their assistance in producing the figures.
Chapter 1.1 • The global and regional burden of cancer 53
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .1
References
1. Parkin DM, Stjernswärd J, Muir CS (1984). Estimates of the worldwide frequency of twelve major cancers. Bull World Health Organ, 62:163–182. PMID:6610488
2. Ferlay J, Bray F, Pisani P, Parkin DM (2001). GLOBOCAN 2000: Cancer Incidence, Mortality and Prevalence Worldwide. IARC Cancer Base No. 5. Lyon: IARC.
3. Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
4. Ferlay J, Shin HR, Bray F et al. (2010). Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer, 127:2893–2917. http://dx.doi.org/10.1002/ ijc.25516 PMID:21351269
5. Parkin DM, Ferlay J, Curado MP et al. (2010). Fifty years of cancer incidence: CI5 I-IX. Int J Cancer, 127:2918–2927. http://dx.doi.org/10.1002/ijc.25517 PMID: 21351270
6. Ferlay J, Parkin DM, Curado MP et al. (2010). Cancer Incidence in Five Continents, Volumes I to IX: IARC Cancer Base No. 9 [Internet]. Lyon: IARC. Available at http://ci5.iarc.fr.
7. Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
8. WHO Mortality Database. Available at http://w w w.who.int /healthinfo/statistics/ mortality _rawdata/en/index.html.
9. Bray F, Ren JS, Masuyer E, Ferlay J (2013). Global estimates of cancer preva- lence for 27 sites in the adult population in 2008. Int J Cancer, 132:1133–1145. http://dx.doi.org/10.1002/ijc.27711 PMID: 22752881
10. Abascal W, Esteves E, Goja B et al. (2012). Tobacco control campaign in Uruguay: a population-based trend analy- sis. Lancet, 380:1575–1582. http://dx.doi. o r g /10.1016 / S 014 0 - 673 6 (12) 6 0 8 2 6 - 5 PMID:22981904
11. Dondog B, Lise M, Dondov O et al. (2011). Hepatitis B and C virus infections in hepatocellular carcinoma and cirrhosis in Mongolia. Eur J Cancer Prev, 20:33–39. ht t p: //d x.d o i.o r g /10.10 97/C E J.0 b 013 e 32833f0c8e PMID:21166097
12. Edwards BK, Brown ML, Wingo PA et al. (2005). Annual report to the nation on the status of cancer, 1975–2002, featuring population-based trends in cancer treat- ment. J Natl Cancer Inst, 97:1407–1427. PMID:16204691
13. Geng EH, Hunt PW, Diero LO et al. (2011). Trends in the clinical characteristics of HIV-infected patients initiating antiretrovi- ral therapy in Kenya, Uganda and Tanzania between 2002 and 2009. J Int AIDS Soc, 14:46. http://dx.doi.org/10.1186/1758-2652- 14-46 PMID:21955541
14. National Center for Health Statistics, Cen- ters for Disease Control and Prevention. Available at http://www.cdc.gov/nchs/.
15. Center MM, Jemal A, Lortet-Tieulent J et al. (2012). International variation in prostate cancer incidence and mortal- ity rates. Eur Urol, 61:1079–1092. http:// dx.doi.org/10.1016/j.eururo.2012.02.054 PMID:22424666
16. Jha P, Ramasundarahettige C, Landsman V et al. (2013). 21st-century hazards of smoking and benefits of cessation in the United States. N Engl J Med, 368:341–350. http://dx.doi.org/10.1056/NEJMsa1211128 PMID:23343063
17. Pellegriti G, Frasca F, Regalbuto C et al. (2013). Worldwide increasing in- cidence of thyroid cancer: update on epidemiology and risk factors. J Cancer Epidemiol, 2013:965212. http://dx.doi. org/10.1155/2013/965212 PMID:23737785
18. Anttila A, Ronco G; Working Group on the Registration and Monitoring of Cervical Cancer Screening Programmes in the European Union; within the European Network for Information on Cancer (EUNICE) (2009). Description of the nation- al situation of cervical cancer screening in the member states of the European Union. Eur J Cancer, 45:2685–2708. http://dx. doi.org/10.1016/j.ejca.2009.07.017 PMID: 19744852
19. Hendricks D, Parker MI (2002). Oesoph- ageal cancer in Africa. IUBMB Life, 53:263–268. http://dx.doi.org/10.1080/152 16540212643 PMID:12121007
20. Doll R (1967). Prevention of Cancer: Pointers from Epidemiology. London: Nuffield Provincial Hospitals Trust.
21. Wild CP (2012). The role of cancer re- search in noncommunicable disease con- trol. J Natl Cancer Inst, 104:1051–1058. h t t p: //d x . d o i . o r g /10 .10 9 3 / j n c i /d j s 2 6 2 PMID:22781435
22. de Martel C, Ferlay J, Franceschi S et al. (2012). Global burden of cancers attribut- able to infections in 2008: a review and synthetic analysis. Lancet Oncol, 13:607– 615. http://dx.doi.org/10.1016/S1470-2045 (12)70137-7 PMID:22575588
23. Brewster DH, Coebergh JW, Storm HH (2005). Population-based cancer regis- tries: the invisible key to cancer control. Lancet Oncol, 6:193–195. http://dx.doi. o r g /10 .1016 / S1470 - 2 0 4 5 (0 5)70 071-1 PMID:15811615
24. Global Initiative for Cancer Registry De- velopment in Low- and Middle-Income Countries. Available at http://gicr.iarc.fr/
54
Summary • From a global perspective, the
cancer burden for each coun- try, and the most common tu- mour types, are associated with the value of the Human Development Index (HDI) for that country. Transitions to- wards higher levels of human development have the effect of increasing the burden of cancer overall, and of specific types.
• Cancers of the lung, breast, prostate, and colorectum are the major incident cancers in countries with high or very high HDI. In countries with low or medium HDI, cancers of the colorectum, breast, and lung have become rather more fre- quent, although there remains an excess of poverty- and infection-related cancers, no- tably cancers of the stomach, liver, cervix, and oesophagus.
• In countries in transition, de- clines in the incidence of cer- vical cancer tend to be offset by concomitant increases in the incidence of female breast cancer. In countries with low or medium HDI, incidence rates of colorectal cancer increase markedly as HDI rises in a giv- en country.
• The predicted global cancer burden is expected to exceed 20 million new cancer cases annually by 2025, compared with the estimated 14.1 million new cases worldwide in 2012.
This chapter reviews the evolution of the cancer burden worldwide in relation to transitions in human development. Societal, economic, and lifestyle changes in a rapidly
1.2 1 WORLDWIDE
Transitions in human development and the global cancer burden Freddie Bray Bernard W. Stewart (reviewer)
Christopher P. Wild (reviewer)
globalizing world continue to have profound effects on the scale and profile of the cancer burden and the need for tailored and effective strat- egies for cancer control and pre- vention. The increasing magnitude of the cancer problem is partly the consequence of population growth and ageing. Such demographic tran- sitions mean that by 2030, well over 20 million new cancer cases will be diagnosed every year. The great- est impact will unquestionably be in countries presently in the midst of
Fig. 1.2.1. In India, an information technology business park stands where agricultural fields previously existed, indicative of the transition towards a higher Human Development Index rating.
Chapter 1.2 • Transitions in human development and the global cancer burden 55
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .2
this developmental transition, many of which are ill-equipped to deal with the escalating numbers of cancer patients expected over the next decades.
As countries transition to higher levels of human development, the underlying populations tend to in- creasingly adopt behavioural and lifestyle habits that have become conventional in prosperous and in- dustrialized countries. A changing prevalence and distribution of sev- eral reproductive, dietary, and hor- monal risk factors has the effect of increasing the risk at the population level of certain cancers associated with affluence; these include female breast cancer, prostate cancer, and colorectal cancer in both sexes. The net effect of this shift to life- styles typical of industrialized coun- tries – occurring largely in countries traditionally classified as “develop- ing” – is a steady increase in the overall incidence rates of cancer, as well as a change in the spec- trum of the most common cancers towards those observed in most of the highly developed industrialized countries.
This chapter explores the link between developmental trends and these dual aspects of cancer transi- tion using the Human Development Index, an indicator introduced through the United Nations Devel- opment Programme [1]. The geo- graphical and temporal patterns of cancer-specific incidence, mortal- ity, and prevalence are examined at the national, regional, and global level. The combined demographic and epidemiological impact of a rising and transforming cancer bur- den in the many countries undergo- ing transition has, of course, major implications for public health and clinical services planning. Some of the main characteristics of this global phenomenon emphasize the need for an increasing focus on the cancers associated with afflu- ence, which invariably rank within the “top five” most frequent cancers presently observed in high-income countries.
Epidemiological transitions, noncommunicable diseases, and cancer Omran’s theory of epidemiological transition focused on how changing health and disease patterns interact with complex demographic, social, and economic determinants [2]. Omran described how, in the third stage of transition, “pandemics of in- fection” are superseded by “degen- erative and man-made diseases” as the major causes of morbidity and mortality. Thus, there is some anal- ogy with the rather recent documen- tation of the increasing burden of noncommunicable diseases, which are displacing infection-related dis- eases as major causes of morbidity and mortality in many parts of the world. Noncommunicable diseases are the foremost cause of total mor- tality in the world today, responsi- ble for two thirds of the estimated 57 million deaths worldwide in 2008 [3]. This shift in disease burden has been gaining political momen- tum and, subsequent to the United Nations Resolution on noncommu- nicable diseases in 2011, govern- ments have approved a global mon- itoring framework and an updated Global Action Plan 2013–2020 for the prevention and control of
noncommunicable diseases, aimed at tackling the four major diseases – cardiovascular disease, diabetes, chronic respiratory disease, and cancer – at the national, regional, and global level.
That cancer has emerged as such an important cause of death in human populations is, and remains, striking [4]. Yet in compilations of global mortality by cause, the com- mon practice of distinguishing be- tween neoplasms according to the major anatomical sites has tended to obscure cancer’s importance as a leading cause of death, in some instances almost to the point of its exclusion. As an example, WHO statistics for 2011 place ischaemic heart disease, stroke, lower respi- ratory infections, chronic obstruc- tive pulmonary disease, diarrhoea, and HIV/AIDS above lung cancer as leading causes of death worldwide [5]. However, when presented as a single entity, the 7.9 million deaths due to cancer outrank deaths from every other major cause of death, as estimated, in that year (Fig. 1.2.2).
Countries in rapid societal, eco- nomic, and lifestyle transition are journeying towards an ever-greater impact of cancer. A global view of cancer patterns and trends in
Fig. 1.2.2. The 10 leading causes of death worldwide in 2011. COPD, chronic obstruc- tive pulmonary disease.
56
relation to national level of develop- ment becomes a vital consideration as a matter subject to understand- ing, prediction, and planning in re- lation to anticipated changes in the cancer burden and the spectrum of major cancer causes.
The Human Development Index: a marker of growth and development Before studying cancer patterns and trends against socioeconomic markers, it is important to consider what constitutes human develop- ment, how it may be measured, and how it is changing with time. There is no automatic link between eco- nomic growth and human progress; thus, policy issues concern the ex- act process through which growth translates, or fails to translate, into human development under differ- ent developmental conditions. The Human Development Index (HDI) is a summary measure developed by the United Nations Development Programme and first published in 1990 [1]. For the 2012 estimates, the HDI is a composite index of three basic dimensions of human development – a long and healthy life, education, and a decent stan- dard of living – combined into a unit-free index with a value between 0 and 1. The first dimension is mea- sured by life expectancy at birth. This component is calculated using a minimum value of 20 years and a maximum value of 83.57 years; these are the observed extremes of this determination for countries over the period 1980–2012. Access to knowledge, the educational com- ponent, is indicated by the average duration of schooling that has been provided to adults aged 25 years and the expected years of school- ing for children of school-entry age. The third component, a decent standard of living, is measured by gross national income per capita (in purchasing power parity in United States dollars).
Using country-specific HDI val- ues, one can go beyond the histori- cal construct of characterizing the
dichotomy between a “developed” and a “developing” country or area (Fig. 1.2.3A). A two-level construct based on HDI estimates for 2012 is shown in Fig. 1.2.3B. This still rep- resents a rather crude dichotomy; therefore, to provide greater dis- crimination, four levels of HDI are used. The 2012 HDI is divided into quartiles, indicating “very high”,
“high”, “medium”, and “low” levels of human development. The global breakdown of countries according to the HDI quartiles is shown in Fig. 1.2.3C. The low HDI category is concentrated within sub-Saharan Africa, although several African countries have transitioned to the medium HDI category, as have many Asian countries, including the
Fig. 1.2.3. Global map of development level of individual countries. (A) Historical approach: developing and developed countries. (B) Two levels of Human Development Index (HDI): countries with high or very high HDI versus those with low or medium HDI, 2012. (C) Four levels of HDI: countries with low, medium, high, and very high HDI, 2012.
Chapter 1.2 • Transitions in human development and the global cancer burden 57
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .2
very densely populated countries of India and China. Many countries in Latin America and Central Asia now lie within the high HDI band.
Without exception, all coun- tries worldwide have a higher HDI in 2012 than they had in 2000 (Fig. 1.2.4). Most countries experi- enced steady rises in the HDI deter- mination over time. Highlighted are temporal patterns in eight countries that represent a broad spectrum of HDI values. The rates of change are such that a HDI of 0.4 was reached in Uganda 24 years after China, a value of 0.7 was attained in China 22 years after Argentina, and a val- ue of 0.8 was reached in Argentina 30 years after Norway. In these ex- amples, two to three decades were required to transition from one HDI category to the next.
Global cancer statistics by human development level in 2012 To examine cancer patterns in 2012 by levels of socioeconomic development, the country-specific GLOBOCAN estimates from IARC [6] were linked to the correspond- ing HDI, as estimated for 2012 by the United Nations Development Programme [7]. Methodological details of the data sources and methods of estimation used to com- pile global estimates of incidence, mortality, and prevalence are well documented, alongside more de- tailed presentation of these results [6,8–10]; the methods are largely dependent on the availability and the accuracy of the data sources locally. Generally, there is a pauci- ty of high-quality cancer incidence and mortality data in areas with low and medium HDI. For exam- ple, there are national or regional population-based cancer registries in slightly more than one third of the 184 countries examined here; less than 20 of these systems are in countries with low or medium HDI, whereas there are no vital statistics available in more than 80 countries, predominantly those with low or medium HDI.
Cancer burden by two-level Human Development Index The most common cancers in terms of new cases and deaths accord- ing to two levels of HDI – high and very high HDI versus low and me- dium HDI – are shown in Fig. 1.2.5 for 2012. After lung cancer, breast, prostate, and colorectal cancers are the most frequent cancers in countries with high or very high HDI. Whereas colorectal, breast, and lung cancers have also be- come rather more frequent in coun- tries with low or medium HDI, there remains a large excess of poverty- and infection-related cancers in these countries, notably cancers of the stomach, liver, cervix, and oesophagus.
Lung cancer is the major cause of cancer death in both HDI areas, and colorectal cancer ranks a clear second in countries with high or very high HDI. Liver and stomach can- cers remain as major causes of can- cer death in areas with low or medi- um HDI, a result of a high incidence of these cancers in these areas as well as an extremely poor prognosis for patients after diagnosis.
The extent to which cancers more strongly associated with life- styles typical of industrialized coun- tries (colorectal, breast, and pros- tate cancers combined) dominate the distribution of tumour types compared with those cancers wholly or partly related to infectious causes (cervical, liver, and stomach cancers
Fig. 1.2.4. Trends in Human Development Index (HDI) in 1980–2012, colour-coded by HDI category: red, low; orange, medium; turquoise, high; blue, very high.
58
and Kaposi sarcoma combined) is captured in Fig. 1.2.6. Colorectal, breast, and prostate cancers con- stitute more than one third of the 7.9 million new cancers annually in countries with high or very high HDI, whereas cervical, liver, and stom- ach cancers and Kaposi sarcoma in combination constitute less than 10% of the total cancer burden in these areas. Conversely, in coun- tries with low or medium HDI, infec- tion-related cancers form a greater proportion of the 6.2 million esti- mated new cases annually, with one quarter due to infection – a higher relative proportion than the one fifth due to cancers of the colorectum, breast, and prostate.
Cancer burden by four-level Human Development Index A closer examination of the five most common cancers contributing
to incidence, mortality, and 5-year prevalence by sex and by four lev- els of HDI is depicted in Fig. 1.2.7. Overall, in countries with high or very high HDI, the same five neo- plasms rank as the five most fre- quently diagnosed cancers: lung and stomach cancers and the so- called cancers of affluence, breast, prostate, and colorectal cancers. Lung cancer occupies a high rank in terms of incidence and is the most frequent cause of cancer death in all HDI categories except low HDI. The disease has a lesser rank in terms of prevalence, how- ever, because of poor survival after diagnosis.
Incidence rates of breast, pros- tate, and colorectal cancers in 2012 are a function of level of develop- ment, and risk of developing these cancers increases as countries transition to higher HDI levels, while
risk of mortality varies consider- ably less, because of proportion- ately higher case fatality in coun- tries with lower HDI (Fig. 1.2.8). For female breast cancer, there is a gradient in the magnitude of the cumulative incidence with HDI level (Fig. 1.2.8A), with cumulative risks of 2.7%, 4.9%, and 8.4% es- timated in 2012 in countries with medium, high, and very high HDI, respectively; cumulative risk is also relatively high in areas with low HDI (3.4%). However, this gradient is much less consistent for corre- sponding mortality; the estimates of 1.6% and 1.5% in areas with high and very high HDI, respectively, are elevated compared with those for countries with medium HDI. Risk of breast cancer death is highest in countries with low HDI (1.8%), due to incidence combined with a pre- dominantly poorer survival prospect
Fig. 1.2.5. Total burden of the 27 most common cancers in 2012: (A) incidence and (B) mortality according to two-level Human Development Index (HDI).
Chapter 1.2 • Transitions in human development and the global cancer burden 59
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .2
after diagnosis in these areas. Despite affecting women almost exclusively, breast cancer (close to 1.7 million new cases worldwide in 2012) is among the five most com- mon incident and prevalent cancers in each of the four HDI categories, and it is also the leading cause of cancer death in women in countries with low HDI (Fig. 1.2.7).
Similarly, prostate cancer also shows rapidly diverging incidence rates at medium to very high HDI, with lifetime risk of diagnosis rang- ing from 0.7% to 9.3% in coun- tries with medium to very high HDI (Fig. 1.2.8B). As with breast cancer, the cumulative incidence of prostate cancer is higher in countries with low HDI than in those with medium HDI, while the corresponding risk of death in areas with low HDI (1.1%) is comparable to the risk estimated
in countries with high HDI (1.3%), which is elevated relative to coun- tries with medium or very high HDI.
Perhaps the clearest gradient of increasing risk of cancer with hu- man development is for colorectal cancer, with almost a 2-fold incre- ment in cumulative incidence in both sexes combined: from 0.6% in countries with low HDI to 1.3%, 2.1%, and 3.6% in countries with medium, high, and very high HDI, respectively. A similar gradient emerges for mortality, although the estimated risk of death from colorectal cancer is close to equal in countries with high and very high HDI. Colorectal cancer now ranks within the five most frequent cancers in countries with medium, high, or very high HDI, is the sec- ond most common cause of cancer death in countries with high and
very high HDI, and ranks among the top five prevalent cancers in all four HDI areas (Fig. 1.2.7).
Stomach cancer appears with- in the top five cancers in term of both incidence and mortality in all four HDI categories (Fig. 1.2.7). Whereas stomach cancer is the fifth most frequent cancer in coun- tries with high or very high HDI, it is the third most common cancer in countries with medium HDI. Liver cancer ranks above stomach can- cer in countries with medium HDI, surpassed only by lung cancer, with the impact of cancers common in China influencing the cancer pro- file in this HDI category. Cervical cancer is the second most common cause of cancer incidence and mor- tality burden in countries with low HDI, where it is the cause of almost one in five of the cancer deaths es- timated among women.
Evidence of cancer transitions: some temporal examples Colorectal cancer as a marker of development Fig. 1.2.9 shows trends in colo- rectal cancer incidence rates plot- ted concomitantly with HDI levels for men in selected countries with high-quality population-based can- cer registries. Across most coun- tries, increasing incidence rates have paralleled increases in HDI. However, the magnitude of the rates does not necessarily corre- spond with HDI levels attained by a given year: for example, rates are much lower in Colombia than in China in recent years, when their respective HDI levels were broadly similar. HDI has also increased in some countries where increases in colorectal cancer incidence are not evident, as is the case in Uganda. Finally, it is notable that rates tend to eventually stabilize and decline in countries that have attained very high HDI, as can be observed in the USA from 1985, Japan from the early 1990s, and Australia from 1995.
Fig. 1.2.6. Percentage of cancers related to industrialized lifestyles (colorectal cancer, female breast cancer, and prostate cancer), infection-related cancers (cervical, liver, and stomach cancers and Kaposi sarcoma), and all other cancers (A) in countries with high or very high Human Development Index (HDI) and (B) in countries with low or medium HDI, 2012.
60
Fig. 1.2.7. The five most frequent cancers in terms of incident cases, cancer deaths, and 5-year cancer prevalence in 2012 (ages 15 years and older), according to four-level Human Development Index (HDI).
Chapter 1.2 • Transitions in human development and the global cancer burden 61
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .2
The rise of cancer in women: breast cancer compared with cervical cancer The evolution of cancers in women shows a consistent and very striking pattern during the epidemiological transition, with rapid declines in the incidence of cervical cancer offset by concomitant increases in the inci- dence of breast cancer (Fig. 1.2.10). The year in which the two cancers are equally common, with one trending up and the other down, is a marker of the extent of transition in a given country; this time is in the distant past for Australia, Spain, and the USA and is much more recent for Colombia, Costa Rica, and India. However, the picture in Uganda is markedly different, with rates of both cancers still rising, and cervical cancer incidence rates twice those of breast cancer. This is in major
contrast to the 10–20-fold variations in breast and cervical cancer inci- dence rates in countries with very high HDI across Europe, Oceania, and North America.
Regional cancer profiles and human development Chapter 1.1 includes a regional and inter-regional analysis of the geo- graphical variations in the cancer burden according to continent. To link such variations to level of hu- man development, the ecological comparisons of the levels of HDI in the countries of Africa and Latin America and the corresponding age-standardized rates of cervical, breast, and prostate cancer are pre- sented in Fig. 1.2.11.
In Africa, the HDI varies from about 0.3 in the Democratic Re- public of the Congo, Niger, and
Mozambique to about 0.7 in Botswana and Gabon, with the highest values in the North African countries of Algeria, the Libyan Arab Jamahiriya, and Tunisia. Rates of cervical cancer tend to be very high in countries with the lowest HDI (< 0.4), including Mali, Malawi, Mozambique, and Zimbabwe, yet rates are low in Niger, the country with the lowest HDI ranking world- wide in 2012, as well as among the highest HDI ranked countries on the continent, in North Africa.
For breast cancer, the situa- tion tends to be reversed, with the highest rates observed where HDI is high and cervical cancer rates are low, as in Algeria and Egypt, or intermediate, as in Ethiopia and Nigeria. Yet breast cancer incidence rates are relatively low in countries with elevated cervical cancer rates, such as Malawi and Mozambique, whereas in Botswana breast cancer rates are low and cervical cancer rates intermediate, contrasting with the country’s relatively high HDI. Of course, this observation may partly relate to heterogeneity in the preva- lence and distribution of risk factors for breast cancer within these coun- tries, as well as the considerable lag between average changes in human development in the national popu- lation and anticipated increases in breast cancer incidence, linked to the etiological factors and the natu- ral history of the disease.
For prostate cancer, rates tend to be less driven by HDI levels, with rates elevated in countries with low to intermediate HDI, such as the United Republic of Tanzania, Uganda, Zimbabwe, Guinea, and the Republic of the Congo, but also in countries with higher HDI, such as South Africa; rates are consistently low in North Africa.
In Latin America, the HDI does not vary as much as on the African continent, from about 0.6, as in Guatemala and Nicaragua, to just over 0.8, as in Chile and Argentina. However, some of the disparities in burden are in line with cancer tran- sitions, with relatively low rates of cervical cancer, and relatively high
Fig. 1.2.8. Cumulative risk of (A) female breast cancer, (B) prostate cancer, and (C) colorectal cancer incidence and mortality (ages 0–74 years) in 2012, according to four-level Human Development Index (HDI).
62
Fig. 1.2.9. Trends in age-standardized (World) incidence rates (ASR) for colon cancer in men in 1978–2010 in 12 countries versus concomitant trends in Human Development Index (HDI) in 1980–2012. Incidence is smoothed using Loess regression.
Chapter 1.2 • Transitions in human development and the global cancer burden 63
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .2
Fig. 1.2.10. Trends in age-standardized (World) incidence rates (ASR) for breast cancer versus cervical cancer in 1978–2010 in 12 countries. Incidence is smoothed using Loess regression.
64
rates of breast cancer, in countries with high HDI, such as Argentina, Uruguay, and Brazil, and the oppo- site situation in countries with lower HDI, such as Nicaragua and Bolivia. Yet for other Latin American coun- tries such simple relations appear not to hold: the rates of both cancers are relatively low in Chile, while for cervical cancer, rates in Peru are twice those of neighbouring Brazil, despite approximately similar HDI levels in 2012. Prostate cancer inci- dence is high in countries with higher HDI such as Brazil and Costa Rica, but in countries with analogous de- velopment levels such as Argentina and Peru, rates are intermediate to low in relative terms, respectively.
This description of the cancer burden for the three cancers by re- gion and level of human develop- ment provides some evidence of cancer transition at the continental level but reveals the complexity lim- iting a broad interpretation, given the counter-intuitive examples provided above, and indeed the possible lag between HDI transitions and cancer risk. Certainly, the extent and modal- ities of early detection and screening vary from country to country and be- tween and within the two regions ex- amined. Increasing levels of screen- ing at the population level are linked to human development, and where precursor lesions are not the target of the intervention, such interven- tions have the potential to artificially increase the incidence burden of fe- male breast cancer due to mammo- graphic screening programmes and, particularly, prostate cancer due to prostate-specific antigen testing of asymptomatic individuals.
The extent to which cytological screening of the cervix has been introduced is partly linked to devel- opment levels and national cancer control policies and will have had an impact on the rates of cervical cancer in some countries, while other factors may explain some of the variation, including rates of high-risk human papillomavirus in- fection at the population level, as influenced by religious and cultural practices in given countries.
Fig. 1.2.11. Regional maps of Human Development Index (HDI) value versus age- standardized (World) incidence rates (ASR) of cervical, breast, and prostate cancer, (A) in Africa and (B) in Latin America, 2012.
Chapter 1.2 • Transitions in human development and the global cancer burden 65
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .2
There are regional determinants for prostate cancer. Incidence rates are influenced by the diagnosis of latent cancers mainly in countries with higher HDI, while the distri- bution of mortality rates is less affected by the effects of early di- agnosis. Evidently the elevated mortality rates in several sub-Saha- ran countries, such as Uganda and Zimbabwe, suggest that prostate cancers commonly diagnosed in these populations are developed as a result of largely unknown inherited or environmental factors linked to the underlying risk of invasive dis- ease, rather than being determined by interventions aimed at revealing evidence of asymptomatic cancers.
As with breast cancer, differenc- es within countries in the unknown determinants and the extensive de- lay between developmental chang- es and an increasing risk of pros- tate cancer may help to illuminate the apparent paradoxes. The inher- ent lack of high-quality registry data underpinning national incidence estimation, most notably in Africa, is likely to have hampered the infer- ence attempted here.
Cancer incidence by 2025: demographic and trend-based predictions The global population, 7 billion in 2012, will reach 8.3 billion by 2025 [11]. Slow changes in fertility and
increasing life expectancy are hav- ing a major impact on population growth and ageing in countries in developmental transition, with the greatest increases in cancer burden in the very same countries where trends in overall cancer rates are also rising. Updating an exercise based on the observed time trends in cancer-specific incidence by HDI level [9], with annual increases in rates of colorectal cancer, female breast cancer, and prostate cancer worldwide – and, in areas with high to very high HDI, for lung cancer in women – alongside annual de- creases in rates of stomach cancer and cervical cancer worldwide – and, in areas with high to very high HDI, for lung cancer in men – a pre- dicted global burden of more than 20 million new cancer cases can be anticipated by 2025. Countries with medium HDI and, particularly, those with low HDI will experience the greatest proportional popula- tion growth during this time, and thus the greatest increases in the future cancer burden (Fig. 1.2.14). Notably, this scenario predicts an excess of almost 2 million cancers in men over cancers in women.
Fig. 1.2.12. A woman in China tends a farm with her child at her side. Generally, there is a relative paucity of cancer incidence and mortality data in areas with low and medium Human Development Index.
Fig. 1.2.13. Children look out at Kampala’s skyline during a celebration of Uganda’s 50th anniversary of independence. Currently, all countries worldwide have a higher Human Development Index (HDI) than they had in 2000, but countries differ in the rate of transition to a higher HDI level.
66
Overview The predicted global burden is ex- pected to surpass 20 million new cancer cases by 2025, compared with an estimated 14.1 million new cases in 2012. Demographic chang- es are the key drivers of the unprec- edented growth in the cancer bur- den, but the profiles of cancer are changing as populations increas- ingly adopt behavioural and lifestyle habits that are common in affluent, industrialized countries. Elevated rates of female breast cancer, pros- tate cancer, and colorectal cancer incidence are linked to an attain- ment of higher levels of human de- velopment, with colorectal cancer the most markedly associated with HDI. A close to 2-fold increase in cumulative incidence of colorectal cancer is estimated as countries transition from low to higher cat- egories of HDI. This, coupled with the increases in incidence rates of colorectal cancer in countries where HDI is rapidly rising, may be related to the increasing prevalence of unhealthy dietary habits and de- creased levels of physical activity in populations undergoing develop- mental transition.
Lung cancer is the leading cause of cancer death in countries with medium, high, and very high HDI and is determined very largely by tobacco smoking, with risk be- ing influenced by the number of cigarettes smoked, the duration of the habit, and the composition of the tobacco used. Given the limited nature of long-term data accumula- tion on tobacco-related cancer in- cidence and mortality in countries with low and medium HDI, along- side the relatively recent smoking histories in these populations, cur- rent lung cancer rates are often low and not indicative of impending epidemics in either sex. The extent of the projected increases in lung cancer and other tobacco-related diseases is, however, inextricably linked to the global tactics of to- bacco companies aiming to expand their sales [12]. A smoking epidemic is unfolding in many countries with lower HDI that are in transition,
potentially impeding human devel- opment by consuming scarce re- sources, increasing pressures on already weak health-care systems, and inhibiting national productivity. Discernible increases in smoking- attributable risks have recently been reported in Bangladesh [13], China [14], India [15], and South Africa [16]. In many countries, in keeping
with the tobacco epidemic model, the habit has been acquired rela- tively recently in women; in some countries, in a departure from this model, women have not acquired the habit at all. Perhaps there is more encouraging evidence of the latter scenario, at least in China where a recent epidemiological study of asbestos workers reported
Fig. 1.2.14. Population estimates for 2012 and predictions for 2025: (A) changes in total population size by four-level Human Development Index (HDI); (B) incident cancer burden based on demographic changes and demographic + incidence rate changes, by sex and four-level HDI.
Chapter 1.2 • Transitions in human development and the global cancer burden 67
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .2
Fig. 1.2.15. Ranking of premature mortality from cancer compared with cardiovascular disease and diabetes (combined) and chronic obstructive pulmonary disease (ages 30–69 years, both sexes), estimated for 2011.
that of the relevant workforce, 79% of men smoked compared with less than 1% of women [17].
Cancer is becoming the major cause of premature deaths among noncommunicable diseases. A rap- id assessment of the ranking of pre- mature mortality (ages 30–69) from major noncommunicable diseases (Fig. 1.2.15) reveals how cancer has surpassed cardiovascular disease
(here combined with diabetes) and chronic obstructive pulmonary dis- ease in countries with very high HDI. It can therefore be expected that cancer will become a leading cause of premature death not only compared with other noncommuni- cable diseases but across the spec- trum of causes, as countries head towards the higher human develop- ment threshold. In turn, therefore,
noncommunicable diseases are recognized as a barrier to human development and as such should be a central part of the post-2015 agenda, after the target date for the Millennium Development Goals.
The author thanks Mathieu Laversanne for statistical analysis and generation of the graphics.
68
References
1. United Nations Development Programme (1990). Human Development Report 1990. Concept and Measurement of Human Development. New York: UNDP. Available at http://hdr.undp.org/en/reports/global/hdr 1990/.
2. Omran AR (1971). The epidemiologic transition. A theory of the epidemiol- ogy of population change. Milbank Mem Fund Q, 49:509–538. http://dx.doi.org/10. 2307/3349375 PMID:5155251
3. WHO (2011). Global Status Report on Non- Communicable Diseases 2010. Geneva: WHO. Available at http://www.who.int/ nmh/publications/ncd_report2010/en/.
4. Gersten O, Wilmoth JR (2002). The cancer transition in Japan since 1951. Demogr Res, 7:271–306. http://dx.doi.org/ 10.4054/DemRes.2002.7.5
5. WHO (2013). The top 10 causes of death. Available at http://www.who.int/ mediacentre/factsheets/fs310/en/.
6. Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
7. United Nations Development Programme (2013). Human Development Report 2012. The Rise of the South: Human Progress in a Diverse World. New York: UNDP. Available at http://www.undp.org/ content /undp/en/home/librar ypage/hdr/ human-development-report-2013/.
8. Bray F, Ren JS, Masuyer E, Ferlay J (2013). Global estimates of cancer prevalence for 27 sites in the adult population in 2008. Int J Cancer, 132:1133–1145. http://dx.doi. org/10.1002/ijc.27711 PMID:22752881
9. Bray F, Jemal A, Grey N et al. (2012). Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol, 13:790–801. http://dx.doi.org/10.1016/S14 70-2045(12)70211-5 PMID:22658655
10. Ferlay J, Shin HR, Bray F et al. (2010). Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer, 127:2893–2917. http://dx.doi.org/10.1002/ ijc.25516 PMID:21351269
11. United Nations Population Division (2007). World Population Prospects: The 2008 Revision. New York: United Nations. Available at http://www.un.org/en/develop m e n t / d e s a / p o p u l a t i o n / p u b l i c a t i o n s / trends/population-prospects.shtml.
12. O’Connor RJ, Wilkins KJ, Caruso RV et al. (2010). Cigarette characteristic and emission variations across high-, middle- and low-income countries. Public Health, 124:667–674. http://dx.doi.org/10.1016/j. puhe.2010.08.018 PMID:21030055
13. Alam DS, Jha P, Ramasundarahettige C et al. (2013). Smoking-attributable mortal- ity in Bangladesh: proportional mortality study. Bull World Health Organ, 91:757– 764. http://dx.doi.org/10.2471/BLT.13.120 196 PMID:24115799
14. Gu D, Kelly TN, Wu X et al. (2009). Mortality attributable to smoking in China. N Engl J Med, 360:150–159. http:// d x .d o i .o r g /10 .10 5 6 / N E J M s a 0 8 0 2 9 0 2 PMID:19129528
15. Jha P, Jacob B, Gajalakshmi V et al.; RGI- CGHR Investigators (2008). A nation- ally representative case-control study of smoking and death in India. N Engl J Med, 358:1137–1147. http://dx.doi.org/10.1056/ NEJMsa0707719 PMID:18272886
16. Sitas F, Egger S, Bradshaw D et al. (2013). Differences among the coloured, white, black, and other South African pop- ulations in smoking-attributed mortality at ages 35–74 years: a case-control study of 481,640 deaths. Lancet, 382:685–693. ht t p: //d x.d o i.o r g /10.1016 /S 014 0 - 673 6 (13)61610-4 PMID:23972813
17. Wang X, Lin S, Yu I et al. (2013). Cause- specific mortality in a Chinese chrysotile textile worker cohort. Cancer Sci, 104: 245–249. http://dx.doi.org/10.1111/cas.120 60 PMID:23121131
Chapter 1.3 • Childhood cancer 69
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .3
Summary • Childhood cancers are distinct
entities from cancers occurring in adults. Overall annual inci- dence rates vary between 50 and 200 per million in children and between 90 and 300 per million in adolescents. Reliable data on cancer in children and adolescents are available for only a small fraction of the world’s population.
• Over the past 50 years, 5-year survival has improved in high- income countries from less than 30% to more than 80%, al- though this trend may have pla- teaued recently. Treatment and follow-up must be adapted to the age of the patient and must embrace the patient’s family and its circumstances.
• The growing population of survivors requires specialized follow-up and care. Innovation in clinical research is required to ensure continued improve- ment of prognosis and reduced occurrence and severity of the late effects of treatment.
• The mortality-to-incidence ratio remains high in countries with low human development in- dex, although some diagnoses
are highly curable at low cost. National investment and inter- national collaboration are re- quired to improve outcomes.
Incidence rates The term “childhood cancer” is most commonly used to designate cancers
1.3 1 WORLDWIDE
Childhood cancer Eva Steliarova-Foucher A. Lindsay Frazier
Bernard W. Stewart (reviewer) Christopher P. Wild (reviewer)
Fig. 1.3.1. Cancer cases and deaths in children (aged < 15 years) as a proportion of the total numbers in the population, in 2012. HDI, Human Development Index.
that arise before the age of 15 years. Cancer in childhood represents be- tween 0.5% and 4.6% of the total number of cancer cases in a popu- lation (Fig. 1.3.1) [1]. The proportion of cancers that occur in children is higher in countries with low Human Development Index (HDI) than in countries with high HDI. This is
70
because children make up a larger percentage of the overall population (up to 50%) and concomitantly the elderly population, which exhibits the highest risk of cancer, is quite small in countries with low HDI.
Population-based cancer reg- istries around the world report overall incidence rates that vary between 50 and 200 per million children per year. The spectrum of tumour types differs across popu- lations (Fig. 1.3.2) [2–12]. Acute lymphoblastic leukaemia is the most common diagnosis, except in sub-Saharan Africa, where chil- dren are more prone to develop
non-Hodgkin lymphomas (most commonly Burkitt lymphoma) and Kaposi sarcoma, due to the specific infectious exposures in that region (notably Plasmodium, Epstein–Barr virus, and HIV) [11]. Brain tumours are the second most common ma- lignancy in regions that can afford to implement non-invasive diag- nostic technologies [13]. Improved diagnostics may also influence the incidence of neuroblastoma; inci- dental findings of indolent tumours discovered in the countries with highly developed medical surveil- lance may account for observed increases in incidence.
All childhood cancers are indi- vidually rare, with incidence rates varying from 1 per million for a very rare cancer such as hepatoblastoma to 50 per million for the most com- mon subgroup, lymphoid leukaemia. Haematological malignancies repre- sent 40–60% of tumours in the first 15 years of life. Embryonal tumours (e.g. retinoblastoma, neuroblastoma, nephroblastoma) constitute about 20% of childhood malignancies and virtually never occur except in chil- dren. Carcinomas represent less than 5% of childhood tumours, while these are the most frequent histolo- gies in adults. The characteristic morphological appearances imply a distinct developmental origin and etiology for tumours in childhood. Because of the age-specific histol- ogy, the International Classification of Childhood Cancer [14] is used to describe the occurrence and out- comes of cancer in childhood.
Increasingly, adolescents (aged 15–19 years) and young adults (aged 20–24 years) with cancer are being considered as a group re- quiring special consideration, simi- lar to childhood cancer, due to the unique composition of cancer types (Fig. 1.3.3) [15] and their health-care requirements. Although embryonal neoplasms in adolescents are rare, haematological neoplasms continue to be common, and most are lym- phomas. Central nervous system tu- mours still represent a large propor- tion of cancers, and bone tumours peak in the adolescent age group in many populations. Malignant melanoma is more common than in children, particularly in female ado- lescents. The leading cancer site in male adolescents is the testis, with age-specific rates rising until ages in the early forties. The prominent types in female adolescents are thyroid cancer, ovarian germ cell tumours, and cervical cancer. The worldwide incidence rates vary ap- proximately 3-fold in male (9–30 per 100 000) and in female (9–27 per 100 000) adolescents [16].
Fig. 1.3.2. Incidence rates of cancer in children (aged 0–14 years) in the 1990s and 2000s. CNS, central nervous system; PNET, primitive neuroectodermal tumours; SEER, Surveillance, Epidemiology and End Results national statistics; SNS, sympa- thetic nervous system.
Chapter 1.3 • Childhood cancer 71
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .3
The overall incidence of child- hood cancer increased by about 1% per year over the last three decades of the 20th century in Europe, North America, Australia, and elsewhere, although the rate of increase seems to have levelled off over the most recent decade [10,17–19]. Among adolescents, incidence has also increased (Fig. 1.3.4), at a similar rate [17,20]. The observed temporal trends might partly reflect improved diagnosis and reporting of cancers [13], but the impact of changes in ex- posures and lifestyles over the same period may have also contributed to the changes in incidence. For exam- ple, the recent study that showed a reduction of incidence rates of Wilms tumour and some other rare childhood tumours in the USA after supplementation of the food supply with folic acid [21] provides hope that the predisposing environmental factors will be identified and child- hood cancers prevented.
It is estimated that by 2035 the annual number of new can- cers across all ages will grow by 70% compared with 2012 esti- mates [1] (see also Chapter 1.2). In
comparison, the annual numbers of cancer cases and deaths among children are expected to increase by only 7%, based on the assump- tion of a medium fertility variant of the population growth [22] and sta- ble incidence and mortality rates.
However, from the international data it appears that the rate of childhood leukaemia, the most common can- cer in many childhood populations, is positively correlated with socio- economic status [23], and several hypotheses suggest that the risk factors associated with a high level of development, such as delayed exposure to common infections, population mixing, and changes in reproductive behaviour, influence the incidence of acute lymphoblastic leukaemia. In sub-Saharan African countries, where childhood leukae- mia is reported relatively rarely, sig- nificant increases in rates of child- hood leukaemia may be expected in the near future, both due to improv- ing diagnostic facilities and due to adoption of industrialized lifestyles.
Outcomes Over the past 40 years, outcomes of cancer in children have im- proved dramatically. In the United Kingdom, the 5-year survival rate increased from less than 30% to al- most 80% on average, thus reduc- ing the risk of death by 68% overall (Fig. 1.3.5) [24], and similar chang- es were observed in other high-in- come countries. These spectacular
Fig. 1.3.3. Cancer incidence rates in (A) male and (B) female adolescents (aged 15– 19 years) and young adults (aged 20–24 years) in Europe, in 2003–2007. Data from 84 cancer registries contributing data for the selected calendar years were retrieved from the European Cancer Observatory on 18 November 2013. CNS, central nervous system.
Fig. 1.3.4. Cancer trends in adolescents and young adults (aged 15–24 years) in Europe. Pooled data from the European Cancer Observatory from all cancer registries with data covering the period shown: Finland, Germany (Saarland), Iceland, Italy (Varese), Norway, Slovakia, Sweden, Switzerland (Geneva, St Gallen-Appenzell) and the United Kingdom (Scotland).
72
improvements are the results of therapeutic advances, fostered through worldwide collaboration of childhood cancer study groups and the International Society of Paediatric Oncology (www.siop- online.org).
Because of the age proxim- ity, the overlapping spectrum of tu- mours, and the need for a very sen- sitive approach, adolescent patients
Fig. 1.3.5. Five-year survival rates of children diagnosed before the age of 15 years in Britain during the indicated periods.
with selected malignancies are man- aged within a paediatric oncology practice. Indeed, adolescents may benefit from inclusion in paediatric therapeutic protocols compared with those developed for adults, at least for “paediatric-like” cancers such as acute lymphoblastic leukaemia [25]. The overall 5-year survival of ado- lescents is similar to that of children, reaching 84% [4,26].
Table 1.3.1. Five-year survival rates (%) in children (aged 0–14 years) with cancer diagnosed during the indicated periods
Cancer type
Location (period) [source of data]
Australia (1997–2006)
[18]
Shanghai, China (2002–2005)
[27]
Chennai, India (1990–2001)
[9]
Thailand (2003–2004)
[7]
All cancers 79.6 55.7 40.0 54.9
Leukaemias 80.6 52.2 36.3 57.4
Lymphomas 89.9 58.8 55.3 59.5
Central nervous system tumours 71.0 41.2 26.8 41.7
Neuroblastoma 67.8 — 36.9 33.6
Retinoblastoma 98.4 75.0 48.1 73.1
Renal cancers 88.6 86.7 58.0 70.4
Liver cancers 76.0 33.3 10.5 44.5
Malignant bone tumours 68.9 52.6 30.6 33.7
Soft tissue sarcomas 72.1 54.1 36.3 50.1
Germ cell tumours 89.4 78.4 38.0 70.6
Carcinomas and melanoma 93.3 88.9 35.1 —
Other 72.2 — — —
The spectacular improvement in outcomes is, however, limited to high- income countries. The results from rare reports of childhood cancer sur- vival in middle-income countries are compared with data from Australia in Table 1.3.1 [7,9,18,27]. Despite the dif- ferent reporting periods, it is evident that the proportion of 5-year survivors in India is much lower than that seen in high-income countries. The low survival figures may be explained by late presentation at diagnosis, treat- ment abandonment, and the absence of sophisticated multidisciplinary care and adequate resources.
In the absence of globally compa- rable data on cancer patient survival, estimates are used to describe the ultimate cancer burden in terms of mortality. Compared with the estimat- ed 163 000 new cancers in children worldwide in 2012, the 80 000 deaths represented about a half of the new cases. In 2012, 82% of the new cases and 93% of the deaths occurred in the less developed countries [1]. In Africa, the mortality represents 60% of the incidence, while in North America this ratio is less than 15% (Fig. 1.3.6) [1,16]. National investment and inter- national collaborations are required to improve this situation.
Chapter 1.3 • Childhood cancer 73
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .3
Whereas low- and middle-income countries need to build capacity to reach the 80% 5-year survival cur- rently achievable in the countries with the most resources, the reduction in mortality observed in Europe and the USA has plateaued over the most recent period (Fig. 1.3.7) [28,29]. There remains, therefore, an urgent need for new major breakthroughs in therapies to re-accelerate the current rate of improvement in survival [30].
The greatly improved survival in high-income countries has given rise to a growing population of sur- vivors of cancer in childhood or adolescence, a large proportion of whom face significant late effects of their primary disease and cancer treatment. Survivors experience an excess risk of death and second primary cancer long past their diag- nosis. In a Nordic study of 21 984 5-year survivors of childhood cancer,
the overall standardized mortality ratio was 8.3 and the absolute ex- cess risk was 6.2 per 1000 person- years. The causes of death included primary cancer (60%), subsequent malignant neoplasms (12%), and non-cancer causes (27%) [31]. Other sequelae of cancer treatment include neurocognitive impairment, cardiovascular disease, and other organ dysfunction, but also the psy- chosocial impact of the disease and its treatment on patients, their family members, and their future lives. It is clear that the usual 5-year survival statistics reflect a short-term, rather than a long-term, prognosis for chil- dren with cancer. Lifelong follow-up is important and has to take into ac- count the inevitable transition into the non-specialized, and often un- aware, primary health-care system.
Data sources Worldwide, there are approximately 50 countries with established na- tional population-based cancer reg- istries, which provide their data for comparative international studies. Most of these registries collect data on cancers occurring in individuals of all ages (a general cancer registry), but some countries are also covered by a separate, national paediatric cancer registry. Paediatric cancer registries often collect more details on the patient, diagnosis, treatment, and outcome. Regional population- based cancer registries operate in other countries and provide valuable information on cancer burden, in- cluding that in children. Comparative incidence data on childhood cancer were reported in the two volumes of the IARC Scientific Publications series International Incidence of Childhood Cancer; a third volume of the series, describing incidence in children and adolescents over the past two decades, is in preparation (http://iicc.iarc.fr/).
Childhood cancer is a rare dis- ease, requiring that large populations are covered to collect sufficient num- bers of cases for meaningful analy- ses. National cancer registration is the ideal option for rare cancers such
Fig. 1.3.6. Estimated mortality-to-incidence ratio for cancer in children (aged 0–14 years) in 2012, and percentage of population coverage by cancer registration (all ages).
Fig. 1.3.7. Cancer mortality in children (aged 0–14 years) in three European countries (Ireland, the Netherlands, and the United Kingdom combined) for 1950–2009 and in the USA for 1950–2007.
74
as those occurring in childhood and possibly also in adolescence, and is a workable model, as shown by the examples of Argentina, South Africa, and the Islamic Republic of Iran. In low-income settings, however, a more appropriate strategy would be to start a regional population-based cancer registry. In this respect, exist- ing hospital-based cancer registries may aim to become population- based if they succeed in determining the relevant geographically defined population at risk. It is no coinci- dence that population coverage by cancer registration is correlated with high incidence rates and low mortal- ity rates (Fig. 1.3.6). Although found- ing a cancer registry will not auto- matically reduce the mortality rates in a population, cancer registration is an essential element of a cancer control plan, and is bound to serve as the evidence base that will inform other services to cancer patients and eventually lead to improvements in survival.
The value of the statistics pro- duced by cancer registries depends on the quality, accuracy, and com- pleteness of the data generated in clinical practice. The better the cooperation between the partners, the more useful the produced out- put will be. For optimal functioning of cancer registries, it is essential to
ensure access to the relevant data. Regulations imposing requirements of informed consent, trusted third parties, or various encryption tech- niques would make it virtually im- possible to use data from a cancer registry and thus hinder the achieve- ment of better outcomes for current survivors and future patients.
Several initiatives have been un- dertaken in Europe recently to en- large the information base available within population-based cancer reg- istries and allow a more specific inter- pretation of the outcome in addition to a description of the geographical and temporal patterns. This work is funded by the European Union within the Seventh Framework Programme of the European Commission. The European Network for Cancer Re- search in Children and Adolescents is a Network of Excellence with the overall aims to improve survival and quality of life for children and ado- lescents with cancer, to facilitate ac- cess to the best therapies, to develop biology-guided therapies, and to ef- ficiently structure and enhance col- laboration within the field of paediat- ric oncology in Europe (http://www. encca.eu/). PanCare Childhood and Adolescent Cancer Survivor Care and Follow-up is a research proj- ect of the PanCare Network with a
mission to provide every childhood and adolescent cancer survivor with better access to care and better long- term health (http://www.pancare surfup.eu/). The member registries of the European Network of Cancer Registries (http://www.encr.com.fr/) are involved in the European Network for Cancer Research in Children and Adolescents and in PanCare Childhood and Adolescent Cancer Survivor Care and Follow- up to investigate the feasibility of collecting additional clinically rel- evant information on children and adolescents with cancer to facilitate routine short-term clinical follow-up and long-term follow-up for selected late effects. The additional variables of interest may include pre-existing patient conditions (e.g. genetic syn- drome), diagnostic data (e.g. biologi- cal markers or stage), information on treatment received (e.g. inclusion in clinical trials), referral and follow-up (e.g. date of relapse), and possible late effects in long-term survivors (e.g. cause of death, cardiac con- dition). Sharing expertise between clinical and population-based re- search will simultaneously enhance the potential of data collected in both types of resources and provide a cost-effective solution to advance research in paediatric oncology.
Fig. 1.3.8. A young patient is treated at the 57357 Children’s Cancer Hospital in Cairo, Egypt.
Fig. 1.3.9. A child receiving chemotherapy at the Uganda Cancer Institute (Lymphoma Treatment Centre). Burkitt lymphoma is common in Uganda.
Chapter 1.3 • Childhood cancer 75
1 W
O R
L D
W ID
E C
H A
P T
E R
1 .3
References
1. Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
2. Moreno F, Loria D, Abriata G, Terracini B; ROHA network (2013). Childhood cancer: incidence and early deaths in Argentina, 2000–2008. Eur J Cancer, 49:465– 473. http://dx.doi.org/10.1016/j. ejca.2012.08.001 PMID:22980725
3. Moradi A, Semnani S, Roshandel G et al. (2010). Incidence of childhood cancers in Golestan province of Iran. Iran J Pediatr, 20:335–342. PMID:23056726
4. Howlader N, Noone AM, Krapcho M et al., eds (2013). SEER Cancer Statistics Review, 1975–2010. Bethesda, MD: National Cancer Institute. Available at http://www.seer.cancer.gov/csr/1975_ 2010/browse_csr.php?section=29&page =sect_29_table.02.html.
5. Lacour B, Guyot-Goubin A, Guissou S et al. (2010). Incidence of childhood cancer in France: National Children Cancer Registries, 2000–2004. Eur J Cancer Prev, 19:173–181. http://dx.doi. o r g /10.10 97/C E J. 0 b 013 e 3 2 8 3 3 876 c 0 PMID:20361423
6. Kaatsch P, Spix C (2012). German Childhood Cancer Registry Annual Report 2011 (1980 –2010). Mainz: Institute of Medical Biostatistics, Epidemiology and Informatics at the University Medical Center of the Johannes Gutenberg University. Available at http:// w w w. k i n d e r k r e b s r e g i s t e r. d e /e x t e r n / v e r o e f f e n t l i c h u n g e n / j a h r e s b e r i c h t e / aktueller-jahresbericht/index.html?L=1.
7. Wiangnon S, Veerakul G, Nuchprayoon I et al. (2011). Childhood cancer incidence and survival 2003–2005, Thailand: study from the Thai Pediatric Oncology Group. Asian Pac J Cancer Prev, 12:2215–2220. PMID:22296359
8. Fajardo-Gutiérrez A, Juárez-Ocaña S, González-Miranda G et al. (2007). Incidence of cancer in children residing in ten ju- risdictions of the Mexican Republic: im- portance of the Cancer registry (a pop- ulation-based study). BMC Cancer, 7:68. PMID:17445267
9. Swaminathan R, Rama R, Shanta V (2008). Childhood cancers in Chennai, India, 1990–2001: incidence and survival. Int J Cancer, 122:2607–2611. http://dx.doi. org/10.1002/ijc.23428 PMID:18324630
10. Baade PD, Youlden DR, Valery PC et al. (2010). Trends in incidence of childhood cancer in Australia, 1983–2006. Br J Cancer, 102:620–626. http://dx.doi.org/ 10.1038/sj.bjc.6605503 PMID:20051948
11. Parkin DM, Ferlay J, Hamdi-Chérif M et al. (2003). Cancer in Africa: Epidemiology and Prevention. Chapter 5: Childhood cancer. Lyon: IARC (IARC Scientific Publications Series, No. 153), pp. 381– 396.
12. Bao PP, Zheng Y, Wang CF et al. (2009). Time trends and characteristics of child- hood cancer among children age 0–14 in Shanghai. Pediatr Blood Cancer, 53:13– 16. http://dx.doi.org/10.1002/pbc.21939 PMID:19260104
13. Kroll ME, Carpenter LM, Murphy MF, Stiller CA (2012). Effects of changes in diagnosis and registration on time trends in recorded childhood cancer in- cidence in Great Britain. Br J Cancer, 107:1159–1162. http://dx.doi.org/10.1038/ bjc.2012.296 PMID:22898786
14. Steliarova-Foucher E, Stiller CA, Lacour B, Kaatsch P (2005). International Classification of Childhood Cancer, third edition. Cancer, 103:1457–1467. h t t p: //d x . d o i . o r g /10 .10 0 2 /c n c r. 2 0 910 PMID:15712273
15. Steliarova-Foucher E, O’Callaghan M, Ferlay J et al. (2012). European Cancer Observatory: Cancer Incidence, Mortality, Prevalence and Survival in Europe, ver- sion 1.0. European Network of Cancer Registries, IARC. Available at http:// eco.iarc.fr.
16. Curado MP, Edwards B, Shin HR et al. (2007). Cancer Incidence in Five Continents, Vol. IX. Lyon: IARC (IARC Scientific Publications Series, No. 160).
17. Kohler BA, Ward E, McCarthy BJ et al. (2011). Annual report to the nation on the status of cancer, 1975–2007, featur- ing tumors of the brain and other nervous system. J Natl Cancer Inst, 103:714–736. h t t p: //d x . d o i . o r g /10 .10 9 3 / j n c i /d j r 0 7 7 PMID:21454908
18. Baade PD, Youlden DR, Valery PC et al. (2010). Population-based survival esti- mates for childhood cancer in Australia during the period 1997–2006. Br J Cancer, 103:1663–1670. http://dx.doi.org/10.1038/ sj.bjc.6605985 PMID:21063404
19. Baba S, Ioka A, Tsukuma H et al. (2010). Incidence and survival trends for child- hood cancer in Osaka, Japan, 1973–2001. Cancer Sci, 101:787–792. http://dx.doi. org/10.1111/j.13 49 -70 0 6. 20 0 9.014 43.x PMID:20132215
20. Stiller CA, Desandes E, Danon SE et al. (2006). Cancer incidence and survival in European adolescents (1978–1997). Report from the Automated Childhood Cancer Information System project. Eur J Cancer, 42:2006–2018. http://dx. d o i . o r g / 1 0 .1 0 1 6 / j . e j c a . 2 0 0 6 . 0 6 . 0 0 2 PMID:16919767
21. Linabery AM, Johnson KJ, Ross JA (2012). Childhood cancer incidence trends in association with US folic acid fortification (1986–2008). Pediatrics, 129:1125–1133. http://dx.doi.org/10.1542/ peds.2011-3418 PMID:22614769
22. UN World Population Prospects, 2010 revision. Available at http://www.un.org/ esa/population/unpop.htm.
23. Kroll ME, Stiller CA, Richards S et al. (2012). Evidence for under-diagnosis of childhood acute lymphoblastic leukae- mia in poorer communities within Great Britain. Br J Cancer, 106:1556–1559. ht tp://dx.doi.org /10.103 8/ bjc. 2012.102 PMID:22472883
24. Stiller CA, Kroll ME, Eatock EM (2007). Survival from childhood cancer. In: Stiller CA, ed. Childhood Cancer in Britain: Incidence, Survival, Mortality. Oxford: Oxford University Press, pp. 131–204.
25. Ram R, Wolach O, Vidal L et al. (2012). Adolescents and young adults with acute lymphoblastic leukemia have a better outcome when treated with pediatric- inspired regimens: systematic review and meta-analysis. Am J Hematol, 87:472– 478. http://dx.doi.org/10.1002/ajh.23149 PMID:22388572
26. Gatta G, Zigon G, Capocaccia R et al.; EUROCARE Working Group (2009). Survival of European children and young adults with cancer diagnosed 1995– 2002. Eur J Cancer, 45:992–1005. http:// dx.doi.org /10.1016 / j.ejc a. 20 0 8.11.0 42 PMID:19231160
27. Bao PP, Zheng Y, Wu CX et al. (2012). Population-based survival for child- hood cancer patients diagnosed during 2002–2005 in Shanghai, China. Pediatr Blood Cancer, 59:657–661. http://dx.doi. org/10.1002/pbc.24043 PMID:22302759
28. WHO Mortality Database. Available at http://www.who.int /healthinfo/mortality _ data/en/ and http://www-dep.iarc.fr.
29. Pritchard-Jones K, Pieters R, Reaman GH et al. (2013). Sustaining innovation and improvement in the treatment of child- hood cancer: lessons from high-income countries. Lancet Oncol,14:e95–e103. http://dx.doi.org/10.1016/S1470-2045(13) 70010-X PMID:23434338
30. Sullivan R, Kowalczyk JR, Agarwal B et al. (2013). New policies to address the global burden of childhood cancers. Lancet Oncol, 14:e125–e135. http://dx.doi.org/ 10.1016/S1470-2045(13)70007-X PMID: 23434339
76
31. Garwicz S, Anderson H, Olsen JH et al.; Association of the Nordic Cancer Registries; Nordic Society for Pediatric Hematology Oncology (2012). Late and very late mortality in 5-year survivors of childhood cancer: changing pattern over four decades–experience from the Nordic countries. Int J Cancer, 131:1659– 1666. http://dx.doi.org/10.1002/ijc.27393 PMID:22170520
Websites CANCERMondial: http://www-dep.iarc.fr
European Cancer Observatory: http://eco.iarc.fr
European Network for Cancer Research in Children and Adolescents: http://www.encca.eu/
European Network of Cancer Registries: http://www.encr.com.fr/
German Childhood Cancer Registry Annual Report 2011 (1980–2010): http://www.kinderkrebsregister.de/extern/ veroeffentlichungen/jahresberichte/ aktueller-jahresbericht/index.html?L=1
GLOBOCAN: http://globocan.iarc.fr
International Incidence of Childhood Cancer: http://iicc.iarc.fr/
International Society of Paediatric Oncology: www.siop.nl
PanCare Childhood and Adolescent Cancer Survivor Care and Follow-up: http://www.pancaresurfup.eu/
Surveillance, Epidemiology and End Results Program: http://seer.cancer.gov/
UN Department of Economic and Social Affairs, Population Division: http://www.un.org/esa/population/unpop.htm
WHO Mortality Data and Statistics: http://www.who.int/healthinfo/statistics/ mortality/en/index.html
Reconciling stress and cancer: insights from telomeres 77
Summary Various conditions throughout geo- graphically and societally diverse parts of the world share in common the potential to inflict severe, often prolonged, psychological stresses on human populations. Telomere maintenance, which is emerging as a central biological process that plays roles in both prevention and progression of human cancers and other ageing-related diseases, is adversely affected by psychological stress. Telomeres are the dynamic structures that cap and protect chromosome ends. Telomeric DNA attrition commonly occurs in hu- man cells during ageing or because of inherited “telomere syndromes”. If unchecked, such attrition can eventually lead to cellular malfunc- tioning. This in turn contributes to multiple, often co-morbid, diseases of ageing, including a spectrum of cancers. Clear links already are known to exist between stress, telomere attrition, and several other common diseases of ageing be- sides cancers. Therefore, the les- sons already learned from such studies should inform strategies and policies in efforts to intercept human cancers and reduce their worldwide burden.
Telomeres, the protective terminal regions of our chromosomes, are dynamic structures that play multi- ple roles to ensure correct cellular functions and prevent genomic in- stability. Both biologically internal and external factors that influence human conditions can compromise telomere maintenance, and as I dis- cuss in this Perspective, this knowl- edge has implications for reducing cancer risks.
First, a short background on telo- meres and telomerase. Telomeres, found at every end of every sta- ble eukaryotic chromosome, form “caps” made up of elaborately con- trolled DNA–protein complexes. The telomere structure consists of a scaffold of a thousands-of-base- pairs-long stretch of a simple re- peated DNA sequence at the ex- treme end of the chromosome that directly binds a suite of telomeric, DNA sequence-specific, protective proteins. These proteins in turn in- teract with several other proteins that, through multiple mechanisms (collectively called capping), aid in protecting the telomeres (reviewed in [1]).
Crucially, a minimum length of telomeric DNA is required to sus- tain chromosome stability and cel- lular functions. However, net telo- mere erosion commonly occurs in humans over the lifespan [2]. Action
Reconciling stress and cancer: insights from telomeres
Elizabeth H. Blackburn,
a professor at the University of
California San Francisco, was
awarded the Nobel Prize in Physiology
or Medicine 2009 for her discovery
of how chromosomes are protected
by telomeres and the enzyme
telomerase. The pioneering work
of Dr Blackburn and her colleagues
has far-reaching implications and
has advanced the knowledge of the
impact of telomeres and telomerase in
human health and disease, particularly
in cancer, inherited diseases,
and ageing. Born in Australia,
Dr Blackburn earned B.Sc. and M.Sc.
degrees in biochemistry from the
University of Melbourne, received
her Ph.D. in molecular biology from
the University of Cambridge, and
completed postdoctoral work in
molecular and cell biology at Yale
University. Beyond the laboratory,
Dr Blackburn’s contemplation of the
implications of scientific research for
society underpins her recognized
contributions to bioethics and
public policy discourse.
Elizabeth H. Blackburn
78
of the cellular enzyme telomerase (a specialized ribonucleoprotein re- verse transcriptase) can elongate telomeric DNA and thus counteract telomere erosion, replenishing telo- mere length and functional integrity. The telomeric complex is a dynamic entity, with the capacity to be con- stantly shortened, built, shortened, and rebuilt. Yet the general rule in humans as they age appears to be that telomerase action over- all in somatic cells is insufficient to guarantee continued telomere maintenance in the face of attrition throughout the human lifetime.
If the telomeric repeat DNA tract becomes too short, the chromo- some end behaves as though it is a broken DNA end and becomes subject to inappropriate responses in the cell. These responses include premature cessation of cell divi- sion as well as reprogramming of transcriptional and other pathways to produce pro-inflammatory and tumour-promoting factors (reviewed in [3]). Furthermore, deleterious at- tempts by the cell to repair the telo- mere lead to genomic instabilities of various kinds that can promote cancers [4]. In humans the net re- sult is to cause or spur various dis- ease processes, including cancer progression [3,5].
Telomere maintenance and telomerase Clinical, genetic, cellular, and mo- lecular research of the past sev- eral years has recently converged to show that either too much or too little telomerase, even over quite a narrow range, can greatly increase lifetime risks of a range of different cancers, including many common cancers. Such findings indicate that humans appear to live on a knife- edge with respect to the telomere maintenance machinery and can- cer risks. Therefore, external fac- tors that influence telomere main- tenance and telomerase constitute potentially significant contributors to cancer risks. Importantly, these factors include psychosocial inputs that bear on chronic psychological
stress. In this Perspective, I briefly discuss current understanding of these issues, and their implications.
Can telomere shortening con- tribute to diseases of ageing, in- cluding cancers? The answer is yes. Inherited disorders that com- promise telomere maintenance establish a causal relationship be- tween functional compromise of telomeres (usually caused by their over-shortening to below-functional lengths) and a wide range of dis- eases, collectively dubbed “telo- mere syndromes” [3]. These inher- ited syndromes include pulmonary fibrosis, diabetes, compromised cardiovascular function, and im- mune system disorders such as aplastic anaemia. They also include high risks for various haematologi- cal malignancies and gastrointes- tinal tract and squamous cell can- cers, which can account for about 10% of the overall mortality in indi- viduals with such syndromes. Such telomere maintenance mutations have cropped up sporadically in in- dividual families from different parts of the world and are often well stud- ied, as are mouse genetic models of telomere deficiency. Therefore, the causal link between this large inherited risk of cancers and very short, and hence non-functional, telomeres (caused by inadequate telomerase action) is compelling.
Can there be too much telome- rase? Again the answer is yes, in abnormal situations. In addition to the cancer-causing mutations that diminish telomerase and cause overly short telomeres, even small changes in telomerase activity in the upward as well as the down- ward direction can greatly augment risks for some cancers and appear to act early in cancer progression. Point mutations (somatic or inher- ited) that increase expression of a telomerase component by only about 2-fold were recently shown to be associated with dramatically increased risks of melanoma in hu- mans and are highly recurrent in sporadic melanomas [6,7]. Similar mutations were found also in liver and bladder cancers. It has been
known for many years that highly upregulated telomerase is a hall- mark of many advanced (especially highly malignant) human cancers. This upregulation can be rational- ized primarily as the need for can- cer cells to maintain their telomeres to sustain their continued prolifera- tion, a central property of cancer cells. However, interestingly, the melanoma-causing mutations de- scribed above bear the hallmarks of sunlight-induced mutations. This suggests that small but abnormal increases in telomerase may drive melanoma at even early stages [6,7].
Risks of age-related diseases In human populations as well as pa- tients with inherited telomere mainte- nance diseases, telomere shortness is linked to risks of some cancers. Indeed, in multiple population and cohort studies, the degree of telo- mere shortness is quantitatively linked to future risks of acquiring many age-related diseases, includ- ing some (but not all) cancers [8]. Impaired telomere maintenance leading to eventual over-shortening of telomeres has defined cellular and pathogenic consequences, as shown by laboratory research and the inherited telomere syndromes described above. Strikingly, these early-onset telomere syndrome dis- eases echo the diseases that become common with ageing in general hu- man populations. After correcting for age and many other factors, links repeatedly have been found be- tween shortness of telomeres in normal cells and the most common diseases of ageing – poor immune function, various forms of cardio- vascular disease, diabetes, pul- monary disease, certain cancers, dementia, depression, and some other mental disorders. Such con- ditions are often co-morbid. Thus, it is plausible and indeed likely that in general populations, telomere shortening can contribute a caus- al role (but obviously not the sole role) in the development and patho- genesis of these diseases. Being
Reconciling stress and cancer: insights from telomeres 79
systemic, and hence common across all organs and tissues, telo- mere shortening (with eventual loss of telomere function) begins to pro- vide a compelling explanation for the co-morbidity of many age-related diseases and conditions.
Since most cancers in general populations are not obviously inher- ited, and since compromised telo- mere maintenance can increase cancer risks, factors that can affect telomere maintenance are of great interest. Although telomeres gen- erally shorten with age in humans, chronological age can account for only 10% or less of the total variance in telomere length in the general population [9]. While telo- mere length shows some degree of heritability, common genetic varia- tions in populations (as opposed to rare mutations) account for only a very limited portion of variability in telomere length. Furthermore, the heritability of telomere length dimin- ishes with age. In a telling example, an interaction between smoking history and a common single-nucle- otide polymorphism associated with telomere length has been reported for risk of bladder cancer (reviewed in [10]). Hence, it is important to un- derstand other influences on telo- mere maintenance, and eventually to understand how genetic and non- genetic influences interact to affect cancer risks.
Effects of psychological stress One important factor is severe psy- chological stress. This can have many inputs, including societal con- ditions of poverty, war, and social strife, and childhood abuse, trau- ma, and neglect [11]. Psychological stress has long been established as an important risk factor for some common diseases of ageing – prom- inently including cardiovascular dis- ease, diabetes, and depression – but its relationship to cancers is
less well understood. Chronic psy- chological stress is also associated with telomere shortness. A growing body of studies links various types of chronic psychological stress, dis- tress, and poor lifestyle (e.g. smok- ing and lack of exercise) to shorter telomeres [12,13]. In addition, social factors that cause chronic psycho- logical stress are also established as risk factors for at least some chronic diseases of ageing and are also linked to telomere shortness in both adults and children. Some of these studies indicate that the psy- chological stress causes the telo- mere shortening, but this does not exclude two-way effects. Hence, stress, telomere shortness, and dis- ease risks are quantifiably related through three pairwise links: stress with telomere shortness, stress with disease risks, and telomere short- ness with risks of these same dis- eases (reviewed in [14]). The fact that telomere compromise is linked to (and in some situations is known to cause) diseases similar to those linked to (and in some cases caused by) stress leads directly to the hy- pothesis that telomere shortness at least partly explains how stress causes many common, often co- morbid, diseases of ageing.
Effects of severe stress on telo- mere shortness can begin even early in life, as shown in direct stud- ies of children exposed to adverse socioeconomic situations and vio- lence. Significantly, the effects can be lasting: past histories of child- hood adversity, and even exposure of developing fetuses to maternal stress, have been quantitatively re- lated to telomere shortness in hu- man adults (reviewed in [14]). It is also well established that adverse events that occurred in childhood predict early onset of chronic dis- eases, including mental disorders, in adults. Thus, it will be important to understand how much telomere compromise underlies or contributes
to the effects of childhood stressors on risks of diseases that continue to reverberate throughout life. An important unanswered question is whether such diseases include cancers. While compromised telo- mere maintenance has been linked to cancer risks, less is known about the impact of chronic psychological stress on cancer. Nonetheless, be- cause both telomere shortness and chronic stress have independently been linked to several other chronic diseases of ageing, it will be impor- tant to extend the lessons learned from these other chronic diseases into a fuller understanding of can- cer risks and cancer progression. Going forward, policies aimed at lessening the burden of cancer and other diseases throughout the world [15] cannot afford to ignore these connections.
Conclusions A tsunami of overwhelming health and social issues is looming for the world’s ageing population, can- cers being prominent among them. We need to know how stress af- fects cancers. Basic biomedical research has identified telomeres, and correctly controlled telomere maintenance mechanisms, as cru- cial for replenishment of cells and tissues and prevention of cancers. Compromised telomere mainte- nance is emerging as a potential root factor underlying and interact- ing with mechanisms and etiolo- gies of multiple co-morbid diseases of ageing. Psychological stress, including that related to societal and life factors, has independently been linked to these same chronic diseases and to telomere compro- mise. What we have learned about this underlying disease process should be applied to strategies to mitigate cancer risks, intercept cancer progression, and improve cancer survival.
80
References 1. Sfeir A, de Lange T (2012). Removal
of shelterin reveals the telomere end- protection problem. Science, 336:593–597. http://dx.doi.org/10.1126/science.1218498 PMID:22556254
2. Aubert G, Lansdorp PM (2008). Telomeres and aging. Physiol Rev, 88:557–579. http:// dx.doi.org/:10.1152/physrev.00026.2007 PMID:18391173
3. Armanios M, Blackburn EH (2012). The telomere syndromes. Nat Rev Genet, 13:693 –70 4. ht tp://dx.doi.org/10.1038/ nrg3246 PMID:22965356
4. Chin L, Artandi SE, Shen Q et al. (1999). p53 deficiency rescues the adverse effects of telomere loss and cooperates with telomere dysfunction to accelerate carcinogenesis. Cell, 97:527–538. http://dx.doi.org/10.1016/ S0092-8674(00)80762-X PMID:10338216
5. Hills M, Lansdorp PM (2009). Short telomeres resulting from heritable mutations in the telomerase reverse transcriptase gene predispose for a variety of malignancies. Ann N Y Acad Sci, 1176:178–190. http://dx.doi. org/10.1111/j.1749 - 6 632. 20 09.0 4565.x PMID:19796246
6. Huang FW, Hodis E, Xu MJ et al. (2013). Highly recurrent TERT promoter mutations in human melanoma. Science, 339:957– 959. http://dx.doi.org/10.1126/ science.1229259 PMID:23348506
7. Horn S, Figl A, Rachakonda PS et al. (2013). TERT promoter mutations in familial and sporadic melanoma. Science, 339:959 – 961. http://dx.doi.org/10.1126/ science.1230062 PMID:23348503
8. Willeit P, Willeit J, Kloss-Brandstätter A et al. (2011). Fifteen-year follow-up of association between telomere length and incident cancer and cancer mortality. JAMA, 3 0 6:42 – 4 4. ht t p: //d x.d o i.o r g /10.10 01/ jama.2011.901 PMID:21730239
9. Lin J, Epel ES, Blackburn EH (2009). Telomeres, telomerase, stress, and aging. In: Berntson GG, Cacioppo JT, eds. Handbook of Neuroscience for the Behavioral Sciences. Hoboken, NJ: Wiley, pp. 1280–1295.
10. Blackburn EH (2011). Walking the walk from genes through telomere maintenance to cancer risk. Cancer Prev Res (Phila), 4:473–475. http://dx.doi.org/10.1158/1940- 6207.CAPR-11-0066 PMID:21460394
11. Adler N, Pantell MS, O’Donovan A et al. (2013). Educational attainment and late life telomere length in the Health, Aging and Body Composition Study. Brain Behav Immun, 27:15–21. http:// d x .d o i .o r g /10 .1016 / j . b b i . 2 012 . 0 8 . 014 PMID:22981835
12. Surtees PG, Wainwright NW, Pooley KA et al. (2011). Life stress, emotional health, and mean telomere length in the European Prospective Investigation into Cancer (EPIC)-Norfolk population study. J Gerontol A Biol Sci Med Sci, 66:1152–1162. http://dx.doi.org/10.1093/gerona /glr112 PMID:21788649
13. Lin J, Epel E, Blackburn E (2012). Telomeres and lifestyle factors: roles in cellular aging. Mutat Res, 730:85–89. http:// dx.doi.org/10.1016/j.mrfmmm.2011.08.003 PMID:21878343
14. Blackburn EH, Epel ES (2012). Telomeres and adversity: too toxic to ignore. N a t u r e, 4 9 0 :16 9 –171. h t t p: //d x .d o i . org/10.1038/490169a PMID:23060172
15. Marmot M, Friel S, Bell R et al.; Commission on Social Determinants of Health (2008). Closing the gap in a generation: health equity through action on the social determinants of health. Lancet, 372:1661– 1669. http://dx.doi.org/10.1016/S0140 - 6736(08)61690-6 PMID:18994664
Most cancers are associated with risks from en- vironmental, lifestyle, or behavioural exposures. Comparing the highest with the lowest incidence recorded for different communities or examining changes over time provides clues as to which factors determine the distribution of cancer globally and how they differ in their impact. For the most important fac- tor – tobacco smoking – differences worldwide reflect timing. The epidemic of lung cancer has peaked in high-income countries, whereas in China, for exam- ple, the epidemic is starting. Chronic infections con- tinue to play a major role in common cancers in parts of Africa and Asia but are far less common in Europe and North America. Diet exerts an influence on can- cer development; in parts of Africa and Asia, this is sometimes related to the impact of potent carcino- gens such as aflatoxins, but in high-income countries this is primarily because of energy-rich food intake and associated overweight or obesity, linked in part to low levels of physical activity. Personal choices about alcohol intake and deliberate sun exposure are reflected in the incidence of particular cancer types. Worldwide, exposure to hazardous workplaces and environmental pollution causes cancer in low- and middle-income countries despite experience that may have been gained from other communities.
Cancer etiology
2
82
Summary • Worldwide, tobacco use kills
approximately 6 million people annually, from cancer and other diseases.
• Consumption of manufactured cigarettes has peaked and is declining in most high-income countries but is increasing or persisting at high levels in many low- and middle-income countries.
• The historical progression of the cigarette epidemic that unfolded first in high-income countries is now occurring worldwide, espe- cially in men.
• The application of effective to- bacco control measures such as large periodic increases in excise taxes and the elimination of advertising can provide un- paralleled opportunities for the prevention of cancer and other diseases worldwide.
At least 1.3 billion people world- wide use some form of tobacco [1]. In 2000, 4.83 million (uncer- tainty range, 3.94–5.93 million) premature deaths globally were attributable to smoking, 2.41 mil- lion (1.80–3.15 million) in low-
and middle-income countries and 2.43 million (2.13–2.78 million) in high-income countries. Of these deaths, 3.84 million were in men. The leading causes of death from smoking were cardiovascular dis- eases (1.69 million deaths), chronic obstructive pulmonary disease (0.97 million deaths), and lung cancer (0.85 million deaths) [2]. By 2013, the total number of deaths worldwide attributable to tobacco use was approximately 6 million [3].
Manufactured cigarettes are the predominant tobacco product used globally; nearly 20% of the world’s
adult population (about 800 million men and 200 million women) smoke cigarettes [4]. Other smoked prod- ucts (cigars, pipes, bidis, kreteks, and water pipes) and smokeless forms of tobacco (snuff, chewing tobacco, betel quid, and paan) are also commonly used regionally [5]. Reliable nationally representa- tive data on smoking prevalence have been available for certain high-income countries since the mid-20th century, but coverage of low- and middle-income countries has expanded only during the past decade. The Global Adult Tobacco
2.1 2 ETIOLOGY
The global tobacco epidemic Michael J. Thun Jane R. Nilson Alex C. Liber Evan Blecher
Prabhat Jha (reviewer) Pramil N. Singh (reviewer)
Fig. 2.1.1. In Togo, a group of schoolboys huddle together on a street to smoke cigarettes.
Chapter 2.1 • The global tobacco epidemic 83
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
Survey has now surveyed 19 coun- tries that include half of the world’s adult population [6]; the Global Youth Tobacco Survey has studied students aged 13–15 years in more than 160 countries [7]. Whereas in most high-income countries cigarette smoking has now peaked and is declining, in many low- and middle-income countries it has become entrenched, especially in men. More than 80% of all smokers reside in low- and middle-income countries; more than 60% live in just 10 countries (in decreasing order): China, India, Indonesia, the Russian Federation, the USA, Japan, Brazil, Bangladesh, Ger- many, and Turkey [8].
Historical progression The large geographical and tem- poral variations in tobacco use observed currently are best un- derstood in terms of the historical progression that unfolded first in high-income countries such as the United Kingdom and the USA [9,10] and that is now occurring worldwide, especially in men. Manufactured cigarettes were introduced early in the 20th century in the USA and the United Kingdom. They rapidly dis- placed traditional tobacco products such as cigars, pipes, hand-rolled cigarettes, and snuff and by 1920 were the product of choice among new users. The distribution of free cigarettes to soldiers in the First and Second World Wars, combined with mass advertising, greatly in- creased cigarette sales. Women
did not begin smoking regularly un- til the Second World War, when ad- vertising targeted to women and the entry of women into the workforce overcame social and cultural norms against female smoking.
Successive birth cohorts of men, and later women, smoked more cigarettes per day and initi- ated smoking at progressively ear- lier ages. Cigarette sales continued to increase sharply in the United Kingdom until the 1950s and in the USA until the early 1960s, when studies of smoking and mortality and authoritative government re- ports increased concern about the health hazards of smoking, causing a decline in consumption and smok- ing prevalence [10,11]. Continuing publicity about the harmful effects of smoking and the introduction of effective tobacco control policies such as smoke-free laws and in- creased taxes on cigarettes further discouraged smoking. However, the number of deaths caused by smoking continued to increase for several decades after the decline in cigarette sales, due to the ageing of continuing smokers with the great- est lifetime exposure to actively in- haled tobacco smoke.
Stages of the epidemic The progression of the cigarette ep- idemic, first observed in developed countries, has been conceptual- ized as a sequence of four stages that apply more broadly worldwide. These stages (Fig. 2.1.2) are con- ventionally defined by increasing
or decreasing trends in smoking prevalence, and then in smoking- attributable deaths, in a given country. However, most low- and middle-income countries have only cross-sectional data on smoking prevalence and no formal estimates of smoking-attributable deaths. In these countries, the stages can be approximated using cross-sectional measures of adult prevalence in re- lation to longitudinal trends in per capita cigarette sales. The staging of countries discussed below refers only to cigarettes or cigarette-like
0
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
% of deaths caused by smoking
% of smokers among adults
STAGE 1 STAGE 2 STAGE 3 STAGE 4
1960 1900 1920 1940 1980 2000
% male smokers
% male deaths
2020
A
0
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
% of deaths caused by smoking % of smokers
among adults STAGE 1 STAGE 2 STAGE 3 STAGE 4
1960 1900 1920 1940 1980 2000
% female smokers
% female deaths
2020
B
Fig. 2.1.2. Stages of the tobacco epidemic in (A) men and (B) women. For details, see text.
Fig. 2.1.3. Smoking home-grown tobacco wrapped in green leaves or stuffed in a pipe is a tribal tradition in Phak Nam in north-eastern Cambodia, among mem- bers of the Kreung ethnic minority.
A B
84
products such as bidis, although other tobacco products are also dis- cussed. The stages typically differ for men and women in their severity and timing.
Stage 1 represents the begin- ning of the epidemic. At this stage, the prevalence of cigarette smok- ing in adult men or women is low (< 20%), and as yet few deaths can be attributed to smoking. Per capita consumption is flat or only begin- ning to increase, and there is no history of greater smoking in the past. In stage 2, the prevalence of cigarette smoking in adults rises above 20% and may increase to a peak of between 40% and 80%. Per capita cigarette consump- tion also rises. In countries where estimates of tobacco-attributable deaths are available, this percent- age begins to rise as a fraction of all deaths. Stage 3 is characterized by a flattening or downturn in smoking prevalence (and usually in per cap- ita consumption), coinciding with a continuing steep increase in smok- ing-attributable deaths. Stage 4 is characterized by a decline in both smoking prevalence and smoking- attributable deaths [11].
Several factors influence the stage of the epidemic within a given country or region. These include the affordability and availability of cigarettes and the intensity of to- bacco industry marketing efforts. The uptake of smoking among women may be deterred, possibly
indefinitely, by religious and so- cial norms that discourage female smoking, hence limiting the appli- cability of the stages of the epidem- ic model to countries that observe a significant increase in female smoking prevalence. Public edu- cation about the harmful effects of smoking and second-hand smoke and implementation of effective to- bacco control policies also curtail tobacco use. Even within a given region there may be considerable variation in tobacco use because of these factors.
Geographical variations Most of the stages of the cigarette epidemic can be observed in Africa. Women in all countries are in stage 1, as are men in many countries in sub-Saharan Africa [12]. Men in several sub-Saharan African coun- tries, such as Namibia, Botswana, and Sierra Leone, are in stage 2 (adult prevalence > 20%), as are men in all countries in North Africa [4]. The prevalence of cigarette smoking among adult men in North Africa ranges from 28% in Morocco to 53% in Tunisia (stage 2 or 3), not counting the use of water pipes (shisha and hookah) [4]. In Egypt, 38% of adult men smoke cigarettes, and per capita consumption has nearly doubled over the past 20 years [13,14]. Most sub-Saharan African countries, with the excep- tion of South Africa, are positioned on the upslope rather than the downslope of the epidemic curve and remain especially vulnerable to industry marketing [12,15].
In Central and South America, the prevalence of cigarette smok- ing in men generally increases from north to south, ranging from 3% in Suriname and Belize to 30% in Chile [4]. The smoking preva- lence in individual countries does
Fig. 2.1.4. Tobacco advertising has always been common in Formula One racing. However, as tobacco advertising bans take effect, teams have started using an alternative livery that alludes to the tobacco sponsor or eliminates its name entirely.
Fig. 2.1.5. Preparation of paan in Varanasi, India. Paan is a stimulating, psychoactive preparation of betel leaf combined with areca nut and/or cured tobacco. It is chewed and then spat out or swallowed.
Chapter 2.1 • The global tobacco epidemic 85
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
not necessarily follow this gen- eral pattern. Male smoking preva- lence exceeds 40% in Cuba, Peru, and Bolivia and exceeds 30% in Venezuela, Uruguay, and Argentina [4]. Brazil is the only country in South America that may be in stage 4 for men, with decreases in per capita consumption and smoking prevalence in men (22%) and wom- en (13%) due to rigorous tobacco control efforts [13,16]. Cuba has the highest female smoking prevalence in the Caribbean (29%) [4] and also has higher lung cancer death rates in women than those in most of Europe [17].
In North America, northern Europe, Australia, and New Zea- land, men are in stage 4 of the epi- demic, whereas women are gener- ally transitioning from stage 3 to stage 4. In all of these countries, male cigarette smoking prevalence exceeded 50% in past decades [16]. Currently, Sweden has the lowest smoking prevalence in men (13%) and is the only country in Europe where prevalence is higher
in women (16%) than in men [4]. The use of smokeless tobacco (snus) is most common in Sweden but is also seen in other Scandinavian coun- tries [4]. Greece has the highest smoking prevalence in men (63%), and Austria has the highest preva- lence in women (45%) [4].
Countries in the former Soviet Union have some of the highest male cigarette smoking preva- lence rates in the world, whereas smoking prevalence in men has decreased substantially in some countries in eastern Europe. Male smoking prevalence exceeds 50% in the Russian Federation (59%) and Georgia (57%) and remains at 50% in Latvia [4]. Smoking preva- lence among men has decreased in Poland and Ukraine [13]. Although female prevalence does not gen- erally exceed 25% in these coun- tries, it is increasing rapidly in the Russian Federation and to a lesser extent in Ukraine [4]. Per capita con- sumption in the Russian Federation has almost doubled in the past 20 years [14].
Tobacco smoking is uncom- mon among women in the Middle East, except in Lebanon [4]. Men are in stage 1 in Qatar, Oman, Saudi Arabia, and the United Arab Emirates and in stage 3 or late stage 2 in Turkey (48%) [14], Jordan (42%) [4], and the Syrian Arab Republic (36%) [4]. Use of water pipes (shisha, hookah, and narghile) is common throughout the region.
Smoking is uncommon (< 10%) among women in nearly all coun- tries in East and South-East Asia [4]. Among men the highest smok- ing prevalence estimates are in Indonesia (57%) [4] and China (53%) [13]. Per capita cigarette con- sumption is still increasing in many countries (China, Bangladesh, Ne- pal, Viet Nam, Indonesia, and the Philippines) where men account for most of the smoking [14]. Smoking prevalence is decreasing among older women in China. A concern, based on limited anecdotal infor- mation, is that smoking prevalence may be beginning to increase
Fig. 2.1.6. Cigarette consumption and prices in (A) France (1980–2011), (B) Thailand (1990–2011), and (C) South Africa (1961– 2009). THB, Thai baht; ZAR, South African rand.
0
1
2
3
4
5
6
7
0
50
100
150
200
250
300
350
1980 1985 1990 1995 2000 2005 2010
Ci ga
re tt
es C
on su
m ed
P er
A du
lt P
er D
ay
Pr ic
e (%
R el
at iv
e to
1 98
0)
Prices Consumption
A
500
550
600
650
700
750
800
850
900
10
15
20
25
30
35
40
45
50
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
A nn
ua l P
er C
ap it
a Ci
ga re
tt e
Co ns
um pt
io n
(S ti
ck s)
Pr ic
e Pe
r Pa
ck o
f C ig
ar et
te s
(2 00
0 TH
B)
Figure 2b: Cigarette Consumption and Prices in Thailand (1990-2011)
Prices Consumption
B
0
500
1000
1500
2000
2500
0
5
10
15
20
25
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
A nn
ua l P
er C
ap it
a Ci
ga re
tt e
Co ns
um pt
io n
(S ti
ck s)
Pr ic
es P
er P
ac k
(c on
st an
t Z A
R 20
09 )
Figure 2c: Cigarette Consumption and Real Price in South Africa (1961-2009)
Prices Consumption
C
A B
C
86
among young urban Chinese wom- en [18]. Smoking typically begins at young ages, and Chinese women represent a vast untapped market for the tobacco companies.
Tobacco smoking is rare among women in India, Pakistan, and Bangladesh, but the use of smoke- less tobacco by women in these countries is the highest in the world [4]. Nearly half of adult men in Bangladesh (45%) smoke either cig- arettes or bidis [13]. The prevalence of male tobacco smoking is lower in India (24%) and Pakistan (28%) [4], but the use of smokeless products (paan, gutkha, and betel quid) ex- ceeds 30% [13]. Per capita consump- tion of manufactured cigarettes is low in all three countries [14].
Strategies to curtail the epidemic Education about the harms of to- bacco use and policy interventions
advocated by the WHO Framework Convention on Tobacco Control have reduced tobacco use in many countries. For example, increases in excise taxes on cigarettes con- sistently decrease cigarette sales [19,20]. This is exemplified by the decrease in per capita cigarette consumption in Thailand, South Africa, and France, which coincided with large increases in the real price of cigarettes (Fig. 2.1.6) [14,21–26]. Other important strategies that promote cessation and decrease initiation include smoke-free laws, public service counter-advertising, and restrictions on or banning of tobacco industry advertising [27]. Many countries have implemented graphic health warnings on ciga- rette packs. Australia is leading the way on plain packaging (removal of all graphic branding from ciga- rette packs). The widespread ap- plication of a few powerful tobacco
control interventions could prevent a large proportion of the anticipated 450 million deaths from tobacco over the next few decades by en- abling those who are current smok- ers to quit and by persuading youth to never start smoking in the first place [28].
In conclusion, tobacco use re- mains an enormous health prob- lem worldwide. Use is currently decreasing in most high-income countries but increasing or persist- ing at high levels in many low- and middle-income countries. Effective tobacco control strategies such as large periodic increases in excise taxes and the elimination of ad- vertising provide unparalleled op- portunities for cancer prevention worldwide.
Chapter 2.1 • The global tobacco epidemic 87
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
References
1. Glynn T, Seffrin JR, Brawley OW et al. (2010). The globalization of tobacco use: 21 challenges for the 21st century. CA Cancer J Clin, 60:50–61. http://dx.doi. org/10.3322/caac.20052 PMID:20097837
2. Ezzati M, Lopez AD (2003). Estimates of global mortality attributable to smoking in 2000. Lancet, 362:847–852. http://dx.doi. o r g /10.1016 / S 014 0 - 673 6 (0 3)14 3 3 8 - 3 PMID:13678970
3. WHO (2013). WHO Report on the Global Tobacco Epidemic, 2013: Enforcing Bans on Tobacco Advertising, Promotion and Sponsorship. Geneva: WHO. Available at http://apps.who.int/iris/bit stream/10665/85380/1/9789241505871_ eng.pdf.
4. Eriksen M, Mackay J, Ross H (2012). The Tobacco Atlas, 4th ed. Atlanta, GA: American Cancer Society, World Lung Foundation. Available at http://www. tobaccoatlas.org/.
5. IARC (2012). Personal habits and in- door combustions. IARC Monogr Eval Carcinog Risks Hum, 100E:1–575. PMID:23193840
6. WHO (2012). Global Adult Tobacco Survey. Available at http://www.who.int/ to b a c c o /s u r ve i l l a n c e /s u r vey/g at s /e n / index.html.
7. WHO (2012). Global Youth Tobacco Survey. Available at http://www.who.int/ tobacco/surveillance/gyts/en/index.html.
8. WHO (2008). MPOWER: A Policy Package to Reverse the Tobacco Epidemic. Geneva: WHO. Available at h t t p: // w w w.w h o. i n t / t o b a c c o / m p o w e r/ mpower_english.pdf.
9. Pirie K, Peto R, Reeves GK et al.; Million Women Study Collaborators (2013). The 21st century hazards of smoking and benefits of stopping: a prospective study of one million women in the UK. Lancet, 381:133–141. http://dx.doi.org/10.1016/S01 40-6736(12)61720-6 PMID:23107252
10. U.S. Public Health Service (1964). Smoking and Health: Report of the Advisory Committee to the Surgeon General of the Public Health Service. Washington, DC: U.S. Department of Health, Education, and Welfare, Public Health Service (Public Health Service Publication No. 1103).
11. Thun M, Peto R, Boreham J, Lopez AD (2012). Stages of the cigarette epi- demic on entering its second century. Tob Control, 21:96–101. http://dx.doi. org/10.1136/tobaccocontrol-2011-050294 PMID:22345230
12. Pampel F (2008). Tobacco use in sub- Sahara Africa: estimates from the de- mographic health surveys. Soc Sci Med, 66:1772–1783. http://dx.doi.org/10.1016/ j.socscimed.2007.12.003 PMID:18249479
13. Giovino GA, Mirza SA, Samet JM et al.; GATS Collaborative Group (2012). Tobacco use in 3 billion individuals from 16 countries: an analysis of nationally representative cross-sectional household surveys. Lancet, 380:668–679. http://dx. doi.org/10.1016/S0140-6736(12)61085-X PMID:22901888
14. ERC (2010). World Cigarette Reports 2010. Suffolk, UK: ERC Group.
15. Blecher E (2010). Targeting the afford- ability of cigarettes: a new benchmark for taxation policy in low-income and middle- income countries. Tob Control, 19:325–330. http://dx.doi.org/10.1136/tc.2009.030155 PMID:20530141
16. OECD (2012). Non-medical determinants of health. OECD Health Statistics data- base. Paris: Organisation for Economic Co-operation and Development. http:// dx.doi.org/10.1787/20758480
17. WHO (2012). Cancer Mortality Database. Available at http://www-dep.iarc.fr/WHO db/WHOdb.htm.
18. Ho MG, Ma S, Chai W et al. (2010). Smoking among rural and urban young women in China. Tob Control, 19:13–18. http://dx.doi.org/10.1136/tc.2009.030981 PMID:19822528
19. WHO (2003). WHO Framework Conven- tion on Tobacco Control. Available at http://www.who.int/fctc/en/.
20. IARC (2011). IARC Handbooks of Cancer Prevention, Vol. 14: Tobacco Control: Effectiveness of Tax and Price Policies for Tobacco Control. Lyon: IARC.
21. Economist Intelligence Unit (2011). Worldwide Cost of Living Survey. London: The Economist Group.
22. Economist Intelligence Unit (2011). Marlboro cigarette and local cigarette prices, Worldwide Cost of Living Survey. London: The Economist Group.
23. International Monetary Fund (2011). World Economic Outlook Database, April 2011 edition. Available at http://www. imf.org /ex ter nal /pubs / f t / weo/ 2011/ 01/ weodata/index.aspx.
24. Levy DT, Benjakul S, Ross H, Ritthiphakdee B (2008). The role of to- bacco control policies in reducing smok- ing and deaths in a middle income na- tion: results from the Thailand SimSmoke simulation model. Tob Control, 17:53–59. http://dx.doi.org/10.1136/tc.2007.022319 PMID:18218810
25. Blecher EH (2011). The Economics of Tobacco Control in Low- and Middle- Income Countries [thesis]. Cape Town, South Africa: School of Economics, University of Cape Town.
26. Guérin S, Hill C (2010). Cancer epidemiol- ogy in France in 2010: comparison with the USA [in French]. Bull Cancer, 97:47–54. http://dx.doi.org/10.1684/bdc.2010.1013 PMID:19995688
27. Blecher E (2008). The impact of to- bacco advertising bans on consumption in developing countries. J Health Econ, 27:930–942. http://dx.doi.org/10.1016/j. jhealeco.2008.02.010 PMID:18440661
28. Jha P (2009). Avoidable global cancer deaths and total deaths from smoking. Nat Rev Cancer, 9:655–664. http://dx.doi. org/10.1038/nrc2703 PMID:19693096
88
Summary • Tobacco smoke contains more
than 7000 chemical compounds, of which many are known carcin- ogens. Components of tobacco smoke contribute to carcinogen- esis through multiple pathways, including DNA binding and mu- tations, inflammation, oxidative stress, and epigenetic changes.
• Smokeless tobacco products contain more than 3000 chemi- cals and numerous carcinogens. DNA binding and mutations are among the mechanisms clearly implicated in carcinogenesis due to smokeless tobacco use.
• Epidemiological studies have shown causal associations be- tween tobacco smoking and at least 14 different types of cancer to date. Accumulated evidence from epidemiological studies so far has been suf- ficient to show that smokeless tobacco causes cancers of the oral cavity and pancreas.
• Comprehensive tobacco con- trol strategies are in place at the global level through provi- sions in the WHO Framework Convention for Tobacco Control to combat the epidemic of to- bacco use.
Worldwide, tobacco is used in di- verse smoked or smokeless prod- ucts, all delivering nicotine to their users. The worldwide epidemic of tobacco-caused diseases comes largely from widespread smoking of manufactured cigarettes, now mostly manufactured and distrib- uted by a small number of multi- national corporations and, in the case of China, the China National Tobacco Corporation. In some countries, particularly India and Bangladesh, smokeless tobacco use is prominent, especially among women, even though smoking is rare among them. The potency of tobacco products, particularly cig- arettes, as a cause of cancer and their role in causing cardiovascular and lung diseases make tobacco use the leading cause of avoidable premature mortality worldwide, causing an estimated 6 million deaths annually [1].
Tobacco smoking Manufactured cigarettes are highly engineered for the purpose of deliv- ering nicotine-containing smoke to their users. They burn at high tem- perature, generating thousands of chemicals, and consequently smok- ers inhale a highly toxic mixture con- taining many known carcinogens and toxins, such as benzo[a]pyr- ene and other polycyclic aromatic
hydrocarbons, tobacco-specific ni- trosamines (N-nitroso derivatives of nicotine and its metabolites), benzene (a cause of leukaemia), formaldehyde (an irritant and car- cinogen), carbon monoxide and cy- anide (asphyxiants), acrolein (an ir- ritant), and polonium (a radioactive carcinogen) [2]. The smoke inhaled by the smoker is referred to as mainstream smoke, and that emit- ted from the smouldering cigarette is called sidestream smoke. In the presence of smoking, nonsmokers inhale second-hand smoke, made up largely of sidestream smoke and also some exhaled mainstream smoke. Pipes, cigars, and water pipes deliver smoke with mixtures of components comparable to those from cigarettes, although there are differences in smoke characteris- tics across products.
Smokeless tobacco Smokeless tobacco is consumed in ways apart from the tobacco being burnt; relevant products are most commonly used orally and some- times nasally. Oral smokeless tobac- co use delivers nicotine more slowly than cigarette smoking does – the peak increase in plasma levels of nicotine in smokeless tobacco users is 30 minutes after intake – but the level remains higher longer because it declines slowly while the product is
2.2 2 ETIOLOGY
Tobacco smoking and smokeless tobacco use Jonathan M. Samet Prakash C. Gupta Cecily S. Ray
Catherine Sauvaget (reviewer) Deborah M. Winn (reviewer)
Chapter 2.2 • Tobacco smoking and smokeless tobacco use 89
2 E
T IO
LO G
Y C
H A
P T
E R
2 .2
held in the mouth. Over the course of a day, a smokeless tobacco user may ingest twice as much nicotine as a smoker. Smokeless tobacco comes in many forms, with names specific to the countries of origin, but the products may be generally classified as loose leaf for chew- ing or twist (sweetened, flavoured strands), flakes, shredded or fine- cut tobacco for chewing or holding in the mouth, solid compressed tobac- co products in the form of chunks or sticks, viscous pastes, dry or moist ground tobacco (snuff) for oral or nasal use, liquid snuff (an extract) for nasal use in Eastern Africa, and tobacco-smoke water for gargling in South Asia. Newer products include fine-cut or powdered tobacco (i.e. snuff) in teabag-like pouches, dis- solvable tobacco lozenges, and to- bacco toothpicks.
Some smokeless tobacco prod- ucts are plain tobacco leaf. Most products are flavoured: sweetened with sugar, molasses, or artificial sweeteners, salted, and/or aro- matized with substances like es- sences, spices, and perfumes. Menthol is often used to prevent irritation. Smokeless tobacco is used along with areca nut by many people of South-East Asia, both in their countries of origin and wher- ever they have migrated. The user may mix tobacco with areca nut or may purchase a pre-mixed product. Areca nut contributes its own set of carcinogens, mainly through areca nut-specific nitrosamines [3].
Tobacco-specific nitrosamines are relevant to both smoking and smokeless tobacco use but are of particular prominence in the use of smokeless tobacco because poly- cyclic aromatic hydrocarbons are generated by combustion and are therefore not implicated in smoke- less tobacco carcinogenesis. These compounds are formed by the nitro- sation of nicotine and its metabolites. The nature and amount of these compounds depends on the tobacco species, the growing conditions, the amount of fermentation and ageing, and how finely the tobacco is cut or powdered. Smokeless tobacco
products, as with many cigarette brands, usually have an ingredient that increases the pH, to promote the availability of nicotine for absorption. Such agents include ammonia, am- monium carbonate, calcium hydrox- ide, potassium hydroxide, potassium carbonate, and sodium carbonate. Nicotine availability also increases with finer powdering of the tobacco [3]. Products with different levels of free nicotine are marketed to differ- ent types of users; novices usually start with low-nicotine products and may graduate to brands with higher levels [4].
Patterns of use Cigarettes The Global Adult Tobacco Survey and the Global Youth Tobacco Survey, both implemented by WHO and the United States Centers for Disease Control and Prevention in collaboration with participating countries, track the use of tobacco worldwide. The first global report based on the Global Adult Tobacco Survey was published in 2012, drawing on findings from 14 low- and middle-income countries and from the USA and the United Kingdom [5]. The findings showed a wide range
of smoking prevalence among the 19 countries and far higher rates of smoking among men than women, although in several countries the prevalence of smoking among women approached that among men (Fig. 2.2.1). In some countries, exemplified by China and India, only a small percentage of ever- smokers had stopped smoking. This survey, which covered most of the high-burden countries, documented tobacco use by about 850 million people. Among those aged 24– 35 years who were interviewed for this survey, the mean age of smoking initiation was 17.7 years in men and 18.0 years in women [5]. For those aged 35 years or older, the average age of initiation was generally older in women than men.
The Global Youth Tobacco Survey data collected from adoles- cents aged 13–15 years in schools in most countries show that youth con- tinue to smoke worldwide [6]. In some countries more than 30% of boys and girls in this age range have smoked in the past month and the prevalence of smoking among girls is approach- ing that among boys. Countries with the highest rate of smoking among boys include Papua New Guinea (50.6% prevalence), Timor-Leste
0 10 20 30 40 50 60 70
USA United Kingdom
Uruguay Mexico
Brazil Turkey Egypt
Poland Ukraine
Russian FederaHon Thailand
Philippines Viet Nam
Bangladesh India China
Males
Females
Fig. 2.2.1. Prevalence of current tobacco smoking among male and female adults aged ≥ 15 years in 14 Global Adult Tobacco Survey countries, the United Kingdom, and the USA.
90
(50.6%), and Tonga (37.5%), whereas the highest rates of smoking among girls were reported in Chile (39.9%), Papua New Guinea (35.8%), and the Czech Republic (32.7%) [6].
Smokeless tobacco Smokeless tobacco is used orally in all regions of the world, but its use is most prevalent in the Pacific Islands, South and South-East Asia, Africa, and Europe [6]. Countries with the highest known prevalence of smoke- less tobacco use include Marshall Islands, Myanmar, Bangladesh, India, and Madagascar (prevalence 20% and above). These are fol- lowed by Bhutan, Nepal, Sweden, Sri Lanka, Sudan, Turkmenistan, Federated States of Micronesia, Uzbekistan, Yemen, and Norway (prevalence 10–19%). Fig. 2.2.2 shows the relative position of coun- tries according to national prevalence of adult smokeless tobacco use. Such data are not yet available for many countries, especially in Africa. Products are exported from the USA, Sweden, and India to many destina- tions around the world. Several of the largest cigarette manufacturers are now marketing smokeless tobacco products to smokers as extensions of their cigarette brands, pitching them as products to be used in smoke- free environments [6]. Prevalence figures for youth who use smokeless tobacco are not available from most countries at present. However, the Global Youth Tobacco Survey shows a high proportion of non-cigarette tobacco use among youth in many countries, and a large part of this use is believed to be smokeless tobacco [7]. In Congo in 2006, 18.0% of youth aged 13–15 years had used smoke- less tobacco in the previous 30 days [8]. In India in 2006, the correspond- ing prevalence was 8.1% [9].
Tobacco as a cause of cancer Mechanisms of cancer causation Tobacco smoking Tobacco smoke is a mixture of gas- es and particles sufficiently small to
reach and deposit in the bronchioles and alveoli. Tobacco smoke compo- nents, including nicotine, move from the lungs into the circulation and reach throughout the body, resulting in various tissues being exposed to differing levels of carcinogens, not simply in organs where primary ab- sorption occurs but throughout most of the body.
Tobacco smoke contains more than 7000 chemical compounds, including numerous known car- cinogens, as evaluated for carcino- genicity by IARC [10]. Broad classes of carcinogens in tobacco smoke include polycyclic aromatic hydro- carbons, N-nitrosamines, and aro- matic amines, together with a range of tobacco-specific toxins, includ- ing volatile aldehydes and phenolic amines.
Fig. 2.2.3 provides a general schema for the causation of can- cer by the carcinogens in tobacco smoke. There are both specific and nonspecific pathways by which smoking causes cancer, and there is very extensive experimental evi- dence documenting these pathways [2,10]. The figure begins with initia- tion of cigarette smoking and addic- tion; it is the maintained contact with tobacco smoke carcinogens result- ing from addiction that leads to the very high risk of cancer in those who smoke daily throughout their lives, the general pattern of the ad- dicted smoker. It is not necessary to smoke across the full lifetime to
have excess risk of cancer from smoking. Even a decade or two of smoking increases risk, and cancer risk for former smokers, although re- duced, does not drop to that of nev- er-smokers. The figure highlights the multiple processes that lead to un- controlled cell growth and malignan- cy and the multiple points in these processes at which tobacco smoke components contribute to carcino- genesis. The major pathway to can- cer is considered to be DNA binding and consequent mutations (central pathway), but tobacco smoke also contributes to increased cancer risk through inflammation (top pathway) and epigenetic mechanisms (bottom pathway) [10].
Smokeless tobacco Smokeless tobacco contains more than 3000 chemicals and at least 28 carcinogens, many the same as those contained in cigarette smoke, and the general scheme of Fig. 2.2.3 is applicable. Smokeless tobacco use results in exposure to tobacco-specific nitrosamines, vol- atile N-nitrosamines, N-nitrosamino acids, and volatile aldehydes such as formaldehyde and acetalde- hyde, as well as metals including cadmium, lead, arsenic, nickel, and chromium, and radioactive elements. Certain tobacco-relat- ed nitrosamines (e.g. nicotine- derived nitrosamine ketone and N′-nitrosonornicotine), their me- tabolites, and benzo[a]pyrene are
0
10
20
30
40
50
60
70
M ar sh all Is lan
ds
M ya nm
ar
Ba ng lad
es h
Ind ia
M ad ag as ca r
Bh ut an
Ne pa l
Sw ed en
Sr i L an ka
Su da n
Tu rkm
en ist an
M icr on es ia
Uz be kis ta n
Ye me
n
No rw ay
La o P
DR
Sie rra
Le on e
Pa ra gu ay
M oz am
biq ue
M ald
ive s
So ut h A
fri ca
Al ge ria
Le so th o
Pa kis ta n
Th ail an d
US A
Pr ev al en
ce (%
)
Fig. 2.2.2. Prevalence of smokeless tobacco use among adults aged ≥ 15 years.
Chapter 2.2 • Tobacco smoking and smokeless tobacco use 91
2 E
T IO
LO G
Y C
H A
P T
E R
2 .2
Genetic susceptibility to tobacco-related cancers James D. McKay
The first-degree relatives of pa- tients with tobacco-related cancers have an excess risk of developing the same type of cancer. While a shared environmental exposure such as tobacco could influence such observations, this finding also suggests a role for genetic suscep- tibility [1]. Susceptibility genes have indeed been identified and encom- pass both rare and common genet- ic variants.
Several rare cancer syndromes, such as Li–Fraumeni syndrome (OMIM 151623) and dyskeratosis congenita (OMIM 127550), linked with mutations in the TP53 and TERT genes, respectively, include tobacco-related cancers in their disease spectrum, which implicates such genes in genetic susceptibil- ity. Similarly, a rare familial form of lung adenocarcinoma appears to be linked with germline mutations in EGFR [2].
G e n o m e - w i d e a s s o c i a t i o n study (GWAS) approaches have identified several other intriguing loci with much more common popu- lation frequencies (Fig. B2.2.1). At chromosome 15q25, a locus that contains several nicotinic acetyl- choline receptors, the same vari- ants linked with lung cancer are also linked with propensity to smoke, leading to the debated suggestion that this association could be medi- ated by behavioural changes [3,4]. Two independent common variants at 5p15.33, a region containing the TERT gene, have been linked with
lung cancer [5]. These variants, al- though in close proximity, appear to differ in their effects in terms of histology, with one being more rel- evant to adenocarcinoma [6]. This same locus has been subsequently linked with genetic susceptibility to many other cancer types [7].
GWAS and other approaches have identified genetic variants at several other lung cancer suscepti- bility loci; the chromosome 6 major histocompatibility complex (MHC) region [3], which is critical to im- mune response; and several genes involved in DNA repair and cell-cy- cle regulation, notably CHEK2 [8], RAD52 [9], and CDNK2A [9], all of which appear particularly relevant to lung squamous cell carcinoma (Fig. B2.2.1). Several additional lung cancer susceptibility loci have been identified in Asian populations [10], although their involvement in lung cancer in other populations re- mains unclear. Within cancers of the head and neck, the genes involved in alcohol metabolism (ADH and ALDH2) appear particularly relevant [11].
It is also notable that the ge- netic studies carried out so far have identified fewer susceptibility loci in tobacco-related cancers compared with other common cancers, de- spite the comparably sized genetic studies and similar estimations of hereditary components. The reason for this remains unclear, but some of the unexplained risk may be ex- plained by components of genetic
variation not tested extensively to date, such as rare population fre- quencies and copy number repeats, with the latter potentially important particularly for metabolism genes. Further studies are also required on gene–environment interactions.
References 1. Brennan P et al. (2011). Lancet Oncol,
12:399 – 408. http://dx.doi.org/10.1016/ S1470-2045(10)70126-1 PMID:20951091
2. Ohtsuka K et al. (2011). J Clin Oncol, 29:e191–e192. http://dx.doi.org/10.1200/ JCO.2010.31.4492 PMID:21172876
3. Hung RJ et al. (2008). Nature, 452:633– 637. http://dx.doi.org/10.1038/nature06885 PMID:18385738
4. Chen D et al. (2011). Cancer Epidemiol Biomarkers Prev, 20:658–664. http:// dx.doi.org/10.1158/1055 -9965.EPI-10 - 1008 PMID:21335511
5. McKay JD et al. (2008). Nat Genet, 40:1404–1406. http://dx.doi.org/10.1038/ ng.254 PMID:18978790
6. Truong T et al. (2010). J Natl Cancer Inst, 102:959–971. http://dx.doi.org/10.1093/ jnci/djq178 PMID:20548021
7. Zou P et al. (2012). BMC Cancer, 12:7. http://dx.doi.org/10.1186/1471-2407-12-7 PMID:22221621
8. Brennan P et al. (2007). Hum Mol Genet, 16:1794–1801. http://dx.doi.org/10.1093/ hmg/ddm127 PMID:17517688
9. Timofeeva MN et al. (2012). Hum Mol Genet, 21:4980–4995. http:// d x . d o i . o r g / 1 0 .1 0 9 3 / h m g / d d s 3 3 4 PMID:22899653
10. Dong J et al. (2013). PLoS Genet, 9:e1003190. http://dx.doi.org/10.1371/ journal.pgen.1003190 PMID:23341777
11. McKay JD et al. (2011). PLoS Genet, 7:e1001333. http://dx.doi.org/10.1371/ journal.pgen.1001333 PMID:21437268
Fig. B2.2.1. Manhattan plot of lung cancer genome-wide association study results, overall (A) and restricted to adenocarcinoma (B) and squamous cell carcinoma (C).
92
known to attach to cellular DNA, leading to certain mutations in on- cogenes and tumour suppressor genes, which are found in oral pre- malignant lesions associated with smokeless tobacco use. Smokeless tobacco also generates reactive oxygen species, oxidative stress, and associated DNA fragmenta- tion in laboratory experiments. Inflammation of the oral mucosa at the site of tobacco quid placement may begin as early as 2–7 days after the initial application by regular us- ers on a new site. A precise role for inflammation caused by smokeless tobacco use is not clear at present.
Epidemiological evidence Tobacco smoking The epidemiological evidence on smoking and cancer comes from numerous case–control and cohort studies carried out since the mid- 20th century. The epidemiological evidence on smoking and cancer is consistent in identifying cigarette smoking as a cause of many types of cancer [11,12]. Fig 2.2.4 identifies the cancer sites and types for which IARC [12] and/or the United States Surgeon General [11] have concluded that the relationship is causal. The affected sites include those where smoke is directly deposited (e.g. the orophar- ynx and lung) and distal sites that are reached by circulating tobacco smoke components (e.g. the pancreas and
urinary bladder). Overall, risks of can- cers caused by smoking increase with duration of smoking and with number of cigarettes smoked daily; cancer risk falls after successful ces- sation of smoking, but for longer-term smokers, the risks do not completely drop to those of never-smokers [11]. Associations between active smok- ing and breast cancer continue to be investigated [13].
Table 2.2.1 provides estimates of relative risks of death from smoking-re- lated cancers for major sites in two co- hort studies carried out in the USA by the American Cancer Society: Cancer Prevention Study I (CPS-I) (enrolment in 1959, follow-up through 1965) and CPS-II (enrolment in 1982, follow-up through 1988) [14]. Questionnaires administered at enrolment provided the data on participants’ smoking sta- tus. Several general findings are no- table: (i) the relative risks for current and former smokers compared with never-smokers are remarkably high for some sites (e.g. lung and laryngeal cancer), (ii) former smokers uniformly have decreased relative risks com- pared with current smokers, (iii) rela- tive risks have tended to be lower in women than in men, and (iv) relative risks tended to increase for women over the two decades between the two studies [14]. Findings of more recent studies suggest that relative risks have continued to rise as more recent cohorts of women have started to smoke at a similarly young age as men and have smoked at comparable rates for similar periods [15]. The rela- tive risks in Table 2.2.1 reflect those in high-income countries. In low- and middle-income countries, the relative risks tend to be lower because heavy smoking beginning at a young age – the pattern in high-income countries – has not been typical of smokers to date.
Epidemiological evidence also shows that involuntary smoking, the inhalation of second-hand smoke by nonsmokers, causes cancer. The first epidemiological studies on in- voluntary smoking and lung cancer risk in nonsmokers were published in 1981; by 1986, there was sufficient evidence, particularly in the context of the already extensive literature on active smoking, to conclude that invol- untary smoking causes lung cancer in nonsmokers [16,17]. Exposure to involuntary smoking increases lung cancer risk by about 25%, a finding replicated worldwide [10].
Smokeless tobacco Numerous epidemiological studies, including cohort and case–control
DNA adducts
Mutations in oncogenes and tumour suppressor
genes
Loss of normal growth control
mechanisms
Cancer
Protein kinase A and B activation and other changes
Tumour suppressor gene inactivation and
other changes
Uptake of co-carcinogens and tumour promoters
Gene promoter hypermethylation
Uptake of carcinogens
Initiation of cigarette smoking/ nicotine addiction
Receptor binding
Regular cigarette
smoking
Persistence
Miscoding
Metabolic
activation
Excretion
Metabolic detoxification
Normal DNA
Apoptosis
Repair
Fig. 2.2.3. Pathway for causation of cancer by carcinogens in tobacco smoke.
Fig. 2.2.4. Cancer sites and types for which IARC [12] and/or the United States Surgeon General [11] have concluded that the relationship with tobacco smoking is causal. Note: For cancers of the liver, colon, rectum, ovary (mucinous), and paranasal sinuses, causal conclusions reached by IARC only.
Chapter 2.2 • Tobacco smoking and smokeless tobacco use 93
2 E
T IO
LO G
Y C
H A
P T
E R
2 .2
studies adjusted or stratified for smoking and alcohol intake, have also shown a causal relationship of smokeless tobacco use with can- cer [3]. At this time, the evidence is strongest for cancers of the oral cavity and pancreas. Traditional products from the USA, South Asia, and Sudan have been found to contain very high amounts of tobacco-specific nitrosamines, and users have high risk of oral cancer, as shown in several studies. The level of risk experienced by us- ers depends largely on the type of product they are using. Risk esti- mates for cancer associated with smokeless tobacco use in case– control studies increase with fre- quency of use per day and with du- ration of use in years, and women smokeless tobacco users tend to have higher risks, compared with non-users, than men. Smokeless tobacco users often have pre- cancerous oral mucosal lesions, such as leukoplakia. These white lesions tend to disappear within 1–2 months after discontinuation of tobacco use. These lesions are more common in smokers who also use smokeless tobacco, and such individuals also have a higher risk of cancer.
Prevention Tobacco smoking Tobacco control has had a lengthy evolution that has been closely linked to the increasing evidence on the health effects of active and involuntary smoking and on which tobacco control modalities are ef- ficacious [4,18]. Historically, the initial findings on lung cancer and smoking were followed by efforts to educate the public about the risks of smoking, with the expectation that they would stop. Since then, we have learned that tobacco con- trol requires far more complex ap- proaches that acknowledge the hi- erarchy of factors that determine the use of tobacco and the interplay of these factors across the life-course, as health is damaged by smoking from conception onward. At each age, the emphasis of tobacco con- trol shifts, moving from preventing initiation to promoting successful cessation. Particular issues include avoiding exposure to second-hand smoke and discouraging pregnant women from smoking. In addition, tobacco control efforts need to be dynamic in time, changing as the to- bacco industry attempts to counter control measures.
Many nations have now imple- mented tobacco control programmes.
Most importantly, there is now the WHO Framework Convention on Tobacco Control (FCTC), the first trea- ty negotiated under the auspices of WHO, which brings a global approach to the global epidemic of tobacco use. The WHO FCTC has been in force since 2005, and most nations of the world have become Parties. Key ele- ments of most national tobacco con- trol policies coincide with core WHO FCTC provisions, including a com- prehensive ban on tobacco advertis- ing, promotion, and sponsorship; a ban on misleading descriptors such as “light”; and a mandate to place rotating warnings that cover at least 30% of tobacco packaging and en- couragement for even larger, graphic
Table 2.2.1. Age-adjusted relative risks of death from smoking-related cancers from the Cancer Prevention Study I (CPS-I) and CPS-II, adults aged ≥ 35 years
Disease category (ICD-9 code)
CPS-I (1959–1965) CPS-II (1982–1988)
Men Women Men Women
CS FS CS FS CS FS CS FS
Lip, oral cavity, pharynx (140–149) 6.3 2.7 2.0 1.9 10.9 3.4 5.1 2.3
Oesophagus (150) 3.6 1.3 1.9 2.2 6.8 4.5 7.8 2.8
Stomach (151) 1.8 1.7 1.0 1.0 2.0 1.5 1.4 1.3
Pancreas (157) 2.3 1.3 1.4 1.4 2.3 1.2 2.3 1.6
Larynx (161) 10.0 8.6 3.8 3.1 14.6 6.3 13.0 5.2
Trachea, bronchus, lung (162) 11.4 5.0 2.7 2.6 23.3 8.7 12.7 4.5
Cervix uteri (180) NA NA 1.1 1.3 NA NA 1.6 1.1
Urinary bladder (188) 2.9 1.8 2.9 2.3 3.3 2.1 2.2 1.9
Kidney, other urinary (189) 1.8 1.8 1.4 1.5 2.7 1.7 1.3 1.1
Acute myeloid leukaemia (204–208) 1.6 1.6 1.0 1.0 1.9 1.3 1.1 1.4
CS, current smokers; FS, former smokers; ICD-9, International Classification of Diseases, 9th Revision; NA, not applicable.
Fig. 2.2.5. Chewing tobacco has been proven to cause oesophageal, oral, blad- der, and pancreatic cancers.
94
warnings. The WHO FCTC also has provisions requiring Parties to imple- ment smoke-free workplace laws, address tobacco smuggling, and in- crease tobacco taxes.
In 2008, WHO identified six ev- idence-based tobacco control mea- sures that are the most effective in reducing tobacco use [19]. Known as “MPOWER”, these measures correspond to one or more of the de- mand reduction provisions included in the WHO FCTC: Monitor tobacco use and prevention policies, Protect people from tobacco smoke, Offer help to quit tobacco use, Warn peo- ple about the dangers of tobacco, Enforce bans on tobacco advertis- ing, promotion, and sponsorship, and Raise taxes on tobacco. WHO is tracking in-country implementa- tion of MPOWER and coverage of the world’s population by its provi- sions. To date, its reach is still lim- ited, although increasing coverage of the world’s population by its ele- ments can be anticipated [1]. Global tobacco control has been supported by governments as well as from nongovernment sources, including funding from the Bloomberg Family Foundation and the Bill & Melinda Gates Foundation, first made avail- able in 2007 and now slated to con- tinue through 2016.
Information about the impact of initiatives designed to encourage individuals to cease smoking is pre- sented in Chapter 4.1.
Smokeless tobacco Tobacco control needs to include smokeless tobacco in its purview. Each aspect of tobacco control has to take into consideration both smoking and smokeless tobacco use. Hence, taxation needs to cover both forms equally so that people do not simply switch over to cheaper smokeless forms after an increase in cigarette taxation, as has happened in some countries. Information cam- paigns need to explicitly address the dangers of both types of tobacco, so that smokeless tobacco is not seen as risk-free. The risk of oral cancer is very high for smokeless tobacco users, especially if use is initiated at
an early age or if the same individu- als are also smokers. The impact of unbranded products has been recognized in several countries and warrants action. Prohibition of spit- ting in public places, prohibition of smokeless tobacco use at the workplace, and bans on the sale of all tobacco products in and around educational institutions are options to control smokeless tobacco use. The same stringent requirements on health warnings and prohibition of advertisements, including use of tobacco-related imagery and ex- tending the tobacco brand to other products (i.e. brand stretching), are appropriate for smokeless tobacco as for smoked products. Health-care providers may be informed on how to encourage and support smoke- less tobacco users to quit. Bans on import of smokeless tobacco and control of smuggling are also ap- propriate and required to control smokeless tobacco use.
In areas where smokeless to- bacco is widely used, such as South-East Asia, the acceptance of tobacco use as “normal” needs to be addressed along with rais- ing public awareness of the health hazards and offering help for quit- ting [20]. For example, successful interventions with student youth in India have been formulated based on a study of the characteristics and needs of target groups. These tar- geted interventions have been able to engage the students’ interest. The interventions have involved the stu- dents in interactive classroom and home-based activities (with parents) as well as in activism for tobacco control. Significantly reduced indica- tors of tobacco use compared with non-intervention schools have been achieved [21,22]. Researchers have been sharing their systematically tested and well-documented con- cepts and materials with the gov- ernment for application across the country in government schools [21]. In community interventions, use of multiple types of media with pretest- ed messages to diffuse information at mass level and in small groups as well as by one-to-one interaction with
health educators has been found to be reinforcing and effective [23,24]. Inclusion of oral examination in com- munity and worksite interventions has also been found to be effective for promoting cessation and reduc- tion of tobacco use among adults, as well as for implementing interven- tions that lead to regression of leuko- plakia in tobacco users [25,26].
Conclusion Tobacco use in both smoked and smokeless forms remains the world’s leading cause of cancer morbidity and mortality. Numerous mechanis- tic pathways have been implicated in tobacco-induced carcinogenesis. For multiple sites, including lung cancer in smokers and oral cancer in smokeless tobacco users, risks are remarkably high compared with other circumstances involving cancer de- velopment, and causality is proven. Thus, tobacco control provides a tremendous opportunity for cancer prevention. A global public health treaty, the WHO FCTC, calls for the ratifying nations to implement a set of effective tobacco control policies. In force since 2005 and ratified by 176 nations to date, the FCTC has cre- ated a global tobacco control move- ment that continues to work towards preventing initiation and promoting successful cessation of tobacco use.
Fig. 2.2.6. Areca nut chewing is popular in Papua New Guinea. When the nut’s fibrous inner core is chewed, a red colour is produced that stains the teeth.
Chapter 2.2 • Tobacco smoking and smokeless tobacco use 95
2 E
T IO
LO G
Y C
H A
P T
E R
2 .2
References
1. WHO (2011). WHO Report on the Global Tobacco Epidemic, 2011: Warning about the Dangers of Tobacco. Geneva: WHO.
2. U.S. Department of Health and Human Services (2010). How Tobacco Smoke Causes Disease: The Biology and Behavioral Basis for Smoking-Attributable Disease: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
3. IARC (2007). Smokeless tobacco and some tobacco-specific N-nitrosamines. IARC Monogr Eval Carcinog Risks Hum, 89:1–592. PMID:18335640
4. U.S. Department of Health and Human Services (2000). Reducing Tobacco Use: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
5. Giovino GA, Mirza SA, Samet JM et al.; GATS Collaborative Group (2012). Tobacco use in 3 billion individuals from 16 countries: an analysis of nationally repre- sentative cross-sectional household sur- vey. Lancet, 380:668–679. http://dx.doi. o r g /10.1016 / S 014 0 - 673 6 (12) 610 8 5 -X PMID:22901888
6. Eriksen M, Mackay J, Ross H (2012). The Tobacco Atlas, 4th ed. Atlanta, GA: American Cancer Society, World Lung Foundation.
7. Sinha DN, Palipudi KM, Rolle I et al. (2011). Tobacco use among youth and adults in member countries of South-East Asia region: review of findings from surveys under the Global Tobacco Surveillance System. Indian J Public Health, 55:169– 17 6 . h t t p: //d x . d o i . o r g /10 . 410 3 / 0 019 - 557X.89946 PMID:22089684
8. Rudatsikira E, Muula AS, Siziya S (2010). Current use of smokeless tobacco among adolescents in the Republic of Congo. BMC Public Health, 10:16. http:// d x . d o i . o r g /10 .118 6 /1471 - 2 4 5 8 -10 -16 PMID:20074362
9. Sinha DN, Gupta PC, Gangadharan P (2007). Tobacco use among students and school personnel in India. Asian Pac J Cancer Prev, 8:417– 421. PMID:18159980
10. IARC (2004). Tobacco smoke and in- voluntary smoking. IARC Monogr Eval Carcinog Risks Hum, 83:1–1438. PMID:15285078
11. U.S. Department of Health and Human Services (2004). The Health Consequences of Smoking: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
12. IARC (2012). Personal habits and in- door combustions. IARC Monogr Eval Carcinog Risks Hum, 100E:1–575. PMID:23193840
13. Gaudet MM, Gapstur SM, Sun J et al. (2013). Active smoking and breast cancer risk: original cohort data and meta-anal- ysis. J Natl Cancer Inst, 105:515–525. h t t p: / /d x . d o i . o r g /10 .10 9 3 / j n c i /d j t 0 2 3 PMID:23449445
14. National Cancer Institute (1997). Changes in Cigarette-Related Disease Risks and Their Implications for Prevention and Control. Smoking and Tobacco Control Monograph No. 8. Bethesda, MD: U.S. Department of Health and Human Services, National Institutes of Health, National Cancer Institute (NIH Publication No. 97-4213).
15. Thun MJ, Carter BD, Feskanich D et al. (2013). 50-year trends in smoking- related mortality in the United States. N Engl J Med, 368:351–364. http:// d x . d o i . o r g / 10 .10 5 6 / N E J M s a12 1112 7 PMID:23343064
16. IARC (1986). Tobacco smoking. IARC Monogr Eval Carcinog Risk Chem Hum, 38:1– 421. PMID:3460963
17. U.S. Department of Health and Human Services (1986). The Health Consequences of Involuntary Smoking: A Report of the Surgeon General. Washington, DC: U.S. Department of Health and Human Services, Public Health Service, Office on Smoking and Health.
18. Wipfli H, Samet JM (2009). Global eco- nomic and health benefits of tobacco control: part 1. Clin Pharmacol Ther, 86:263–271. http://dx.doi.org/10.1038/ clpt.2009.93 PMID:19536067
19. WHO (2008). WHO Report on the Global Tobacco Epidemic, 2008: The MPOWER Package. Geneva: WHO.
20. Kakde S, Bhopal RS, Jones CM (2012). A systematic review on the social context of smokeless tobacco use in the South Asian population: implications for public health. Public Health, 126:635–645. http:// dx.doi.org/10.1016/j.puhe. 2012.05.0 02 PMID:22809493
21. Arora M, Stigler MH, Reddy K (2011). Effectiveness of health promotion in pre- venting tobacco use among adolescents in India: research evidence informs the National Tobacco Control Programme in India. Glob Health Promot, 18:9–12. http:// dx.doi.org/10.1177/1757975910393163 PMID:21721292
22. Sorensen G, Gupta PC, Nagler E, Viswanath K (2012). Promoting life skills and preventing tobacco use among low- income Mumbai youth: effects of Salaam Bombay Foundation intervention. PLoS One, 7:e34982. http://dx.doi.org/10.1371/ journal.pone.0034982 PMID:22523567
23. Aghi MB, Gupta PC, Bhonsle RB, Murti PR (1992). Communication strategies for intervening in the tobacco habits of rural populations in India. In: Gupta PC, Hamner JE III, Murti PR, eds. Control of Tobacco-Related Cancers and Other Diseases. Proceedings of an International Symposium, TIFR, Bombay, January 15–19, 1990. Bombay: Oxford University Press, pp. 303–306.
24. Anantha N, Nandakumar A, Vishwanath N et al. (1995). Efficacy of an anti-tobacco community education program in India. Cancer Causes Control, 6:119–129. ht tp://dx.doi.org /10.10 07/ BF0 0 0 52772 PMID:7749051
25. Gupta PC, Mehta FS, Pindborg JJ et al. (1986). Intervention study for primary prevention of oral cancer among 36 000 Indian tobacco users. Lancet, 1:1235– 1239. http://dx.doi.org/10.1016/S0140- 6736(86)91386-3 PMID:2872391
26. Mishra GA, Shastri SS, Uplap PA et al. (2009). Establishing a model workplace tobacco cessation program in India. Indian J Occup Environ Med, 13:97– 10 3 . h t t p: //d x . d o i . o r g /10 . 410 3 / 0 019 - 5278.55129 PMID:20386628
Websites American Cancer Society. Tobacco and Cancer: http://www.cancer.org/cancer/ cancercauses/tobaccocancer/index
Framework Convention Alliance: http://www.fctc.org/
The Tobacco Atlas: http://www.tobaccoatlas.org/
U.S. Centers for Disease Control and Prevention. Smoking & Tobacco Use: http://www.cdc.gov/tobacco/index.htm
WHO. MPOWER measures: http://www.who.int/tobacco/mpower/en/
WHO. Tobacco Free Initiative (TFI): http://www.who.int/tobacco/en/
96
Summary • The relationship between al-
cohol consumption and cancer risk has been known since the beginning of the 20th century. Epidemiological and biological research on the association has established that alcohol consumption causes cancers of the mouth, pharynx, larynx, oesophagus, liver, colorectum, and female breast.
• Typically, for cancer types caused by alcohol consumption, a dose–response association has been established.
• For 2010, alcohol-attributable cancers were estimated to be responsible for 337 400 deaths w o r l d w i d e , p r e d o m i n a n t l y among men, with liver cancer accounting for the largest pro- portion of deaths among the different tumour types.
• Alcoholic beverages are com- plex mixtures, but ethanol, me- diating a genotoxic effect upon metabolism to acetaldehyde, is recognized as the agent predominantly accounting for carcinogenesis.
• The burden of alcohol-attrib- utable cancer can be reduced through alcohol policy measures
such as reduction of availability, increases in price, and marketing bans.
The association between alcohol consumption and risk of cancer was known as early as the beginning of the 20th century, when Lamy ob- served that approximately 8 out of 10 patients with either cancer of the oesophagus or cancer of the cardiac region of the stomach were alcohol misusers [1]. This observation was followed by an ecological study in- dicating that people who were more likely to consume alcohol (such as people involved in the production and distribution of alcoholic bev- erages) had a higher risk of head and neck cancers compared with people who abstained from drink- ing for religious reasons, and that such abstainers had a markedly lower risk of these forms of cancers compared with the population as a whole [2]. After these early ob- servations, several thousand more analytical studies followed to ex- plore the biology and epidemiology of the relationship between alcohol consumption and risk of cancer. As a result, alcoholic beverages were declared “carcinogenic to humans” (Group 1) by the IARC Monographs Programme, first in 1988 [2] and then again in 2007 and in 2010
[3–5]. Tumour types caused by drinking alcoholic beverages include cancers of the oral cavity, pharynx, larynx, oesophagus, liver, colorec- tum, and female breast. For renal cell carcinoma and non-Hodgkin lymphoma, there is “evidence sug- gesting lack of carcinogenicity” for alcohol consumption [4,5].
Most cultures throughout the world have traditionally consumed some form of alcoholic beverages, and local specialty alcoholic bever- ages still account for the majority. Only a small number have evolved into commodities that are produced commercially on a large scale, such as beer from barley, wine from grapes, and certain distilled bever- ages. Further known beverages are other fruit wines, cider, and a broad range of very diverse spirits, includ- ing shochu, sake, lotus- or agave- based spirits, and various types of country-made liquor in India. In many developing countries, vari- ous types of home-made or locally produced alcoholic beverages, such as sorghum beer, palm wine, or sugar-cane spirits, continue to be the main available beverage types. Alcopops – flavoured, often sweet, alcoholic beverages – have received special attention mainly in a European context.
A large variety of substances that are not intended for human
2.3 Alcohol consumption Jürgen Rehm Kevin Shield
Naomi E. Allen (reviewer) Min Dai (reviewer) Elisabete Weiderpass (reviewer)
2 ETIOLOGY
Chapter 2.3 • Alcohol consumption 97
2 E
T IO
LO G
Y C
H A
P T
E R
2 .3
consumption are nevertheless be- ing consumed as alcohol (surrogate alcohol such as hairspray, after- shaves, lighter fluid, medicines, or al- cohol-containing mouthwash). They often contain very high concentra- tions of ethanol and may also con- tain higher alcohols and toxic con- centrations of methanol and other chemicals [5].
Epidemiological evidence Since the association between alco- hol and various forms of cancer is relatively modest at lower levels of consumption, there have been mul- tiple individual observational studies, where the majority of participants consume low to moderate amounts of alcohol, that have found a significant positive association, an absence of a significant association, or a signifi- cant negative association between alcohol consumption and the risk of mortality and morbidity from certain cancers. To determine whether an association exists between alcohol
consumption and various forms of cancer, the results of numerous case–control and cohort studies have been combined using meta- analytical techniques. These meta- analyses establish that a significant positive dose–response association exists between alcohol consumption and cancers of the mouth, pharynx, oesophagus, colorectum, liver, lar- ynx, pancreas, female breast, and prostate, while a significant negative association or no association have been observed for each of Hodgkin lymphoma [4,6], non-Hodgkin lym- phoma [4,7,8], and renal cell carcino- ma [4,9–11]. When epidemiological criteria were examined for causal- ity [12], the association between al- cohol consumption and cancers of the mouth, pharynx, oesophagus, colorectum, liver, larynx, and fe- male breast was found to be causal [4,5,13]. Table 2.3.1 provides a list of cancers that have been found to be associated with alcohol consumption through at least one meta-analysis
and shows the level of evidence sup- porting these associations.
The relationship between aver- age daily alcohol consumption and the risk of mortality from cancers of the upper digestive tract (except can- cers of the mouth and oral cavity) and from cancer of the female breast is ex- ponential. The relationship between alcohol consumption and the risk of mortality from cancers of the lower digestive tract and from cancers of the mouth and oral cavity is linear. For cancers of the upper digestive tract, lower digestive tract, mouth, oral cavity, and female breast, former drinkers (people who have not con- sumed alcohol within the past year but who have consumed it before in their lifetime) were found to have a higher risk of cancer compared with lifetime abstainers. It should be noted that there is evidence that the risk of head and neck cancers decreases with time since cessation of drinking [14]; however, more data are need- ed to explore the effect of drinking
Table 2.3.1. Cancers where alcohol consumption may be a component cause
Disease ICD-10 code Effecta Epidemiological evidencea
Malignant neoplasms
Cancers of the upper aerodigestive tract
Cancer of the oral cavity and pharynx C00–C13 Detrimental Causally related
Cancer of the larynx C32 Detrimental Causally related
Cancer of the oesophagus C15 Detrimental Causally related
Cancer of the colorectum C18–C21 Detrimental Causally related
Cancer of the liver and hepatobiliary tract C22 Detrimental Causally related
Cancer of the stomach C16 Detrimental Insufficient causal evidence
Cancer of the pancreas C25 Detrimental Causality may need to be re-evaluated
Cancer of the lung C33–C34 Detrimental Insufficient causal evidence
Cancer of the female breast C50 Detrimental Causally related
Cancer of the prostate C61 Detrimental Insufficient causal evidence
Cancer of the kidney and cancer of the urinary bladder
C64–C66, C68 (except C68.9)
Beneficialb/no associationa (renal cell carcinoma only)
Insufficient causal evidence
Cancers of the lymphatic and haematopoietic system
Hodgkin lymphoma C81 Beneficialc/no associationa Insufficient causal evidence
Non-Hodgkin lymphoma C82–C85, C96 Beneficiald/no associationa Insufficient causal evidence
ICD-10, International Classification of Diseases, 10th Revision. a For more information, see [4]. b For more information, see [4,9–11]. c For more information, see [4,6]. d For more information, see [4,7,8].
98
cessation on risk over time for other alcohol-related cancer sites. Box 2.3.1 provides plots of the rela- tive risk functions for all cancers that are currently determined to be causally associated with alcohol consumption [13]. The relative risk
0 50 100 150
Alcohol consumption (grams/day)
R el
at iv
e R
is k
0 .0
2 .0
4 .0
6 .0
8 .0
10 .0
0 50 100 150
Alcohol consumption (grams/day)
R el
at iv
e R
is k
0 .0
2 .0
4 .0
6 .0
8 .0
10 .0
0 50 100 150
Alcohol consumption (grams/day)
R el
at iv
e R
is k
0. 00
0. 50
1. 00
1. 50
2. 00
2. 50
0 50 100 150
Alcohol consumption (grams/day)
R el
at iv
e R
is k
0. 00
0. 50
1. 00
1. 50
2. 00
2. 50
0 50 100 150
Alcohol consumption (grams/day)
R el
at iv
e R
is k
0. 00
0. 50
1. 00
1. 50
2. 00
2. 50
Cancers of the mouth and oropharynx
Oesophageal cancer
Colon cancer Rectal cancer Liver cancer
Laryngeal cancer
Cancer of the female breast
Neoplasms of the upper digestive tract
Neoplasms of the lower digestive tract
Other neoplasms
0 50 100 150
Alcohol Consumption (grams/day)
R el
at iv
e R
is k
0 .0
2 .0
4 .0
6 .0
8 .0
10 .0
0 50 100 150
Alcohol consumption (grams/day)
R el
at iv
e R
is k
0. 00
1. 00
2. 00
3. 00
4. 00
Box 2.3.1. The relationship between average daily alcohol consumption and relative risk of cancer.
functions presented in Box 2.3.1 are not separated by cancer mortality and incidence. However, the rela- tive risk functions for alcohol-related cancers may be different for mortal- ity and incidence (as a difference between relative risks of mortality and incidence has been observed for other alcohol-related diseases, such as liver cirrhosis [15]). Thus, more research is needed to system- atically determine whether there is a difference between the relative risks of incidence and mortality for alco- hol-related cancers.
Evidence suggests a synergistic effect of tobacco smoking and con- sumption of alcoholic beverages on the risk of cancer of the oral cavity, pharynx, larynx, and oesophagus, with very high risks observed in
individuals who are both heavy drinkers and heavy smokers [4].
Observational research has suggested that there is a significant positive dose–response association between alcohol consumption and risk of prostate cancer, with multiple meta-analyses confirming this rela- tionship. Biological pathways indicat- ing how alcohol consumption may in- crease the risk of prostate cancer are currently unknown (see below), and thus additional research is needed to clarify a possible causal association.
Conflicting epidemiological evi- dence exists for the association be- tween alcohol consumption and cancers of the bladder, lung, and stomach. A nonsignificant positive association has been observed between alcohol consumption and
Chapter 2.3 • Alcohol consumption 99
2 E
T IO
LO G
Y C
H A
P T
E R
2 .3
cancers of the endometrium and ovary [13].
One pooled analysis and one meta-analysis found that in cohort and case–control studies, heavy al- cohol consumption – drinking more than 30 g of pure alcohol per day – has been consistently related to an increase in the risk of pancreatic can- cer (cited in [4]); however, as men- tioned in [4], confounding of smoking could not be excluded as a possible explanation of these results.
Meta-analyses have found no association or a significant negative association between alcohol con- sumption and the risk of renal cell carcinoma, Hodgkin lymphoma, and non-Hodgkin lymphoma [5,6]. These apparently protective observed ef- fects should be interpreted with cau- tion since the biological mechanisms are not understood and confounding and/or misclassification of abstainers may be responsible for the observa- tions that have been made.
Associations have been reported between alcohol consumption and
cancer of the cervix, endometrium, ovary, vulva and vagina, testis, brain, thyroid, and skin (malignant mela- noma, basal cell carcinoma, and squamous cell carcinoma) as well as leukaemia and multiple myeloma [5]; however, few studies have examined these associations. Epidemiological research is required to support pre- vious study results, and biological studies are needed to determine the mechanisms by which alcohol intake affects the risk of developing these cancers.
Global rates of alcohol- attributable cancers The number of deaths and number of disability-adjusted life years (DALYs) lost (a measure that combines years of life lost due to pre- mature mortality and years of healthy life lost due to disability) from can- cers that are currently determined to be causally associated with alcohol consumption and from other disease conditions and injuries, in each case
for 2010, can be calculated using mortality and morbidity data (see the WHO Global Information System on Alcohol and Health for global alcohol exposure estimates since 1960), with reference to the alcohol-attributable fraction methodology of the 2010 Global Burden of Disease study (see [16]).
In 2010, 337 400 deaths (91 500 of women and 245 900 of men) and 8 670 000 DALYs lost (2 252 000 for women and 6 418 000 for men) were caused by malignant neoplasms at- tributable to alcohol consumption. This burden represents 0.6% of all deaths (0.4% of all deaths of women and 0.8% of all deaths of men) and 0.3% of all DALYs lost (0.2% of all DALYs lost for women and 0.5% of all DALYs lost for men) in 2010. It should be noted that alcohol-attrib- utable cancer data for 2010 reflect the level of drinking in the early 1990s, due to the long time it takes for cancer to develop [17].
The rates of alcohol-attribut- able cancer deaths and of alcohol-
Mouth cancer Cancer of the nasopharynx
Cancer of other part of pharynx and oropharynx
Oesophageal cancer
Cancers of the colon and
rectum Liver cancer
Laryngeal cancer
Cancer of the female breast
All cancer deaths
Total 27,8% 29,3% 26,7% 19,4% 5,5% 10,7% 21,5% 8,8% 4,2%
Women 11,6% 12,6% 9,3% 8,3% 5,2% 6,4% 7,9% 8,8% 2,7%
Men 37,0% 36,3% 32,5% 23,8% 5,8% 12,7% 23,5% 5,4%
0,0%
5,0%
10,0%
15,0%
20,0%
25,0%
30,0%
35,0%
40,0%
Pe rc en
ta ge
o f d
ea th s at tr ib ut ab
le t o
al co ho
l c on
su m pt io n
Fig. 2.3.1. Percentage of deaths from various forms of cancer attributable to alcohol consumption, in 2010.
100
attributable DALYs lost varied glob- ally with reference to regions de- fined by WHO as Global Burden of Disease regions (see [23]). The number of deaths and the number of DALYs lost per 100 000 people are shown in Tables 2.3.2 and 2.3.3, respectively, by Global Burden of Disease region for 2010. The Eastern Europe region experienced the highest alcohol-attributable can- cer burden, with 8.7 cancer deaths per 100 000 people (5.7 per 100 000 women and 12.9 per 100 000 men) and 242.5 DALYs lost per 100 000 people (153.9 per 100 000 women and 357.4 per 100 000 men), while the North Africa/Middle East region experienced the lowest alcohol- attributable cancer burden, with 0.6 cancer deaths per 100 000 peo- ple (0.5 per 100 000 women and 0.8 per 100 000 men) and 18.5 DALYs lost per 100 000 people (14.8 per 100 000 women and 22.0 per 100 000 men).
In 2010, the largest contributors to the burden of alcohol-attributable cancer deaths were: overall, liver
cancer (responsible for 23.9% of all such deaths); for women, breast cancer (responsible for 42.0% of these deaths); and for men, oesoph- ageal cancer (responsible for 27.4% of these deaths). In 2010, the largest contributors to DALYs lost caused by alcohol-attributable cancer were: overall, liver cancer (responsible for 24.7% of all such DALYs lost); for women, breast cancer (responsible for 48.1% of these DALYs lost); and for men, liver cancer (responsible for 28.2% of these DALYs lost).
The percentage of cancer deaths and the percentage of cancer DALYs lost attributable to alcohol consump- tion are shown in Figs 2.3.1 and 2.3.2, respectively, by cause. In 2010, 4.2% of all deaths caused by cancer were attributable to alcohol consumption (2.7% for women and 5.4% for men) and 4.6% of all DALYs lost caused by cancer were attribut- able to alcohol consumption (2.8% for women and 6.0% for men). In 2010, alcohol consumption was re- sponsible for the largest percentage of the total deaths and DALYs lost
from nasopharyngeal cancer com- pared with other cancer types; was responsible for 29.3% of all deaths from cancers of the mouth and oropharynx (12.6% for women and 36.3% for men); and was responsi- ble for 29.3% of all DALYs lost from cancers of the mouth and orophar- ynx (12.7% for women and 36.1% for men).
Contribution of cancer to the total alcohol-attributable burden of disease Alcohol consumption is related to more than 200 three-digit ICD-10 (International Classification of Diseases, 10th Revision) code dis- eases, conditions, and injuries, such as infectious diseases, malignant neoplasms, diabetes, neuropsy- chiatric conditions, cardiovascular disease, digestive diseases, condi- tions arising during the prenatal pe- riod, and injuries (for an overview of the conditions related to alcohol consumption, see [16]). In 2010, al- cohol consumption was responsible for 2 734 200 deaths (910 800 of
Fig. 2.3.2. Percentage of disability-adjusted life years (DALYs) lost from various forms of cancer attributable to alcohol consumption, in 2010.
Mouth cancer Cancer of the nasopharynx
Cancer of other part of pharynx and oropharynx
Oesophageal cancer
Cancers of the colon and rectum
Liver cancer Laryngeal cancer
Cancer of the female breast
All cancer DALYs lost
Total 28,9% 29,3% 27,4% 20,4% 5,6% 11,2% 22,5% 9,0% 4,6%
Women 11,8% 12,7% 8,9% 8,5% 5,3% 6,4% 8,2% 9,0% 2,8%
Men 37,5% 36,1% 33,9% 24,8% 6,0% 13,0% 24,6% 6,0%
0,0%
5,0%
10,0%
15,0%
20,0%
25,0%
30,0%
35,0%
40,0% Pe
rc en
ta ge
o f D
A LY s lo st a tt ri bu
ta bl e to
al co ho
l c on
su m pt io n
Chapter 2.3 • Alcohol consumption 101
2 E
T IO
LO G
Y C
H A
P T
E R
2 .3
women and 1 823 400 of men) and 97 118 000 DALYs lost (22 523 000 for women and 74 595 000 for men). Of the total deaths and DALYs lost attributable to alcohol consumption in 2010, alcohol-attributable cancers were responsible for 12.3% of all alcohol-attributable deaths (10.0% for women and 13.5% for men) and 8.9% of all alcohol-attributable DALYs lost (10.0% for women and 8.6% for men).
Possible biological mechanisms As indicated above, alcohol con- sumption is carcinogenic and caus- es a variety of human cancers, and
there may be different biological pathways depending on the anatom- ical site. Alcoholic beverages are multicomponent mixtures containing several carcinogenic compounds, such as ethanol, acetaldehyde, afla- toxins, and ethyl carbamate (see [18]), and all of these compounds may contribute to increase the risk of cancer due to alcohol consump- tion reported in observational stud- ies. A recent chemical risk analysis concluded that ethanol is the most important carcinogen in alcoholic beverages, with all other carcino- genic compounds contained in al- coholic beverages increasing the risk of cancer below the threshold
normally acceptable for food con- taminants [18].
The biological mechanisms by which alcohol intake increases the risk of cancer are not fully understood, but the main mechanisms are likely to include a genotoxic effect of acetal- dehyde, the induction of cytochrome P450 2E1 and associated oxidative stress, increased estrogen concen- tration, a role as a solvent for tobacco carcinogens, changes in folate me- tabolism, and changes in DNA repair [4,5].
For cancers of the digestive tract, especially those of the upper diges- tive tract, acetaldehyde from alcohol metabolism in the body and from
Table 2.3.2. Cancer deaths attributable to alcohol consumption, by sex and by Global Burden of Disease region, for 2010a
Global Burden of Disease regionb
Sex
Female Male Total
Deaths Deaths per 100 000 people Deaths Deaths per
100 000 people Deaths Deaths per
100 000 people
Asia Pacific, high-income 4 000 2.4 13 400 8.1 17 400 5.2
Asia, Central 1 200 3.3 2 000 6.2 3 300 4.6
Asia, East 20 300 2.7 99 800 11.8 120 100 7.5
Asia, South 6 300 1.1 27 200 4.2 33 400 2.7
Asia, South-East 4 700 1.9 12 200 4.7 16 900 3.3
Australasia 800 4.1 1 200 5.6 2 000 4.9
Caribbean 400 2.0 1 000 4.4 1 400 3.2
Europe, Central 3 600 3.7 9 900 11.2 13 500 7.2
Europe, Eastern 10 400 5.7 17 300 12.9 27 700 8.7
Europe, Western 18 300 4.8 29 700 8.2 48 000 6.5
Latin America, Andean 400 2.0 300 1.4 700 1.7
Latin America, Central 1 800 1.8 2 800 3.0 4 600 2.4
Latin America, Southern 1 700 4.4 2 200 6.3 3 900 5.3
Latin America, Tropical 2 800 2.8 5 900 6.2 8 600 4.4
North Africa/Middle East 800 0.5 1 300 0.8 2 100 0.6
North America, high-income 8 500 3.4 11 300 4.4 19 800 3.9
Oceania 100 4.6 200 5.6 300 5.1
Sub-Saharan Africa, Central 400 1.8 700 2.9 1 100 2.3
Sub-Saharan Africa, Eastern 2 500 2.8 3 300 3.7 5 900 3.3
Sub-Saharan Africa, Southern 600 2.1 1 600 6.8 2 200 4.3
Sub-Saharan Africa, Western 1 700 2.0 2 700 2.9 4 400 2.4
World 91 500 2.7 245 900 7.1 337 400 4.9
a Numbers have been rounded. b For definitions of regions, see [23].
102
Table 2.3.3. Cancer disability-adjusted life years (DALYs) lost attributable to alcohol consumption, by sex and by Global Burden of Disease region, for 2010a
Global Burden of Disease regionb
Sex
Female Male Total
DALYs lost
DALYs lost per 100 000 people
DALYs lost
DALYs lost per 100 000 people
DALYs lost
DALYs lost per 100 000 people
Asia Pacific, high-income 81 000 60.8 284 000 196.5 365 000 128.5
Asia, Central 38 000 98.5 58 000 170.9 95 000 132.1
Asia, East 530 000 69.3 2 690 000 318.9 3 221 000 200.0
Asia, South 182 000 28.8 782 000 115.3 964 000 73.9
Asia, South-East 113 000 41.8 342 000 124.7 455 000 82.7
Australasia 17 000 101.3 24 000 129.6 42 000 115.9
Caribbean 12 000 53.6 25 000 113.2 36 000 83.5
Europe, Central 90 000 103.8 253 000 302.9 342 000 199.6
Europe, Eastern 250 000 153.9 461 000 357.4 711 000 242.5
Europe, Western 374 000 121.5 649 000 206.5 1 023 000 163.9
Latin America, Andean 11 000 48.9 9 000 38.0 20 000 43.4
Latin America, Central 47 000 45.9 66 000 67.1 112 000 56.6
Latin America, Southern 38 000 106.8 49 000 144.4 86 000 124.4
Latin America, Tropical 76 000 75.4 167 000 172.7 243 000 122.7
North Africa/Middle East 28 000 14.8 42 000 22.0 70 000 18.5
North America, high-income 205 000 90.0 263 000 112.0 468 000 100.8
Oceania 4 000 124.0 5 000 153.4 9 000 138.7
Sub-Saharan Africa, Central 13 000 47.6 21 000 81.9 34 000 64.4
Sub-Saharan Africa, Eastern 72 000 71.3 98 000 97.9 170 000 84.3
Sub-Saharan Africa, Southern 17 000 56.6 47 000 182.8 63 000 114.9
Sub-Saharan Africa, Western 54 000 53.0 85 000 80.7 139 000 66.9
World 2 252 000 66.0 6 418 000 184.8 8 670 000 125.9
a Numbers have been rounded. b For definitions of regions, see [23].
ingestion as a component of alco- holic beverages [19] has been high- lighted as a likely and important causal pathway [4,5]. For colorectal cancer, in addition to the genotoxic effect of acetaldehyde, there may be the involvement of folate: alcohol may act through folate metabolism or synergistically with low folate intake (see [5]). The biological mechanisms accounting for how alcohol intake may cause pancreatic cancer are currently unclear; however, potential mechanisms include pancreatitis and the accumulation of fatty acid esters in the pancreas, which may induce
inflammatory responses and fibrosis [4].
The biological effects of alco- hol intake on the risk of digestive tract cancers are also dependent on the genotype of the consumer; individuals with the ALDH2 Lys487 allele (and therefore a deficiency of ALDH2) experience a higher risk of oesophageal cancer for the same amount of alcohol consumed. The ALDH2 Lys487 allele is thought to modify the risk of all cancers that are caused by the metabolites of alcohol [4,5].
In relation to breast cancer, al- cohol intake has been shown to
increase levels of estrogen and plasma insulin-like growth factor produced by the liver, to enhance mammary gland susceptibility to car- cinogenesis (by altering mammary gland structural development and by stimulating cell proliferation), to increase mammary DNA damage, and to facilitate the ability of breast cancer cells to migrate [20].
Implications for prevention The relationship between alco- hol consumption and cancer has been observed as monotonic and without threshold (see above and
Chapter 2.3 • Alcohol consumption 103
2 E
T IO
LO G
Y C
H A
P T
E R
2 .3
[3,4]). Thus, as the amount of alco- hol consumed increases, the risk of developing cancer increases. This means that any reduction in alco- hol consumption will be beneficial for health through the reduction of cancer risk. There are cost-effective ways to reduce alcohol consump- tion: most notably, restrictions in availability; increases in price for alcoholic beverages, which could be achieved by increasing taxation or increasing minimum prices; and marketing bans [21]. In addition, the risk of cancer for high-risk heavy alcohol consumers can be reduced by providing more opportunities for brief interventions and treatment for alcohol use disorders [22].
Fig. 2.3.3. Female drinking has risen steadily for the past 20 years, with more women than men drinking heavily on single occasions.
104
References
1. Lamy L (1910). Clinical and statistical study of 134 cases of cancer of the oe- sophagus and of the cardia [in French]. Arch Mal Appar Dig Mal Nutr, 4:451– 475.
2. IARC (1988). Alcohol drinking. IARC Monogr Eval Carcinog Risks Hum, 44:1– 378. PMID:3236394
3. Baan R, Straif K, Grosse Y et al.; WHO IARC Monograph Working Group (2007). Carcinogenicity of alcoholic beverages. Lancet Oncol, 8:292–293. http://dx.doi. o r g /10 .1016 / S1470 -2 0 4 5 (07 )70 0 9 9 -2 PMID:17431955
4. IARC (2012). Personal habits and in- door combustions. IARC Monogr Eval Carcinog Risks Hum, 100E:1–575. PMID: 23193840
5. IARC (2010). Alcohol consumption and ethyl carbamate. IARC Monogr Eval Carcinog Risks Hum, 96:1–1428. PMID:21735939
6. Tramacere I, Pelucchi C, Bonifazi M et al. (2012). A meta-analysis on alcohol drink- ing and the risk of Hodgkin lymphoma. Eur J Cancer Prev, 21:268–273. http://dx.doi. o r g /10 .10 97/ C E J . 0 b 013 e 3 2 8 3 5 0 b11b PMID:22465910
7. Tramacere I, Pelucchi C, Bonifazi M et al. (2012). Alcohol drinking and non-Hodgkin lymphoma risk: a systematic review and a meta-analysis. Ann Oncol, 23:2791– 2798. PMID:22357444
8. Morton LM, Zheng T, Holford TR et al.; InterLymph Consortium (2005). Alcohol consumption and risk of non-Hodgkin lym- phoma: a pooled analysis. Lancet Oncol, 6:469– 476. http://dx.doi.org/10.1016/ S1470-2045(05)70214-X PMID:15992695
9. Bellocco R, Pasquali E, Rota M et al. (2012). Alcohol drinking and risk of re- nal cell carcinoma: results of a meta- analysis. Ann Oncol, 23:2235–2244. http://dx.doi.org/10.1093/annonc/mds022 PMID:22398178
10. Song DY, Song S, Song Y, Lee JE (2012). Alcohol intake and renal cell cancer risk: a meta-analysis. Br J Cancer, 106:1881– 1890. http://dx.doi.org/10.1038/bjc.2012. 136 PMID:22516951
11. Cheng G, Xie L (2011). Alcohol intake and risk of renal cell carcinoma: a meta- analysis of published case-control stud- ies. Arch Med Sci, 7:648–657. http:// d x .d o i .o r g /10 . 5114 /a o m s . 2 011. 2413 5 PMID:22291801
12. Rothman KJ, Greenland S, Lash TL, eds (2008). Modern Epidemiology, 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins.
13. Shield KD, Parry C, Rehm J (2013). Chronic diseases and conditions related to alcohol use. Alcohol Res, (in press).
14. Lubin JH, Purdue M, Kelsey K et al. (2009). Total exposure and exposure rate effects for alcohol and smoking and risk of head and neck cancer: a pooled analysis of case-control studies. Am J Epidemiol, 170:937–947. http://dx.doi.org/10.1093/aje/ kwp222 PMID:19745021
15. Rehm J, Taylor B, Mohapatra S et al. (2010). Alcohol as a risk factor for liver cirrhosis: a systematic review and meta-analysis. Drug Alcohol Rev, 29:437–445. http://dx.doi. org/10.1111/j.1465 -3362. 20 09.0 0153.x PMID:20636661
16. Rehm J, Baliunas D, Borges GLG et al. (2010). The relation between different di- mensions of alcohol consumption and burden of disease: an overview. Addiction, 105:817–843. http://dx.doi.org/10.1111/ j.1360-0443.2010.02899.x PMID:20331573
17. Holmes J, Meier PS, Booth A et al. (2012). The temporal relationship between per capita alcohol consumption and harm: a systematic review of time lag specifications in aggregate time series analyses. Drug Alcohol Depend, 123:7–14. http://dx.doi. org /10.1016/ j.dr ugalc dep. 2011.12.0 0 5 PMID:22197480
18. Lachenmeier DW, Przybylski MC, Rehm J (2012). Comparative risk assessment of carcinogens in alcoholic beverages using the margin of exposure approach. Int J Cancer, 131:E995–E1003. http://dx.doi. org/10.1002/ijc.27553 PMID:22447328
19. Lachenmeier DW, Kanteres F, Rehm J (2009). Carcinogenicity of acetalde- hyde in alcoholic beverages: risk as- sessment outside ethanol metabolism. Addiction, 104:533–550. http://dx.doi. org/10.1111/j.136 0 - 0 4 43. 20 09.02516.x PMID:19335652
20. Singletary KW, Gapstur SM (2001). Alcohol and breast cancer: review of epidemiologic and experimental evidence and poten- tial mechanisms. JAMA, 286:2143–2151. http://dx.doi.org/10.1001/jama.286.17.2143 PMID:11694156
21. Babor T, Caetano R, Casswell S et al. (2010). Alcohol: No Ordinary Commodity: Research and Public Policy, 2nd ed. Oxford: Oxford University Press. http:dx.doi.org/10.1093/ acprof:oso/9780199551149.001.0001
22. Rehm J, Shield KD, Gmel G et al. (2013). Modeling the impact of alcohol dependence on mortality burden and the effect of available treatment interventions in the European Union. Eur Neuropsychopharmacol, 2 3:8 9 – 97. ht t p: //d x .d o i.o r g /10.1016 / j. euroneuro.2012.08.001 PMID:22920734
23. Murray CJL, Ezzati M, Flaxman AD et al. (2012). GBD 2010: design, definitions, and metrics. Lancet, 380: 2063–2066. http:// dx.doi.org/10.1016/S0140-6736(12)61899-6 PMID:23245602
Website WHO Global Information System on Alcohol and Health: http://apps.who.int/ ghodata/?theme=GISAH
Chapter 2.4 • Infections 105
2 E
T IO
LO G
Y C
H A
P T
E R
2 .4
Summary • Infections with viruses, bacteria,
and macroparasites have been identified as strong risk factors for specific cancers.
• Overall, about 2 million (16%) of the total of 12.7 million new cancer cases in 2008 are attrib- utable to infections. This fraction varies 10-fold by region; it is low- est in North America, Australia, and New Zealand (≤ 4%) and highest in sub-Saharan Africa (33%).
• Helicobacter pylori, hepatitis B and C viruses, and human papillomaviruses are responsi- ble for 1.9 million cancer cases globally, including mainly gas- tric, liver, and cervical cancer, respectively.
• Infection with HIV substantially increases the risk of virus- associated cancers, through immunosuppression.
• Application of existing methods for infection prevention, such as vaccination, safe injection prac- tices, and safe sexual behaviour, or antimicrobial and antiparasite treatments could have a major impact on the future burden of cancer worldwide.
2.4 2 ETIOLOGY
Infections Silvia Franceschi Rolando Herrero
Andrew J. Hall (reviewer) Robert Newton (reviewer) You-Lin Qiao (reviewer)
Few aspects of cancer etiology have undergone as much progress in the past three decades as the relationship between a selected number of chronic infections and the risk of cancer. When Doll and Peto [1] estimated in 1981 the proportion of cancer cases attributable to different causes, only the relation- ships of hepatitis B virus (HBV) with liver cancer and of Epstein–Barr vi- rus (EBV) with Burkitt lymphoma were relatively well understood, and the at- tribution to infections was very uncer- tain (> 1%, but the upper confidence interval limit was not reported). About 30 years later, the IARC Monographs Volume 100B Working Group [2] de- clared 11 infectious agents as well-es- tablished human carcinogenic agents (Table 2.4.1). In addition, de Martel et al. [3] considered infectious agents classified by IARC as carcinogenic to humans and calculated correspond- ing population attributable fractions (PAFs) worldwide and in eight geo- graphical regions, using statistics on estimated cancer incidence from 2008 [4] (Fig. 2.4.1). In this chapter, we use the term “developed regions” to in- clude Japan, North America, Europe, Australia, and New Zealand, i.e. high- income countries, and the term “de- veloping regions” to include all other countries, i.e. low- and middle-income countries. In this context, the PAF is an estimate of the proportion of cases of a cancer that could theoretically be
avoided if exposure to a specific infec- tion were avoided or if the infection were detected early and treated be- fore the development of cancer.
The associations of infections with cancer are nearly always very strong (relative risk > 10), and therefore calculations can often be based on the prevalence of infection in cancer cases (e.g. DNA of human papil- lomavirus [HPV] or EBV in cancer tissue biopsies) rather than on the less-often-available prevalence in the general population [3]. Estimates of prevalence of infection and relative risks are extracted from published data, and a short description of the rationale for each cancer site is pre- sented here for each infectious agent in relation to its associated cancer. HIV infection greatly increases the risk of some infection-associated cancer types, especially Kaposi sarcoma and lymphomas, through immunosuppres- sion [2]. To avoid double attribution, the fraction of cancer due to HIV will not be reported separately, but HIV will be mentioned when the infection is strongly implicated in the increase of certain cancer types.
Helicobacter pylori Non-cardia gastric adenocarcinoma A causal association between chron- ic gastric infection with Helicobacter
106
make-up of both the host and the bacterium can modulate the strength of the causal relation between the bacterium and the cancer, no risk factors for non-cardia gastric adeno- carcinoma as strong as H. pylori have been recognized.
A well-known difficulty in esti- mating the prevalence of H. pylori infection is the gradual decline in bacterial load with increasing gas- tric atrophy – a precursor lesion of gastric adenocarcinoma – and the consequent lack of sensitivity of serological markers at the time of cancer diagnosis. Studies in which H. pylori antibodies were measured well before the onset of cancer are considered the most reliable for cal- culating the true prevalence in cases and controls and therefore for esti- mating the relative risk. A review of such studies in 2001 reported that the average prevalence of H. pylori infec- tion in cases is 90%, a figure that was quite homogeneous across studies in various continents despite a highly variable background rate in popu- lation-based controls [5]. A pooled analysis of 11 of these studies yielded a common relative risk of 5.9 (95% confidence interval [CI], 3.4–10.3) for patients who developed cancer more than 10 years after blood draw [5]. This yields a PAF of 75% for H. pylori in non-cardia gastric adenocar- cinoma [3] (Table 2.4.2).
Table 2.4.1. Major cancer sites related to infectious agents classified as carcinogenic to humans
Cancer site Well-established human carcinogenic agents
Stomach Helicobacter pylori
Liver Hepatitis B virus Hepatitis C virus Opisthorchis viverrini Clonorchis sinensis
Cervix Human papillomavirus, with or without HIV
Anogenital (penis, vulva, vagina, anus) Human papillomavirus, with or without HIV
Nasopharynx Epstein–Barr virus
Oropharynx Human papillomavirus, with or without tobacco use/alcohol consumption
Non-Hodgkin lymphoma Helicobacter pylori Epstein–Barr virus, with or without HIV Hepatitis C virus Human T-cell lymphotropic virus type 1
Kaposi sarcoma Kaposi sarcoma herpesvirus, with or without HIV
Hodgkin lymphoma Epstein–Barr virus, with or without HIV
Bladder Schistosoma haematobium
Fig. 2.4.1. Number of new cancer cases in 2008 attributable to infection, by infectious agent and development status.
pylori and development of gastric adenocarcinoma is well established [2]. The risk is restricted to the non- cardia part (i.e. body) of the stom- ach; adenocarcinoma of the gastric
cardia shares risk factors with oe- sophageal cancer and is not asso- ciated with H. pylori [2,5]. Although environmental factors such as the di- etary habits of the host or the genetic
Chapter 2.4 • Infections 107
2 E
T IO
LO G
Y C
H A
P T
E R
2 .4
Inflammation and cancer Curtis C. Harris
Inflammation has long been associ- ated with cancer, and it is one of the hallmarks of cancer [1]. Whereas acute inflammation protects against infectious pathogens, chronic in- flammation is associated with DNA and tissue damage, including ge- netic and epigenetic changes lead- ing to cancer [2]. This association can be both inherited and acquired (Table B2.4.1). Both single-gene inheritance (e.g. haemochromato- sis) and complex multiple-gene in- heritance (e.g. inflammatory bowel
disease) have been defined by genetic analysis. Infections by vi- ruses, bacteria, and parasites are well-known examples of acquired etiology. Tobacco smoke contains more than 60 chemical carcinogens and also induces an inflammatory response that may facilitate lung carcinogenesis [3]. The inflamma- tory response to gastric acid reflux in Barrett disease may be respon- sible for the increased incidence of oesophageal cancer. Obesity has been recognized as a chronic
inflammatory condition predispos- ing to cardiovascular disease and cancer [4]. The cellular and mo- lecular basis of this predisposition involves the accumulation of mac- rophages in the adipose tissue, with the establishment of pro-in- flammatory feedback loops among macrophages, pre-adipocytes, and adipocytes to generate inflamma- tory cytokines and free radicals [5]. The prevention of colon polyps and cancer by aspirin and other anti-in- flammatory agents strengthens the
Table B2.4.1. Chronic inflammation or infection increases cancer risk
Disease Type of cancer
Auto-inflammatory/non-infectious
Crohn disease Colon cancer
Ulcerative colitis Colon cancer
Chronic pancreatitis Pancreatic cancer
Endometriosis Endometrial cancer
Haemochromatosis Liver cancer
Thyroiditis Thyroid cancer
α-1-Anti-trypsin deficiency Liver cancer
Acquired
Viral
Hepatitis B virus Liver cancer
Hepatitis C virus Liver cancer
Epstein–Barr virus Hodgkin lymphoma and Burkitt lymphoma
Bacterial
Helicobacter pylori Gastric cancer
Pelvic inflammatory disease Ovarian cancer
Chronic prostatitis Prostate cancer
Parasitic
Schistosoma haematobium Bladder cancer
Schistosoma japonicum Colon cancer
Liver fluke Cholangiocarcinoma and liver cancer
Chemical/physical/metabolic
Alcohol Multiple cancers (including cancers of the liver, pancreas, and head and neck)
Asbestos Mesothelioma
Obesity Multiple cancers
Tobacco smoke and inhalation of other noxious chemicals Lung cancer (and multiple other cancers)
Gastric reflux, Barrett oesophagus Oesophageal cancer
108
causal link between inflammation and cancer in humans [6].
The Human Microbiome Project is defining the microbes on the ex- ternal and internal epithelial surfac- es of the host in health and disease [7]. The gut microbiota is clearly implicated in intestinal inflammation and possibly human colon cancer [8]. Loss of the epithelial barrier integrity, due to mutations or other factors, allows translocation of the microbiota into the mucosa to elicit a cellular inflammatory response and release of cytokines. For example, microbes can elicit a TH17 response promoting IL-17 expression [9] and increase STAT3, which transcribes both coding and non-coding genes to signal proliferation of epithelial cells and their acquisition of epi- genetic and genetic changes [10]. Current research focuses on the mechanistic connections among obesity, the microbiome, chronic in- flammation, and cancer risk.
The molecular mechanisms of chronic inflammation driving can- cer are being discovered [11]. The tumour microenvironment, includ- ing macrophages and senescent
stromal cells, can facilitate tumour progression and metastasis [12]. Significant advances include iden- tifying pro- and anti-inflammatory cytokines and their cellular signal transduction pathways, such as the NF-κB pathway [11]. Discoveries in the chemistry of free radicals gener- ated by inflammation provide insight into their contribution to carcino- genesis. Mutations to DNA by both direct and indirect damage lead to activation of oncogenes and inacti- vation of tumour suppressor genes [13,14]. Animal models are helping to identify the oncogenic function of specific cytokines and free radicals [11,12]. Future research on inflam- mation will continue to improve pre- vention and treatment of cancer.
References 1. Hanahan D, Weinberg RA (2011). Cell,
144:646–674. http://dx.doi.org/10.1016/j. cell.2011.02.013 PMID:21376230
2. Rook GA, Dalgleish A (2011). Immunol Rev, 240:141–159. http://dx.doi.org/10.1111/ j.1600-065X.2010.00987.x PMID:21349092
3. Hecht SS (2012). Int J Cancer, 131:2724– 2732. http://dx.doi.org/10.1002/ijc.27816 PMID:22945513
4. Chaturvedi AK et al. (2013). Am J Epidemiol, 177:14–19. http://dx.doi.org/10.1093/aje/ kws357 PMID:23171878
5. Weisberg SP et al. (2003). J Clin Invest, 112:1796–1808. http://dx.doi.org/10.1172/ JCI19246 PMID:14679176
6. Thun MJ et al. (2012). Nat Rev Clin Oncol, 9:259–267. http://dx.doi.org/10.1038/nrclin onc.2011.199 PMID:22473097
7. Hooper LV et al. (2012). Science, 336:1268–1273. http://dx.doi.org/10.1126/ science.1223490 PMID:22674334
8. Gallimore AM, Godkin A (2013). N Engl J Med, 368:282–284. http://dx.doi.org/10.1056/ NEJMcibr1212341 PMID:23323906
9. Honda K, Littman DR (2012). Annu Rev Immunol, 30:759–795. http://dx.doi.org/ 10.1146/annurev-immunol- 020711-074937 PMID:22224764
10. Tosolini M et al. (2011). Cancer Res, 71:1263– 1271. http://dx.doi.org/10.1158/0008-5472. CAN-10-2907 PMID:21303976
11. Ben-Neriah Y, Karin M (2011). Nat Immunol, 12:715–723. http://dx.doi.org/10.1038/ni. 2060 PMID:21772280
12. Qian BZ et al. (2011). Nature, 475:222– 225. http://dx.doi.org/10.1038/nature10138 PMID:21654748
13. Lonkar P, Dedon PC (2011). Int J Cancer, 128:1999–2009. http://dx.doi.org/10.1002/ ijc.25815 PMID:21387284
14. Hussain SP et al. (2003). Nat Rev Cancer, 3:276–285. http://dx.doi.org/10.1038/nrc1046 PMID:12671666
Gastric lymphoma Non-Hodgkin lymphoma (NHL) in the stomach represents approxi- mately 5% of all NHL and less than 2% of all gastric cancers. Nearly all
mucosa-associated lymphoid tis- sues and an unknown proportion of diffuse large B-cell lymphomas in the stomach are related to H. pylori infection [6]. Based on these limited data, the pooled prevalence of H. pylori in gastric lymphoma is 86% and the pooled relative risk esti- mate is 7.2, giving a PAF of 74% [3] (Table 2.4.2).
Hepatitis B and C viruses Hepatocellular carcinoma Chronic infection with HBV and/or HCV is the most common risk factor for hepatocellular carcinoma world- wide. Some non-viral exposures are also important risk factors for hepatocellular carcinoma, including heavy alcohol consumption and di- etary aflatoxin contamination. The contribution of viral and non-viral risk factors, alone or in association,
varies greatly in different geographi- cal regions. For hepatitis viruses, HBV infection is found in the general population substantially more often than HCV in most Asian and African countries, while the opposite is true in Europe and the USA but also in Japan, Pakistan, and Mongolia [7]. In HBV-endemic areas, HBV is typi- cally acquired in early life, and the prevalence of chronic HBV carriage tends to plateau after adolescence. In contrast, HCV infection can be ac- quired at any age through contami- nated needles and blood, and HCV prevalence increases steadily with age due to the accumulating risk of exposure.
Estimates of HBV and HCV prev- alence in hepatocellular carcinoma cases were derived from a 2007 re- view [7], which was updated using more recent papers [2]. However, there are few data on HBV and HCV
Fig. 2.4.2. Helicobacter pylori bacterium. H. pylori infection plays a key role in the genesis of gastric cancer.
Chapter 2.4 • Infections 109
2 E
T IO
LO G
Y C
H A
P T
E R
2 .4
Table 2.4.2. Number of new cancer cases in 2008 attributable to infectious agents, by anatomical site
Cancer site Number of new cases in 2008a PAF (%)
Number attributable to infection
By sex By development statusb
Male Female Developing regions Developed
regions
Carcinoma
Non-cardia gastric 870 000 74.7 410 000 240 000 470 000 180 000
Liverc 750 000 76.9 400 000 170 000 510 000 69 000
Cervix 530 000 100.0 0 530 000 450 000 77 000
Vulva 27 000 43.0 0 12 000 4 100 7 500
Anus 27 000 88.0 11 000 13 000 12 000 12 000
Penis 22 000 50.0 11 000 0 7 600 3 200
Vagina 13 000 70.0 0 9 000 5 700 3 400
Oropharynx 85 000 25.6 17 000 4 400 6 400 15 000
Nasopharynx 84 000 85.5 49 000 23 000 66 000 5 900
Bladder 260 000 2.3 4 600 1 400 6 000 0
Lymphoma/leukaemia
Hodgkin lymphoma 68 000 49.1 20 000 13 000 23 000 10 000
Non-Hodgkin gastric lymphoma 18 000 74.1 7 400 5 800 6 500 6 700
Burkitt lymphoma 11 000 62.5 4 000 2 800 6 300 530
HCV-associated non-Hodgkin lymphoma
360 000 8.2 17 000 13 000 18 000 11 000
Adult T-cell leukaemia/lymphoma 2 100 100.0 1 200 900 660 1 500
Sarcoma
Kaposi sarcoma 43 000 100.0 29 000 14 000 39 000 4 100
Total from infections 3 200 000 64.4 990 000 1 100 000 1 600 000 410 000
HCV, hepatitis C virus; PAF, population attributable fraction. a Numbers are rounded to two significant digits. b Developed regions include Japan, North America, Europe, Australia, and New Zealand, i.e. high-income countries. Developing regions include all other countries, i.e. low- and middle-income countries. c Including cholangiocarcinoma.
prevalence in large parts of eastern Europe and Central Asia. In HBV- endemic areas, almost all hepato- cellular carcinomas in individuals younger than 50 years are due to HBV. The PAF for HCV is especially challenging as the infection does not show geographical clustering; high- prevalence countries may thus be geographically close to low-preva- lence countries. For these reasons, regional prevalence estimates were derived by statistical modelling [3].
A meta-analysis of 32 case–con- trol studies published in 1992–1997 yielded a summary relative risk of 22.5 (95% CI, 19.5–26.0) for HBV alone and of 17.3 (95% CI, 13.9– 21.6) for HCV alone [8]. The dual impact of HBV and HCV infection
on risk is not well quantified. Recent prospective and retrospective stud- ies are consistent with previous stud- ies in that they yield relative risks for hepatocellular carcinoma of a similar magnitude [2]. The estimated prevalence of either HBV or HCV in hepatocellular carcinoma cases var- ied from 42% in North America to 87% in Japan. Worldwide, the PAF for hepatocellular carcinoma is 77% (Table 2.4.2).
Non-Hodgkin lymphoma HCV infection is a well-established cause of essential mixed cryoglob- ulinemia, a lymphoproliferative disease that can evolve into B-cell NHL. Several studies found a high prevalence of HCV seropositivity
in patients with B-cell lymphoprolif- erative disorders, particularly B-cell NHL, including cases where essen- tial mixed cryoglobulinemia was ab- sent. A meta-analysis of studies of NHL and HCV seropositivity showed a pooled relative risk for NHL of 2.5 (95% CI, 2.1–3.1) based on 15 case– control studies and of 2.0 (95% CI, 1.8–2.2) in three cohort investiga- tions [9]. The number of new cases of NHL attributable to HCV varies by country and therefore can be es- timated very approximately. Based on a summary relative risk of 2.5 and the prevalence of HCV in cases from different regions [9], the HCV- associated PAF for NHL would be 8% (Table 2.4.2).
110
Human papillomaviruses
Cervical cancer and other anogenital cancers HPV infection is a necessary cause of cervical cancer; its genome can be found in nearly all cervical can- cers with the most sensitive detec- tion methods [2]. Comprehensive mechanistic and epidemiological data have firmly established the chain of causality, with relative risks of more than 100 in many case– control studies [2]. It is generally acknowledged that 100% of cervical cancers are attributable to 13 high- risk HPV types classified as carcino- genic or probably carcinogenic (i.e. HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68) [2]. HPV16 and 18 are the most frequently involved types and together account for about
70% of cervical cancer in all world regions [10].
In contrast to the universal pres- ence of HPV in cervical cancers, other anogenital cancer sites show a varying prevalence of HPV infection, depending on the cancer subtype, the age distribution, and the geo- graphical region of the population studied [11]. High-risk HPVs were found in approximately 88% of anal cancers, 70% of vaginal cancers, 50% of penile cancers, and 43% of vulvar cancers [11]. Compared with cervical cancers, a stronger predom- inance of HPV16 is reported. PAFs for each anogenital site are reported in Table 2.4.2.
Head and neck cancers Head and neck cancers represent a large and heterogeneous group of
malignancies, for which tobacco use and alcohol consumption are consid- ered the major causes. Over the past decade, HPV DNA has been reported to be present in a variable fraction of head and neck cancers (0–60%). Among HPV-positive cancers, HPV16 predominates (90% of HPV-positive carcinomas). HPV prevalence is no- tably high in oropharyngeal cancer [12], in which HPV16 is present at high copy number, localized to cell nuclei, frequently integrated into the cell genome, and actively transcrib- ing viral oncoproteins E6 and E7 [13]. Survival differences by HPV DNA presence have also been consis- tently reported exclusively for oro- pharyngeal cancers. These features support, for the moment, a causal role of HPV infection in some oropharyn- geal cancers, whose incidence is cur- rently increasing in some developed
Fig. 2.4.3. Human papillomavirus (HPV)-mediated progression to cervical cancer. HPV is thought to access the basal cells through microabrasions in the cervical epithelium. After infection, the early HPV genes E1, E2, E4, E5, E6, and E7 are expressed and the viral DNA replicates from episomal DNA. In the upper layers of epithelium (the midzone and superficial zone) the viral genome is replicated further, and the late genes L1 and L2, and E4 are expressed. L1 and L2 encapsidate the viral genomes to form progeny virions in the nucleus. LCR, long control region.
Chapter 2.4 • Infections 111
2 E
T IO
LO G
Y C
H A
P T
E R
2 .4
countries [14]. The HPV-related PAF was, therefore, estimated only for cancers of the oropharynx including the tonsils and the base of the tongue (ICD-10 [International Classification of
Diseases, 10th Revision] codes C01 and C09–C10) [3].
High-risk HPVs were found in approximately half of oropharyngeal cancers in developed countries, but this weighted average prevalence var- ied in different areas; it was generally higher in North America and northern Europe than in southern Europe. In the few developing countries with avail- able data, high-risk HPVs were found in 0–30% of oropharyngeal cancers [3]. The global PAF for oropharyngeal cancer is 26% (Table 2.4.2).
Epstein–Barr virus Hodgkin lymphoma EBV is more commonly associated with two Hodgkin lymphoma sub- types: nodular sclerosis and mixed cellular [2]. The EBV-related global PAF for Hodgkin lymphoma is 49% (Table 2.4.2); however, the PAF is
about 40% in adults in developed countries, whereas it is closer to 90% in children, and about 60% in adults in developing countries [3].
Burkitt lymphoma In endemic areas of sub-Saharan Africa where malaria burden is very high, the EBV-related PAF for Burkitt lymphoma is more than 95% [2]. Outside endemic areas, EBV tran- scriptional gene products are detect- ed much less frequently in Burkitt lymphoma. Sporadic forms of Burkitt lymphoma have been described elsewhere, but EBV was detected in only 20–30% of cases [3]. The EBV-related global PAF for Burkitt lymphoma is 63% (Table 2.4.2).
Nasopharyngeal carcinoma Nasopharyngeal carcinoma is a rare cancer worldwide but has an excep- tionally high incidence (10–30 per
Fig. 2.4.4. Precancerous changes caused by human papillomavirus infection can be detected in cervical tissue. Micrograph (400×) of a Pap test showing a low-grade squamous intraepithelial lesion with effects of human papillomavirus.
Do infections play a role in breast cancer? James F. Holland
Human breast cancer has many phenotypes and, often modulated by estrogens, probably results from many different etiologies. Wide variation in the incidence of breast cancer globally is often ascribed to reproductive patterns, ethnicity, and diet. Breast cancer in feral and labo- ratory mice is caused by the mouse mammary tumour virus (MMTV), discovered in 1936. In the 1990s, molecular techniques not previ- ously available were used to find a unique 660 bp sequence of MMTV that does not occur in the human ge- nome or in any other organism. After this sequence was detected using polymerase chain reaction, the en- tire 9.9 kb viral structure of human mammary tumour virus (HMTV), which is 90–95% homologous to MMTV, was isolated [1].
The distribution of HMTV is similar to that of indigenous mouse species. Mus domesticus, the com- mon house mouse, is dominant in countries in western Europe and in all of their former colonies, presum- ably distributed on sailing ships. M.
domesticus has abundant MMTV in its genome. In eastern Europe and Asia, different mouse species, with less MMTV in their genome, are in- digenous. In six countries of west- ern Europe and North and South America as well as in Australia, where breast cancer incidence is high, 30–40% of human breast cancer specimens contain HMTV sequences, but in four countries of Asia where breast cancer inci- dence is low, only 0–12% of speci- mens contain these sequences [2]. Normal tissues from the affected breast do not contain virus, thereby establishing that HMTV is acquired after conception and is not geneti- cally inherited [3,4]. The possibility of MMTV contamination has been excluded in these analyses.
HMTV can be seen budding from breast cancer cells by electron microscopy, and can infect human mammary epithelial cells in vitro, causing changes in gene expression, morphology, and migration, suggest- ing malignant transformation. It also can infect B and T lymphocytes. In
mice, MMTV in breast milk infects lymphocytes, which transfer virus to the breast. HMTV has been detected in the breast milk of 8% of healthy American women; the virus has been passed from mother to daughter via breast milk perhaps for eons, extend- ing the parallelism with mouse biol- ogy. These data raise the possibility that some breast cancers may have an infectious etiology, which would in turn open new perspectives on ther- apy and prophylaxis and raise ques- tions about other “portable DNA” in the etiology of human cancer.
References 1. Liu B et al. (2001). Cancer Res, 61:1754–
1759. PMID:11245493
2. Holland JF et al. (2004). Clin Cancer Res, 10:5647–5649. http://dx.doi.org/10. 1158/ 1078-0432.CCR-04-1234 PMID:15355888
3. Pogo BG et al. (2010). Cancer, 116:2741– 2744. http://dx.doi.org/10.1002/cncr.25179 PMID:20503403
4. Melana SM et al. (2010). J Virol Methods, 163:157–161. http://dx.doi.org/10.1016/j. jviromet.2009.09.015 PMID:19781575
112
100 000 person-years in men) in southern China, Singapore, and Malaysia [2]. For the non-keratinizing subtype, EBV genome or gene prod- ucts have been detected in virtually all cases, irrespective of geographical origin, and thus EBV appears to be a necessary step in the malignant pro- cess [2]. The EBV-associated PAF for nasopharyngeal carcinoma in areas of high or intermediate incidence is near- ly 100%. This figure is probably less (possibly about 80%) in areas of low incidence [3], but the lack of published data does not allow a precise estimate (the global PAF is 86%) (Table 2.4.2).
Other non-Hodgkin lymphomas Due to the heterogeneity of this group of tumours, even a crude es- timation of the PAF is not possible from the published data, although most NHLs arising in HIV-positive individuals and organ transplant recipients are causally related to EBV [15].
Gastric carcinoma Although EBV DNA has been found in 5–10% of gastric carcinomas, the IARC Monographs Volume 100B Working Group [2] concluded that there is insufficient epidemiological evidence for the involvement of the virus. Therefore, we have not calcu- lated the PAF for EBV and gastric cancer.
Kaposi sarcoma herpesvirus Kaposi sarcoma Kaposi sarcoma herpesvirus is a causal factor for the development of Kaposi sarcoma and a few rare lymphoproliferative disorders, in- cluding primary effusion lymphoma and multicentric Castleman disease [2]. Although virtually all cases of Kaposi sarcoma are attributable to Kaposi sarcoma herpesvirus [3], var- ious forms of the disease have been described and suggest that other factors, notably immunosuppres- sion associated with HIV or organ transplantation, have an important contributory role in cancer develop- ment. In some regions of Europe and North America, after years of dra- matically increasing Kaposi sarcoma incidence, a rapid decline of inci- dence has followed the introduction of highly active antiretroviral therapy in 1996 [16]. Conversely, with the large number of people with HIV/ AIDS in sub-Saharan Africa and the still-limited access to treatment, the estimated incidence of Kaposi sar- coma is, for example, about 40 per 100 000 person-years in Zimbabwe, and it is the most common cancer in men in a few sub-Saharan African countries.
Human T-cell lymphotropic virus type 1 Adult T-cell leukaemia A causal association between hu- man T-cell lymphotropic virus type 1 (HTLV-1) and adult T-cell leukaemia, found nearly exclusively in Japan, has been firmly established. Thus, 100% of adult T-cell leukaemia cases in Japan are attributable to HTLV-1 [3] (Table 2.4.2), although the global PAF is unclear [3]. Restriction of the duration of breastfeeding has dimin- ished the circulation of HTLV-1 and the corresponding cancer burden in Japan [2].
Macroparasite infections Chronic infections with the liver flukes Opisthorchis viverrini and Clonorchis sinensis have long been associated with cholangiocarcinoma (cancer of the bile ducts). Human liver fluke infection is endemic in ar- eas of China, the Republic of Korea, the Democratic People’s Republic of Korea, Thailand, Lao People’s Democratic Republic, Viet Nam, and Cambodia, where it is estimat- ed that a total of 24.4 million people are infected [2]. Flukes are acquired from eating contaminated raw fish, either fresh or fermented. A total of
Fig. 2.4.5. An HIV-infected H9 T cell. The image has been coloured to highlight HIV budding (yellow) against the T cell (blue).
Fig. 2.4.6. Unsafe needle use during an immunization day in rural Sylhet, Bangladesh. Contaminated needles have been implicated in the spread of infections such as HIV/ AIDS.
Chapter 2.4 • Infections 113
2 E
T IO
LO G
Y C
H A
P T
E R
2 .4
2000 cases of cholangiocarcinoma were attributable to liver flukes [3] (in- cluded in liver cancer in Table 2.4.2).
The causal association between the macroparasite Schistosoma haematobium and bladder cancer is strong and consistent [2]. Endemic areas are found in sub-Saharan Africa, Sudan, Egypt, and Yemen, where the snail that transmits the macroparasite is present. The num- ber of people infected with S. hae- matobium in sub-Saharan Africa is estimated to be 112 million, and the number of bladder cancers per year attributable to S. haematobium was estimated to be about 6000 [3]. The global PAF for bladder cancer is 2% (Table 2.4.2).
HIV HIV-positive individuals have an in- creased cancer risk. Advanced HIV
infection is characterized by immu- nosuppression, which is a risk factor for many malignancies [16]. There are three AIDS-defining cancers: Kaposi sarcoma, NHL, and cervi- cal cancer. In contrast to the gen- eral population, almost all NHLs in HIV-infected people are caused by EBV. HIV-infected people also have elevated risks of other virus-related cancers, notably anal cancer, hepa- tocellular carcinoma, and Hodgkin lymphoma. As HIV typically coexists with oncogenic viruses, the inclusion of HIV-associated malignancies in the computation of global PAF would mean that some infection-associat- ed cancers would be counted twice. However, HIV-infected people also have a higher prevalence of lifestyle- associated risk factors, including smoking and alcohol consumption, leading to increased risks of can- cers of the head and neck and lung.
Overall, in HIV-infected people the risk of non-AIDS-defining cancers is double that in the general popula- tion [16]. The advent of antiretroviral therapy led to a striking increase in the life expectancy of HIV-infected people. The cancer burden is set to increase as they age, and cancer pre- vention strategies need to be put in place in parallel with efforts to detect and treat HIV infection early enough to prevent immunosuppression.
Conclusions Conservative estimates indicate that about 2 million of the new cancer cases per year (16% of the global cancer burden) are attributable to a few chronic infections [3]. This fraction is substantially larger in de- veloping countries (26%) than in de- veloped countries (8%) (Fig. 2.4.1). Large variations are also seen by
Table 2.4.3. Number of new cancer cases in 2008 attributable to infectious agents, by geographical region
Region Number of new cases in 2008a Number attributable
to infection PAF (%)
Africa
Sub-Saharan Africa 550 000 180 000 32.7
North Africa and West Asia 390 000 49 000 12.7
Asia
India 950 000 200 000 20.8
Other Central Asia 470 000 81 000 17.0
China 2 800 000 740 000 26.1
Japan 620 000 120 000 19.2
Other East Asia 1 000 000 230 000 22.5
Americas
South America 910 000 150 000 17.0
North America 1 600 000 63 000 4.0
Europe
Europe 3 200 000 220 000 7.0
Oceania
Australia and New Zealand 130 000 4 200 3.3
Other Oceania 8 800 1 600 18.2
Developing regionsb 7 100 000 1 600 000 22.9
Developed regionsb 5 600 000 410 000 7.4
World 12 700 000 2 000 000 16.1
PAF, population attributable fraction. a Numbers are rounded to two significant digits. b Developed regions include Japan, North America, Europe, Australia, and New Zealand, i.e. high-income countries. Developing regions include all other countries, i.e. low- and middle-income countries.
114
continent and country (Table 2.4.3). The fraction of cancer attributable to infections is largest in sub-Sa- haran Africa (33%) and smallest in Australia, New Zealand, and North America (≤ 4%). A better under- standing of the association between NHL and infectious agents would lead to a substantial increase in the infection-associated PAF, especially in developed countries.
Prevention or eradication of these infections is a key tool to overcome inequalities in cancer incidence between populations in high-income countries and those in
low- and middle-income countries. Of note, substantial reductions in the occurrence of many cancer-associ- ated infections are now achievable and, in many instances, affordable. Control measures include avoid- ance of contaminated blood and needles (HBV, HCV, HIV), safe sex practices (HBV and HIV), and early detection and treatment of the infec- tion (e.g. detection and treatment of HPV-related precancerous cervical lesions in screening programmes, antibiotic treatment of H. pylori, and praziquantel against liver flukes). Very efficacious vaccines against
HBV and HPV are currently avail- able (see Chapter 4.6). Historically, only clean water has performed bet- ter than vaccination to reduce dis- ease burden. If high coverage can be achieved, vaccines can reduce inequalities in health more than other medical interventions can. Clinicians and public health profes- sionals should better appreciate the importance of infections that cause cancer and support the implementa- tion of currently available prevention strategies, especially in developing countries.
References 1. Doll R, Peto R (1981). The causes of
cancer: quantitative estimates of avoid- able risks of cancer in the United States today. J Natl Cancer Inst, 66:1191–1308. PMID:7017215
2. IARC (2012). Biological agents. IARC Monogr Eval Carcinog Risks Hum, 100B: 1– 441. PMID:23189750
3. de Martel C, Ferlay J, Franceschi S et al. (2012). Global burden of cancers attributable to infections in 2008: a review and synthetic analysis. Lancet Oncol, 13:607–615. http:// dx.doi.org/10.1016/S1470- 2045(12)70137-7 PMID:22575588
4. Ferlay J, Shin HR, Bray F et al. (2010). Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer, 127:2893–2917. http://dx.doi.org/10.1002/ ijc.25516 PMID:21351269
5. Helicobacter and Cancer Collaborative Group (2001). Gastric cancer and Helicobacter pylori: a combined analysis of 12 case control studies nested within prospective cohorts. Gut, 49:347–353. http://dx.doi.org/10.1136/gut.49.3.3 47 PMID:11511555
6. Parsonnet J, Hansen S, Rodriguez L et al. (1994). Helicobacter pylori infection and gastric lymphoma. N Engl J Med, 330:1267–1271. http://dx.doi.org/10.1056/ NEJM199405053301803 PMID:8145781
7. Raza SA, Clifford GM, Franceschi S (2007). Worldwide variation in the relative importance of hepatitis B and hepatitis C viruses in hepatocellular carcinoma: a sys- tematic review. Br J Cancer, 96:1127–1134. http://dx.doi.org/10.1038/sj.bjc.6603649 PMID:17406349
8. Donato F, Boffetta P, Puoti M (1998). A meta-analysis of epidemiological studies on the combined effect of hepatitis B and C virus infections in causing hepatocellu- lar carcinoma. Int J Cancer, 75:347–354. ht tp: //dx.doi.org /10.10 02 /(SICI)10 97- 0215(19980130)75:3<347::AID-IJC4> 3. 0.CO;2-2 PMID:9455792
9. Dal Maso L, Franceschi S (2006). Hepatitis C virus and risk of lymphoma and other lymphoid neoplasms: a meta-analysis of epidemiologic studies. Cancer Epidemiol Biomarkers Prev, 15:2078–2085. http:// dx.doi.org/10.1158/1055 -9965.EPI- 06 - 0308 PMID:17119031
10. Li N, Franceschi S, Howell-Jones R et al. (2011). Human papillomavirus type distribution in 30,848 invasive cervical cancers worldwide: variation by geograph- ical region, histological type and year of publication. Int J Cancer, 128:927–935. h t t p: / / d x . d o i . o r g / 10 .10 0 2 / i j c . 2 5 3 9 6 PMID:20473886
11. De Vuyst H, Clifford GM, Nascimento MC et al. (2009). Prevalence and type distri- bution of human papillomavirus in carci- noma and intraepithelial neoplasia of the vulva, vagina and anus: a meta-analysis. Int J Cancer, 124:1626–1636. http://dx.doi. org/10.1002/ijc.24116 PMID:19115209
12. Herrero R, Castellsagué X, Pawlita M et al.; IARC Multicenter Oral Cancer Study Group (2003). Human papillomavirus and oral can- cer: the International Agency for Research on Cancer multicenter study. J Natl Cancer Inst, 95:1772–1783. http://dx.doi. org/10.1093/jnci/djg107 PMID:14652239
13. Begum S, Cao D, Gillison M et al. (2005). Tissue distribution of human papillomavirus 16 DNA integration in patients with tonsil- lar carcinoma. Clin Cancer Res, 11:5694– 5699. http://dx.doi.org/10.1158/1078-0432. CCR-05-0587 PMID:16115905
14. Chaturvedi AK, Engels EA, Pfeiffer RM et al. (2011). Human papillomavirus and rising oropharyngeal cancer inci- dence in the United States. J Clin Oncol, 29:4294– 4301. http://dx.doi.org/10.1200/ JCO.2011.36.4596 PMID:21969503
15. Hjalgrim H, Engels EA (2008). Infectious aetiology of Hodgkin and non-Hodgkin lym- phomas: a review of the epidemiological ev- idence. J Intern Med, 264:537–548. http:// dx.doi.org/10.1111/j.1365 -279 6. 20 08. 02031.x PMID:19017178
16. Franceschi S, Lise M, Clifford GM et al.; Swiss HIV Cohort Study (2010). Changing patterns of cancer incidence in the early- and late-HAART periods: the Swiss HIV Cohort Study. Br J Cancer, 103:416– 422. http://dx.doi.org/10.1038/sj.bjc.6605756 PMID:20588274
Chapter 2.5 • Reproductive and hormonal factors 115
2 E
T IO
LO G
Y C
H A
P T
E R
2 .5
Summary • Reproductive and menstrual
factors are relevant to the etiol- ogy of breast, endometrial, and ovarian cancers.
• Obesity is a risk factor for these cancers in women as well as for some cancers in men; its influ- ence is most likely mediated through hormonal mechanisms, as further supported by the influ- ence of obesity on the effects of exogenous and endogenous hormones.
• Use of oral contraceptives sub- stantially reduces the risk of en- dometrial and ovarian cancers but appears to increase the risk of breast and of cervical cancers, consistent with grow- ing evidence for a possible role of hormonal factors in cervical carcinogenesis.
• Reproductive and hormonal factors appear to have particu- lar associations for different subtypes of cancers in women, including those defined by either histology or hormone receptor status.
• The role of hormonal factors in the etiology of some cancers in men is being revealed, although
2.5 2 ETIOLOGY
Reproductive and hormonal factors Louise A. Brinton Silvia Franceschi (reviewer)
David B. Thomas (reviewer)
risk relationships remain to be clarified.
Reproductive and hormonal factors are well recognized to play a major role in the etiology of many cancers in women. This is particularly true for breast, endometrial, and ovarian cancers, where such factors likely explain large proportions of disease occurrence. A few cancers in men may also be influenced by hormonal factors, although the relationships are less well defined.
Female breast cancer The role of reproductive factors in the etiology of breast cancer has been recognized for more than 100 years, beginning with the observation by Ramazzini of a high incidence of the disease in nuns. It is now well established that nulliparous women have approximately twice the risk of parous women, and that multipar- ity may be associated with further decreases in risk. Women with early age at first childbirth are at lowest risk, and risk rises steadily with later ages at first birth [1]. Interestingly, women with a first birth at age 30 or later are generally at higher risk than nulliparous women, presum- ably because of promotional effects of pregnancy on previously initiated
cells in older mothers. The effect of pregnancy on breast cancer risk is dependent on it being a full-term pregnancy; there is little evidence for relationships with short-term preg- nancies, including miscarriages and abortions.
The reduced risk associated with parity may be further enhanced if a woman decides to breastfeed. However, protection is likely depen- dent on longer periods of breastfeed- ing; thus, in most developed coun- tries, where numbers of births are limited and each child is breastfed for a relatively short period of time, there is little evidence of a relationship with breastfeeding. The most conclusive findings about the protective effects of breastfeeding derive from stud- ies where women have given birth to multiple children who have been breastfed for long periods of time, leading to long durations of cumula- tive breastfeeding.
Menstrual factors are also predic- tive of risk; early age at menarche and later age at natural menopause are associated with the highest risks, presumably reflecting in part an influ- ence of ovulatory activity (Fig. 2.5.1) [2]. Women who have an early surgi- cal menopause involving removal of both ovaries are at low risk, and those who undergo this operation before the age of 40 have approximately half
116
in endogenous hormonal profiles are involved, but additional research is needed to clarify effects. It is also unclear how hormonally induced changes in breast tissue are in- volved. Recent attention has focused on the effects of parity on involution of lobules, the structures from which the majority of breast cancers are thought to arise (Fig. 2.5.2) [3].
The relationship of obesity with breast cancer risk is complex; obe- sity is inversely related to risk of premenopausal-onset breast cancer and is directly associated with risk of postmenopausal breast cancer. Obesity-associated anovulation has been hypothesized as responsible for the decreased risk, while conver- sion of androgens to estrogens in adipose tissue appears to influence the increased risk. Menopausal hor- mone use has been associated with increased breast cancer risk among postmenopausal women, and the highest risks have been observed among thin women. The type of hor- mones used is also a major risk pre- dictor, with higher risks observed for
the risk of those who have a natural menopause after the age of 55.
Menstrual and reproductive fac- tors are major risk factors and can be used to estimate individual risks via the Gail Model Breast Cancer Risk Assessment Tool (http://www.
cancer.gov/bcrisktool/) and other risk prediction models. Despite the well- recognized role of these factors in breast cancer etiology, studies have been unable to relate them to specific underlying biological mechanisms. It is generally assumed that changes
Fig. 2.5.1. Relative risk of breast cancer by (A) age at menarche and (B) age at meno- pause, based on multiple studies. Calculated stratifying by study, age, year of birth, parity, age at first birth, smoking, alcohol consumption, height, and current body mass index. CI, confidence interval; g-s, group-specific; RR, relative risk.
Fig. 2.5.2. Terminal ductal lobular unit (TDLU) involution. (A) Minimal or no TDLU involution. (B) Marked involution demon- strating TDLUs containing few acini sur- rounded by dense collagen.
A
B
Chapter 2.5 • Reproductive and hormonal factors 117
2 E
T IO
LO G
Y C
H A
P T
E R
2 .5
use of estrogen plus estrogen than for unopposed estrogen therapy. This has been hypothesized as being due to mitotic influences of progestins on breast tissues.
Combined estrogen–progesto- gen oral contraceptives are associ- ated with an increased risk of breast cancer, notably among young wom- en (see Chapter 2.10). Endogenous
Fig. 2.5.4. Endogenous estrogen metabolism in women.
hormones clearly are important predictors of breast cancer risk, although it has been difficult for studies to fully define relationships with either breast cancer risk or patterns of risk factors. This most likely reflects difficulties in measur- ing hormones or the complexity of patterns of many interrelated mark- ers, including not only estrogens but also androgens, progesterone, prolactin, and insulin-like growth factors. In addition, the importance of large inter-individual differences in metabolism, which may have etiological implications, is being in- creasingly recognized. Recent pool- ing efforts have provided evidence that estrogens and androgens are directly related to both hormone- receptor-positive and -negative breast cancers (Fig. 2.5.3) [4], but
Fig. 2.5.3. Sex steroid hormones and sex hormone-binding globulin (SHBG) levels in relation to estrogen recep- tor (ER) breast cancer subtypes, from the European Prospective Investigation into Cancer and Nutrition (EPIC) study. CI, confidence interval; OR, odds ratio. * Test for heterogeneity between ER-positive and ER-negative breast cancer was assessed using a log likelihood ratio test with and without an interaction term for breast cancer subtype outcome. ** Log2 continuous increase in circulating hormone level.
118
Fig. 2.5.5. Age-standardized incidence of endometrial cancer by menopausal hormone therapy use and body mass index, from the United States National Institutes of Health- American Association of Retired Persons (NIH-AARP) Diet and Health Study. EPT, estrogen plus progestin therapy; ET, unopposed estrogen therapy; MHT, menopausal hormone therapy.
further clarity about relationships may derive from additional analyses that use more precise hormone mea- surement techniques. Recently de- veloped liquid chromatography-mass spectrometry assays are now allow- ing measurements of 15 individual estrogen metabolites (Fig. 2.5.4) [5], which are showing different ef- fects on breast cancer risk. A recent analysis within the Prostate, Lung,
Colorectal, and Ovarian Cancer Screening Trial indicated that al- though total estrogens were predic- tive of risk, further risk discrimina- tion was dependent on hydroxylation pathways [6].
Endometrial cancer Endometrial tissue is extremely hor- monally responsive, and endometrial
cancer is believed to arise as a result of estrogen stimulation that is unop- posed by progestins. Both obesity and menopausal hormone therapy are strong risk predictors. Particularly high risks have been noted for unop- posed estrogen use, which has been associated with 2–10-fold increases in risk, depending on the duration of use and the woman’s body size (with higher relative risks observed among thin women).
Much lower risks have been noted for estrogen plus progestin hormone use; in fact, some stud- ies suggest that relative risks may actually be lower among users than among non-users. These risks also appear to be modified by body mass, although in contrast to unopposed estrogens the greatest reductions in relative risks are seen among heavier women. Because of these complexities, more meaningful in- sights can be derived by a focus on absolute risks. The lowest risks are seen among thin women (either non- hormone users or users of continu- ous estrogen plus progestin therapy; these two groups are at similar risk), while the highest risks are observed among obese non-hormone users (who are at higher risk than obese users of continuous estrogen plus progestin therapy) (Fig. 2.5.5) [7]. Combination therapy may also be influenced by how it is prescribed (estrogens given sequentially vs
Table 2.5.1. Associations between history of endometriosis and the histological subtypes of ovarian cancer
Histological subtype Stratified and adjusted ORa (95% CI) P value
Invasive 1.46 (1.31–1.63) < 0.0001
Clear cell 3.05 (2.43–3.84) < 0.0001
Endometrioid 2.04 (1.67–2.48) < 0.0001
Mucinous 1.02 (0.69–1.50) 0.93
High-grade serous 1.13 (0.97–1.32) 0.13
Low-grade serous 2.11 (1.39–3.20) < 0.0001
Borderline 1.12 (0.93–1.35) 0.24
Mucinous 1.12 (0.84–1.48) 0.45
Serous 1.20 (0.95–1.52) 0.12
CI, confidence interval; OR, odds ratio. a Stratified by age (5-year categories) and ethnic origin (non-Hispanic White, Hispanic White, Black, Asian, other) and adjusted for duration of oral contraceptive use (never, < 2 years, 2– 4.99 years, 5 –9.99 years, ≥ 10 years) and parity (0, 1, 2, 3, ≥ 4). Pooled analysis of 13 ovarian cancer case–control studies: 1 in Australia, 3 in Europe, and 9 in the USA.
Chapter 2.5 • Reproductive and hormonal factors 119
2 E
T IO
LO G
Y C
H A
P T
E R
2 .5
to elevated risks of endometrial cancer in premenopausal women, for the more commonly used oral contraceptives, a combination of estrogen and progestin, use has been related to substantial risk reductions. Long-term users have the lowest risk, and reduced risk persists for some time after discontinuation of use. Although the progesterone content of the pills used might affect risk, studies have not been able to definitely confirm this hypothesis.
Despite the recognition of a strong role for hormonal factors in the etiology of endometrial cancers, relatively few studies have under- taken the assessment of the role of endogenous hormones. There is suggestive evidence that both estrogens and androgens deserve specific further attention. In such studies, it will be important to distin- guish patterns of risk according to specific tumour subtypes (e.g. type 1 vs type 2 endometrial cancers), which have been shown to be etio- logically heterogeneous.
Ovarian cancer Nulliparity is a well-recognized risk factor for ovarian cancers, as is infertility. Although there has been
extensive controversy about poten- tial effects of fertility drugs, the lat- est studies show that the indications for use are more important than the drugs themselves. Endometriosis is a well-established predictor of cer- tain types of ovarian cancers, in- cluding clear cell and endometrioid cancers (Table 2.5.1) [8].
Some studies suggest elevated risks with early age at menarche and late age at menopause, but the results are not entirely consistent. Substantially reduced risks have been observed among women who have had a simple hysterectomy or tubal ligation. Although this finding might reflect detection of abnormali- ties and removal of ovaries during either of these procedures, more recent attention has focused on the effects of partial devascularization or partial removal of tubes, given in- creasing evidence of the tubal origin of many serous cancers.
Use of oral contraceptives is related to substantial reductions in the risk of ovarian cancer, par- ticularly when long-term use is in- volved. However, use of menopausal hormones has been linked with in- creases in risk. This has been most clearly demonstrated for unopposed
continuously), although studies are only beginning to pursue this issue.
Whereas nulliparous women have high risks and multiparous women the lowest risks of develop- ing endometrial cancer, no effect on risk has been demonstrated accord- ing to age at first birth. Instead, age at or interval since last birth may be an important contributor to risk, al- though studies are still attempting to understand these relationships. Early age at menarche and late age at menopause are even stronger risk factors for endometrial cancer than for breast cancer, presumably be- cause these parameters indicate an enhanced opportunity for circulating estrogens to influence risk. Despite these patterns of risk, current un- derstanding of the effects of endog- enous hormones on endometrial cancer is still imprecise. Studies are needed to assess the independence of endogenous hormones from the strong effects on risk of obesity and of exogenous hormone use. Further study of estrogen metabolism pat- terns may also provide some clarity.
Whereas use of sequential oral contraceptives (estrogen-only pills followed by a limited number of days of progestin pills) has been related
Table 2.5.2. Associations between menopausal hormone therapy and ovarian cancer risk in the United States National Institutes of Health-American Association of Retired Persons (NIH-AARP) Diet and Health Study cohort, 1995–2006
Number of cancers Person-years RRa (95% CI)
Use of ET only, among 23 584 women with a hysterectomy at baseline
No MHT use 23 55 868 1.00 (reference)
ET only 76 116 139 1.69 (1.05–2.71)
Duration of use of ET only (years)
< 10 27 58 393 1.25 (0.71–2.20)
≥ 10 49 55 878 2.15 (1.30–3.57)
Use of EPT only, among 68 596 women with an intact uterus at baseline
No MHT use 150 316 239 1.00 (reference)
ET only 98 170 556 1.43 (1.09–1.86)
Duration of use of EPT only (years)
< 10 67 126 465 1.33 (0.98–1.79)
≥ 10 31 43 752 1.68 (1.13–2.49)
CI, confidence interval; EPT, estrogen plus progestin therapy; ET, unopposed estrogen therapy; MHT, menopausal hormone therapy; RR, hazard rate ratio. a RR adjusted for continuous age (years), race (White, other/unknown), parity (nulliparous, 1, 2, ≥ 3, unknown), duration of oral contraceptive use (none, < 10 years, ≥ 10 years, unknown), and body mass index (< 25 kg/m2, 25 –29 kg/m2, ≥ 30 kg/m2, unknown); models included terms for use of other MHT formulations.
120
estrogen therapy, but there is grow- ing evidence that combination es- trogen plus progestin therapy may also be linked with elevated risk (Table 2.5.2) [9].
Although many of the identified risk factors for ovarian cancer are consistent with a protective effect of reduced ovulation, this does not appear to entirely explain all of the identified risk factors. Recent atten- tion has focused on the possible role of hormonal and immunological factors [10]. Conflicting results have emerged about the respective role of estrogens, androgens, follicle-stimu- lating hormone, sex hormone-bind- ing globulin, and insulin-like growth factor. Further investigation appears warranted, especially with respect to specific ovarian cancer subtypes, for which there is growing evidence of etiological heterogeneity.
Cervical cancer Although infection with the human papillomaviruses is recognized as a necessary cause of cervical cancer, other co-factors are clearly impor- tant. Epidemiological studies have for some time implicated a variety of factors that could increase risk through hormonal mechanisms,
Fig. 2.5.6. Estrogen signalling pathways (A) and process of cervical carcinogenesis (B). (A) Estrogen-bound receptors dimerize and translocate to the nucleus, where they activate or repress target genes by binding to estrogen response elements (ERE) or to AP1 or Sp1 transcription factors. Estrogen also binds to the membrane receptor GPR30 and transduces various signalling pathways, including phosphatidylinositol 3-kinase (PI3K), mitogen-activated protein kinase (MAPK), and Ca2+ signalling. (B) High-risk human papillomaviruses (HR-HPVs), such as HPV16, infect and persist in multipotent reserve cells, resulting in atypical squamous metaplasia (ASM), which gives rise to cervical intraepithelial neoplasia (CIN) grades 1–3 and eventually cervical cancer. This progressive neoplastic disease process usually takes more than a decade after initial HPV infection.
Fig. 2.5.7. Risk factors for male breast cancer, from the United States National Institutes of Health-American Association of Retired Persons (NIH-AARP) Diet and Health Study. RR, relative risk.
such as oral contraceptive use and multiparity, but the role of hormonal factors in the etiology of cervi- cal cancer is not well understood. Transgenic mice models have re- cently provided evidence that estro- gen and its nuclear receptor promote cervical cancer in combination with human papillomavirus oncogenes (Fig. 2.5.6) [11]. If future studies can
demonstrate that human cervical cancers are dependent on estro- gens, it may be possible to develop hormonal agents to prevent or treat cervical cancers.
Testicular cancer Hormonal factors clearly play a role in the etiology of testicular cancer, as evidenced by the rise in incidence starting at adolescence and a vari- ety of risk factors including height, subfertility, and possibly exposure to endocrine disrupters. Several risk factors also support an influence of exposures received in utero, includ- ing cryptorchidism, hypospadias, in- guinal hernia, low birth weight, short gestational age, and being a twin, some of which may reflect the influ- ence of endogenous hormones [12]. Recent studies have attempted to assess the role of endogenous hor- mones in the etiology of testicular cancer, but further studies are need- ed to fully understand relationships.
Male breast cancer The incidence of breast cancers in men is only about 1% of the rate in women, complicating the evalua- tion of etiological factors. The few available studies, however, appear
Chapter 2.5 • Reproductive and hormonal factors 121
2 E
T IO
LO G
Y C
H A
P T
E R
2 .5
Fig. 2.5.8. Association between risk of prostate cancer and increasing fifths of hormone concentrations. The position of each square indicates the magnitude of the relative risk (RR), and the area of the square is proportional to the amount of statistical information available. The length of the horizontal line through the square indicates the 95% confidence interval (CI). The chi-square 1 degree of freedom statistic for linear trend is calculated by replacing the categorical variables with a continuous variable scored as 0, 0.25, 0.5, 0.75, and 1. The P value was two-sided for statistical significance of the chi-square linear trend statistic. DHEA-S, dehydroepi- androsterone sulfate; DHT, dihydrotestosterone; SHBG, sex hormone-binding globulin.
122
to implicate several hormonally related risk factors, with sugges- tions of increased risks related to gynaecomastia and Klinefelter syndrome and, in the most recent prospective study [13], to obesity, lack of physical activity, prior bone fractures, and current cigarette smoking (Fig. 2.5.7). Studies have not been undertaken to assess relationships with endogenous hormones.
Prostate cancer Prostate cancers respond well to an- ti-androgen therapies, and both sur- gical and medical castration results
in substantial reductions in risk of metastatic disease. Although it has also been assumed that androgens would play a role in the etiology of the disease, studies to date have not provided evidence of a role for any hormones as risk predictors. This includes an absence of association in a large pooling project (Fig. 2.5.8) [14] of testosterone, calculated free testosterone, and conversion prod- ucts; the major conversion product is dihydrotestosterone, to which testosterone is converted within the prostate by 5α-reductase. The only evidence of association observed was an inverse relationship with sex hormone-binding globulin.
Finasteride use reduces pros- tate cancer risk by blocking the conversion of testosterone to dihy- drotestosterone; use has also been associated with increases in estra- diol levels. The Prostate Cancer Prevention Trial has shown substan- tial reductions in prostate cancer in- cidence associated with exposure to finasteride (http://www.cancer.gov/ clinicaltrials/noteworthy-trials/pcpt/ Page1). This has raised concern as to whether estrogen levels might play a role in prostate cancer etiol- ogy. As with androgens, however, no relationships of prostate cancer risk with estrogen levels have been observed. Data from the trial, how- ever, have shown that participants who developed prostate cancer while taking finasteride experienced higher-grade tumours. Thus, it may be worthwhile for future studies to consider effects of estrogens and androgens on subgroups of prostate cancers as well as to consider ef- fects on risk of combined measures or estrogens and androgens as well as hormone metabolites.
Other cancers Although colorectal cancer is not classically considered as a hormon- ally related cancer, there is evidence supporting the possibility of inverse associations of colorectal cancer risk with use of both oral contraceptives and menopausal hormones. Oral contraceptives have also been noted to be a risk factor for liver cancer in the absence of infection with hepa- titis B virus, an important cause of this disease.
Fig. 2.5.9. One mother helps another to breastfeed her baby in Kolkata, India. Generally, reduced risk of breast cancer is associated with prolonged rather than short-term breastfeeding.
Chapter 2.5 • Reproductive and hormonal factors 123
2 E
T IO
LO G
Y C
H A
P T
E R
2 .5
References
1. Bernstein L (2002). Epidemiology of endo- crine-related risk factors for breast cancer. J Mammary Gland Biol Neoplasia, 7:3–15. http://dx.doi.org/10.1023/A:1015714305420 PMID:12160084
2. Collaborative Group on Hormonal Factors in Breast Cancer (2012). Menarche, menopause, and breast cancer risk: indi- vidual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies. Lancet Oncol, 13:1141–1151. http://dx.doi.org/10.1016/ S1470-2045(12)70425-4 PMID:23084519
3. Yang XR, Figueroa JD, Falk RT et al. (2012). Analysis of terminal duct lobular unit involu- tion in luminal A and basal breast cancers. Breast Cancer Res, 14:R64. http://dx.doi. org/10.1186/bcr3170 PMID:22513288
4. James RE, Lukanova A, Dossus L et al. (2011). Postmenopausal serum sex steroids and risk of hormone receptor- positive and -negative breast cancer: a nested case-control study. Cancer Prev Res (Phila), 4:1626–1635. http://dx.doi. org/10.1158/1940 - 6207.CAPR-11- 0090 PMID:21813404
5. Ziegler RG, Rossi SC, Fears TR et al. (1997). Quantifying estrogen metabolism: an evaluation of the reproducibility and validity of enzyme immunoassays for 2- hydroxyestrone and 16alpha-hydroxy- estrone in urine. Environ Health Perspect, 105 Suppl 3:607–614. PMID:9168003
6. Fuhrman BJ, Schairer C, Gail MH et al. (2012). Estrogen metabolism and risk of breast cancer in postmenopausal women. J Natl Cancer Inst, 104:326–339. http://dx.doi. org/10.1093/jnci/djr531 PMID:22232133
7. Trabert B, Wentzensen N, Yang HP et al. (2013). Is estrogen plus progestin meno- pausal hormone therapy safe with respect to endometrial cancer risk? Int J Cancer, 132:417– 426. http://dx.doi.org/10.1002/ ijc.27623 PMID:22553145
8. Pearce CL, Templeman C, Rossing MA et al.; Ovarian Cancer Association Consortium (2012). Association between endometriosis and risk of histological subtypes of ovarian cancer: a pooled analysis of case-control studies. Lancet Oncol, 13:385–394. http:// dx.doi.org/10.1016/S1470-2045(11)70404-1 PMID:22361336
9. Trabert B, Wentzensen N, Yang HP et al. (2012). Ovarian cancer and menopausal hormone therapy in the NIH-AARP Diet and Health Study. Br J Cancer, 107:1181– 1187. http://dx.doi.org/10.1038/bjc.2012.397 PMID:22929888
10. Ness RB, Cottreau C (1999). Possible role of ovarian epithelial inflammation in ovarian cancer. J Natl Cancer Inst, 91:1459–1467. http://dx.doi.org/10.1093/jnci/91.17.1459 PMID:10469746
11. Chung SH, Franceschi S, Lambert PF (2010). Estrogen and ERalpha: culprits in cervical cancer? Trends Endocrinol Metab, 21:504–511. http://dx.doi.org/10.1016/j.tem. 2010.03.005 PMID:20456973
12. McGlynn KA, Trabert B (2012). Adolescent and adult risk factors for testicular cancer. Nat Rev Urol, 9:339–349. http://dx.doi. org/10.1038/nrurol.2012.61 PMID:22508459
13. Brinton LA, Richesson DA, Gierach GL et al. (2008). Prospective evaluation of risk factors for male breast cancer. J Natl Cancer Inst, 100:1477–1481. http://dx.doi. org/10.1093/jnci/djn329 PMID:18840816
14. Roddam AW, Allen NE, Appleby P, Key TJ; Endogenous Hormones and Prostate Cancer Collaborative Group (2008). Endogenous sex hormones and prostate cancer: a collab- orative analysis of 18 prospective studies. J Natl Cancer Inst, 100:170–183. http://dx.doi. org/10.1093/jnci/djm323 PMID:18230794
124
Summary • Excess body fat increases risk
of cancers of the oesophagus, colon, pancreas, endometrium, and kidney, as well as postmen- opausal breast cancer.
• Regular physical activity re- duces risks of multiple can- cers by contributing to weight control, and risks of colorectal and breast cancer by additional mechanisms.
• Among dietary factors related to excess body weight, reduc- tion of consumption of sugar- sweetened beverages should be a high priority.
• High consumption of red meat, especially processed meat, is associated with risk of colorec- tal cancer.
• A diet high in fruits and vegeta- bles and whole grains does not appear to be as strongly protec- tive against cancer as initially believed. However, this dietary pattern is still advisable because of the benefits for diabetes and cardiovascular diseases, and some possible reductions in cancer incidence.
• Additional research is needed on many aspects of diet and
2.6 2 ETIOLOGY
Diet, obesity, and physical activity Walter C. Willett Tim Key Isabelle Romieu
Martin Wiseman (reviewer)
physical activity in relation to cancer, including the effects of these behaviours during child- hood and in early adult life.
Interest in diet as contributing to the cause and prevention of cancer emerged from observations that cancer rates vary greatly between countries and are correlated with dietary factors, and that nutrition can modify cancer incidence in ani- mals. To provide more specific in- formation for human cancer, many retrospective case–control studies, some prospective cohort studies, and a few randomized trials have been conducted. These studies have limitations: measurement of diet and physical activity is always imperfect, the diagnosis of cancer can distort the recall of diet in the ret- rospective studies, and adherence to the assigned diet is often poor in randomized trials (see “Challenges of measuring diet in cancer epide- miological studies and new perspec- tives”). In addition, initial nutritional status may affect response to chang- es in dietary intake. Despite the many methodological challenges, much has been learned during the past 30 years about the relation of diet, obesity, and physical activity to cancer; major findings are briefly de- scribed here. Because prospective
cohort studies are least susceptible to biases, findings from these stud- ies are emphasized, together with the findings from the small number of randomized trials.
Diet Fat Among dietary factors, fat has re- ceived the greatest attention due to strong international correlations with rates of several cancers common in developed countries. However, pro- spective studies have consistently shown little relationship of fat intake with breast cancer risk [1], even with up to 20 years of follow-up. In two large randomized trials of low-fat
Fig. 2.6.1. A man eats a hamburger and fried potato chips and drinks a pint of beer. Drinking alcohol, consumption of red and processed meats, and a high caloric di- etary intake are variously associated with increased risk of cancer at several ana- tomical sites.
Chapter 2.6 • Diet, obesity, and physical activity 125
2 E
T IO
LO G
Y C
H A
P T
E R
2 .6
Challenges of measuring diet in cancer epidemiological studies and new perspectives Nadia Slimani
Among the different environmental and lifestyle risk factors, diet is one of the most complex exposures to investigate in relation to diseases, particularly cancer. Indeed, diet is a universal exposure, and foods are consumed in combinations and with preparation methods that vary greatly between individuals and over the whole lifespan. In addition, the several thousand chemicals (includ- ing contaminants) present in foods may have complex synergistic or an- tagonistic bioactive effects, making it difficult to disentangle individual chemical effects and remove con- founding completely when investi- gating diet–disease relationships and underlying biological mecha- nisms [1]. Finally, diet has strong
social, religious, and psychological aspects that affect study designs and the dietary outcomes of indi- viduals [2].
In addition, the nutrition transi- tion, a phenomenon occurring with an accelerated pace worldwide [3,4], is another under-evaluated chal- lenge to measuring diet–cancer as- sociations. The nutrition transition is the shift in dietary patterns from traditional diets (high in cereal and fibre) to diets more typical of devel- oped countries and characterized by an increased contribution of (highly) industrially processed foods (high in sugars, fats, and animal sources) [4].
Cancer is a multiphase and mul- tifactorial disease that essentially occurs late in life but that might be
affected by different (early) expo- sure windows, which are difficult to evaluate. Most cancer epide- miological studies involve middle- aged populations and use single or limited repeated dietary measure- ments, which are not able to ap- proximate lifelong dietary exposure [5]. Furthermore, the food frequency questionnaire (FFQ) assessment method predominantly used in epi- demiology has been challenged with respect to its validity and reliability [6,7]. The FFQ is a retrospective method that asks respondents to report their usual frequency of con- sumption of each food from a list of foods for a specific period (sev- eral months or a year). In contrast, repeated 24-hour dietary recalls
Fig. B2.6.1. Integrative approaches for measuring diet in large-scale epidemiological studies. 24-HDRs, 24-hour dietary recalls; DQ, dietary questionnaire; FFQ, food frequency questionnaire; FPQ, food propensity questionnaire (non-quantitative FFQ).
126
diets, there was no significant effect on risk of breast cancer. Prospective studies of colorectal cancer have also not supported the positive rela- tionship with fat intake suggested by international comparisons of cancer rates [1,2]. Prospective studies of dietary fat and prostate cancer are fewer but do not generally support a relation with total or specific types of dietary fat [1,3].
For patients with cancer, the effect of diet after diagnosis on survival is of great interest. Two randomized trials of low-fat diets have been conducted among women with breast cancer. In one study, a marginally significantly lower risk of breast cancer was seen, but this might have been explained by a greater weight loss compared with the control group. In the other trial, no effect of the dietary intervention was observed [4].
Meat and dairy products Higher intake of red meat, especial- ly processed meat, has been asso- ciated with greater risk of colorectal cancer in many prospective stud- ies and in a meta-analysis of these studies [1] (see “Single-nucleotide polymorphisms relevant to meat consumption and cancer risk”). Although consumption of meat dur- ing midlife or later has generally not been associated with risk of breast cancer, a positive relation has been seen with intake in adolescence and early adult life [5]. Studies of red meat and risk of prostate cancer have been limited and results in- consistent [1].
Higher consumption of milk or dairy products has been associated with lower risk of colorectal cancer [1]. In contrast, higher consump- tion of dairy products has been associated with increased risks
of total or fatal prostate cancer in many studies [1]; the fat component of milk does not appear to account for these associations.
Fruits, vegetables, and phytochemicals Higher intakes of essential nutrients and other biologically active constit- uents of plants have reduced cancer incidence in laboratory animals and appear to have benefits in isolated cells. Although higher intakes of fruits and vegetables were associ- ated with lower risks of many can- cers in retrospective studies, these findings have generally not been strongly supported by prospective studies [1].
A large protective role for total in- take of fruits and vegetables against overall cancer risk now appears unlikely, but specific phytochemical
(24-HDRs) or food records (FRs), considered to be more precise and reliable than FFQs, are increasingly used or recommended in nutritional cancer research [8]. The 24-HDR is a retrospective assessment method in which an interviewer prompts a respondent to recall and describe all foods and beverages consumed during the preceding 24 hours or the preceding day. The FR is used to record food intake at the time of consumption, over a number of days that are not necessarily sequential.
Based on decades of methodo- logical research [9], front-line nutri- tional research increasingly favours approaches that integrate traditional and more innovative measurements of dietary exposure: (i) repeated 24-HDRs or FRs benefiting from cost-effective new innovative (web) technologies [10,11]; (ii) the FFQ or food propensity questionnaire (FPQ) approach, designed to rank individu- als and/or identify foods infrequently consumed, to complement 24-HDRs and other open-ended methods; and (iii) the use of recovery and/or con- centration nutritional biomarkers
as independent observations or as substitute estimates of dietary expo- sure [12,13] (Fig. B2.6.1). In addition, the rapid development of metabo- lomic research opens a promising perspective for a more holistic ap- proach to measure dietary exposure through individual and population- specific metabolomic profiles (or phenotypes) [14]. These approaches will be maximized when applied in large studies with wide variations in dietary exposures and disease outcomes.
References 1. Penn L et al. (2010). Genes Nutr, 5:205–213.
http://dx.doi.org/10.1007/s12263 - 010 - 0175-9 PMID:21052527
2. Gibney MJ et al., eds (2004). Public Health Nutrition. Oxford: Blackwell Science.
3. Popkin BM et al. (2002). Public Health Nutr, 5:947–953. http://dx.doi.org/10.1079/ PHN2002370 PMID:12633520
4. Popkin BM (2006). Am J Clin Nutr, 84:289– 298. PMID:16895874
5. Kristal AR, Lampe JW (2011). Cancer Epidemiol Biomarkers Prev, 20:725–726. http://dx.doi.org/10.1158/1055-9965.EPI- 10-1349 PMID:21546363
6. Kristal AR, Potter JD (2006). Cancer Epidemiol Biomarkers Prev, 15:1759– 1760. http://dx.doi.org/10.1158/1055-9965. EPI-06-0727 PMID:17021349
7. Willett WC, Hu FB (2006). Cancer Epidemiol Biomarkers Prev, 15:1757– 1758. http://dx.doi.org/10.1158/1055-9965. EPI-06-0388 PMID:17021351
8. Schatzkin A et al. (2009). Cancer Epidemiol Biomarkers Prev, 18:1026– 1032. http://dx.doi.org/10.1158/1055-9965. EPI-08-1129 PMID:19336550
9. Thompson FE, Subar AM (2008). Dietary assessment methodology. In: Coulston AM, Boushey CJ, eds. Nutrition in the Prevention and Treatment of Disease, 2nd ed. San Diego: Elsevier, pp. 3–39.
10. Subar AF et al. (2012). J Acad Nutr Diet, 112:1134–1137. http://dx.doi.org/10.1016/j. jand.2012.04.016 PMID:22704899
11. Touvier M et al. (2011). Br J Nutr, 105:1055–1064. http://dx.doi.org/10.1017/ S0007114510004617 PMID:21080983
12. Freedman LS et al. (2011). Am J Epidemiol, 174:1238–1245. http://dx.doi.org/10.1093/ aje/kwr248 PMID:22047826
13. Illner AK et al. (2012). Int J Epidemiol, 41:1187–1203. http://dx.doi.org/10.1093/ ije/dys105 PMID:22933652
14. Kaput J, Rodriguez RL (2004). Physiol Genomics, 16:166–177. PMID:14726599
Chapter 2.6 • Diet, obesity, and physical activity 127
2 E
T IO
LO G
Y C
H A
P T
E R
2 .6
or botanical subgroups may reduce risks of some cancers. Promising leads include carotenoid-containing vegetables and estrogen-receptor- negative breast cancer, cruciferous vegetables and several cancer sites including prostate, bladder, and lung; allium vegetables and stom- ach cancer; and folate-rich fruits and vegetables and colon cancer. High consumption of soy products, rich in anti-estrogenic compounds, during adolescence, but not during midlife or later, has been associated with lower risk of breast cancer [6]. More research is needed on these associations, and currently none of them is firmly established.
Dietary fibre Fibre has long been hypothesized to reduce risk of colorectal cancer by providing bulk to dilute potential carcinogens and speed their transit through the colon, by binding and thus inactivating carcinogens, or by altering the flora and biochemi- cal environment of the colon (see “Diet and the gut microbiome”). Several large prospective cohort studies of dietary fibre and colon cancer risk have not supported an association [7], although an in- verse relation was seen in the large European Prospective Investigation into Cancer and Nutrition (EPIC)
study and a recent meta-analysis [8]. The variation in findings from prospective studies needs to be better understood; dietary fibre is complex and heterogeneous, and the relation with colorectal cancer could differ by dietary source [9].
Higher intake of fibre has also been hypothesized to reduce risk of breast cancer by interrupting the enterohepatic circulation of estrogens. Although no relation with breast cancer has been seen in most prospective studies, a weak inverse relation was seen in a recent meta-analysis [10] and a new analysis from the large EPIC study [11].
Vitamins and minerals Calcium In a pooled analysis of large cohort studies, persons with the highest calcium intake had a 22% lower risk of colorectal cancer [12]; most of the benefit was achieved with about 800 mg/day. This benefit is sup- ported by reductions in recurrence of colorectal adenomas with calci- um supplementation in some ran- domized trials [1], but not all [13]. However, higher calcium intake has also been associated with a greater risk of total or advanced prostate cancer in several studies [1]. The
possibility of competing risks and benefits associated with higher in- take of dairy products and calcium has made recommendations about consumption difficult in relation to cancer, but adequate intakes of cal- cium are essential for bone growth and other aspects of health.
Vitamin D Higher blood levels of vitamin D, which reflect both intakes and sun- light exposure, have been consis- tently associated with lower risk of colorectal cancer [14], and recently with survival [15], but more research is needed to determine whether this reflects a true protective effect. Inconsistent associations have been seen with other specific cancers.
Folate Folate is important for DNA meth- ylation, repair, and synthesis, and in some prospective studies, higher intakes have been associated with lower risk of colorectal and several other cancers [16]. In a recent analy- sis, a lag of at least 12–14 years was seen between low folate intake and increased risk of colorectal cancer [4]. A role of folate in colorectal car- cinogenesis is supported by the con- firmed association between a genetic polymorphism in MTHFR, an enzyme involved in folate metabolism, and risk of colorectal cancer [17].
In randomized trials among pa- tients with a history of colorectal ade- nomas, supplements of folic acid have not reduced recurrent adenomas, and in one trial an increased risk of recur- rent advanced adenoma or multiple adenomas was seen [18]. These studies suggest that supplemental folic acid is unlikely to be beneficial for those with existing colonic neopla- sia and adequate folate intake, and might even be harmful. This concern was heightened by a report suggest- ing a small and transient increase in incidence of colorectal cancer in the USA almost immediately after the ini- tiation of folic acid fortification of grain products, but this transient increase was likely to have been an artefact
Fig. 2.6.2. In a remote semi-arid region of Laikipia North in Kenya, Masai warriors have exchanged their spears for cricket bats. Physical activity is associated with reduced risk of colorectal and other cancers.
128
Single-nucleotide polymorphisms relevant to meat consumption and cancer risk Rashmi Sinha
Meat is a source of a variety of car- cinogens, including heterocyclic a- mines (HCAs) and polycyclic aro- matic hydrocarbons (PAHs), which are formed in meats cooked at high temperatures, and N-nitroso compounds (NOCs), which are found in processed meats as a re- sult of nitrate or nitrite added during processing, and are formed endog- enously from haem iron. HCAs, PAHs, and NOCs undergo extensive metabolism by phase I (cytochrome P450s) and phase II (sulfotrans- ferases, glucuronosyltransferases, N-acetyltransferases) xenobiotic metabolizing enzymes (XMEs). Single-nucleotide polymorphisms (SNPs) in the genes encoding these XMEs can modify the abil- ity of these enzymes to activate or detoxify these meat carcinogens (Fig. B2.6.2).
While many studies have ob- served modifying effects of SNPs in XME genes on the association be- tween meat carcinogen intake and
colorectal cancer, the evidence for primary associations and interac- tions between XME polymorphisms, meat consumption, and the risk of colorectal adenomas or carcinomas has been mixed. One of the largest analyses to investigate these associ- ations conducted to date, a study of 1205 advanced colorectal adenoma cases and 1387 controls, revealed only one statistically significant in- teraction. The study reported an in- teraction between intake of the HCA 2-amino-3,8-dimethylimidazo[4,5- f ]quinoxaline (MeIQx) and the NAT1 polymorphism rs6586714 (Pinteraction = 0.001); among indi- viduals carrying a GG genotype, high MeIQx intake was positively associated with colorectal ade- noma (odds ratio [OR], 1.43; 95% confidence interval [CI], 1.11–1.85; Ptrend = 0.07), whereas those with the A variant had a decreased risk (OR, 0.50; 95% CI, 0.30–0.84; Ptrend = 0.01) [1]. The study also investigated substrate-oriented
pathway-based approaches, in- corporating a range of XME genes involved in both the activation and detoxification of xenobiotics, but there was no strong evidence for associations.
Two SNPs related to the xenobi- otic MDRI, thought to restrict intesti- nal absorption of HCAs and PAHs, were associated with colorectal cancer in one study [2]. In contrast, 8 genes involved in uptake, absorp- tion, and metabolism of dietary iron (TF, TFRC, HMOX1, SLC40A1, SLC11A2, HAMP, ACO1, HP01) were not associated with colorec- tal adenoma despite a statistically significant association between several SNPs and serum iron indi- ces [3]. The PARP gene in the base excision repair pathway appeared to modify the association between intake of red meat cooked at high temperatures and risk of colorectal cancer as meat carcinogens may induce oxidative DNA damage [4].
Fig. B2.6.2. Simplified schematic of polymorphic genes coding for enzymes involved in metabolism of heterocyclic amines (HCAs), polycyclic aromatic hydrocarbons (PAHs), nitrate/nitrate, and iron. AHR, aryl hydrocarbon receptor.
Chapter 2.6 • Diet, obesity, and physical activity 129
2 E
T IO
LO G
Y C
H A
P T
E R
2 .6
because it corresponded with a large increase in screening colonoscopy and an accelerated decline in colo- rectal cancer mortality rates [1,4].
Vitamin and mineral supplementation The role of vitamin and mineral sup- plements in cancer prevention has been examined in both prospective cohort studies and randomized tri- als. Trials of β-carotene and other single supplements, including vita- min E and selenium, have not shown benefits [1]. In trials using combina- tions of multiple vitamins or minerals at lower doses than those in single supplements, reductions in cancer
rates have been seen in a Chinese population with multiple nutrient deficiencies [19,20], in French men but not women, and in physicians in the USA [21].The study of physicians in the USA was notable because this population was well nourished, and the modest benefit was detectable only after 10 years, which is a period longer than that used in virtually all other studies. Overall, supplements with multivitamins and minerals are likely to be beneficial in populations with multiple nutrient deficiencies, although improvements in general nutrition should be the long-term goal in such populations.
Diet and excess adiposity Because excess adiposity is as- sociated with risk of many cancers [1], the relationship of diet to weight gain and obesity is important for cancer prevention. In trials with similar intensity of intervention and that have lasted at least 1 year, low- fat diets have not been effective in weight loss [22]. A diet low in rap- idly absorbed carbohydrates (such as sugar, jam, and refined cere- als) may facilitate weight loss [23]. A Mediterranean-style, restricted- calorie dietary pattern, which is high in cereals, fruits, and vegetables and low in animal products, has been ef- fective in sustaining weight loss for as long as 6 years [24]. Reduced consumption of soda and other sugar-sweetened beverages has reduced weight among children and adults who were initially overweight [25].
In an analysis of long-term weight gain in three large cohort studies, foods associated with greater weight gain included potato chips, sugar- sweetened beverages, red meat, and processed meat, whereas fruits, vegetables, whole grains, nuts, and yogurt were associated with less weight gain [26]. Among bever- ages, sugar-sweetened beverages and fruit juices were associated with greater weight gain. Because they lack any nutritional value and are directly related to adiposity, dia- betes, and cardiovascular disease, sugar-sweetened beverages are a high-priority focus area for weight control efforts in populations with substantial intakes.
Fig. 2.6.3. Relation of body mass index with risk of developing cancers of the oesopha- gus, colon, pancreas, breast, endometrium, and kidney.
Overall, the inconsistent findings reflect several limitations of existing studies, including the inability to ac- curately estimate the intake of spe- cific meat carcinogens, the limited number of well-characterized cases to ensure sufficient power by geno- type, and our current limited under- standing of the precise genetic vari- ants that account for XME function.
Very large studies that include both comprehensive meat intake informa- tion and numerous markers across multiple XME genes are essential for studying this complex association.
References 1. Gilsing AM et al. (2012). Carcinogenesis,
33:1332–1339. http://dx.doi.org/10.1093/ carcin/bgs158 PMID:22552404
2. Andersen V et al. (2009). BMC Cancer, 9:407. http://dx.doi.org/10.1186/1471-2407- 9-407 PMID:19930591
3. Cross AJ et al. (2011). Cancer Prev Res (Phila), 4:1465–1475. http://dx.doi. org/10.1158/1940-6207.CAPR-11-0103 PMID:21685236
4. Brevik A et al. (2010). Cancer Epidemiol Biomarkers Prev, 19:3167–3173. http:// dx.doi.org/10.1158/1055-9965.EPI-10- 0606 PMID:21037106
130
Overweight and obesity Overweight and obesity are important causes of several types of cancer. Body fatness is usually assessed using body mass index (BMI), which is calculated as the weight in kilo- grams divided by the square of the height in metres and thus provides a simple index of weight relative to height that can be measured eas- ily in population studies. A BMI of 20–24.9 kg/m2 is considered normal, whereas a BMI of 25–29.9 kg/m2 is classed as overweight and a BMI of 30 kg/m2 or more as obese. Waist circumference is another simple mea- sure of adiposity, and shows associa- tions with cancer risk similar to those for BMI.
Epidemiological studies have provided convincing evidence that obesity increases the risk of can- cers of the oesophagus (adeno- carcinoma), colon (in men), pan- creas, breast (postmenopausal), endometrium, and kidney [27] (Fig. 2.6.3); these associations of BMI with cancer risk generally fol- low a dose–response relationship, so that as well as the increased risk with obesity, there is also a smaller increase in risk for people who are overweight but not obese.
The magnitude of the increase in risk varies between cancer sites. For an increase in BMI of 10 kg/m2, rela- tive risks are approximately 2.3 for adenocarcinoma of the oesophagus, 1.5 for colon cancer in men, 1.3 for pancreatic cancer in women, 1.4 for postmenopausal breast cancer, 2.9 for endometrial cancer, and more than 1.5 for kidney cancer.
There is also substantial evidence that obesity increases the risks of cancer of the gall bladder (in wom- en), malignant melanoma, ovarian cancer, thyroid cancer, non-Hodgkin lymphoma, multiple myeloma, and leukaemia [1,28]. Cancer mortality (all sites combined) increases in a linear fashion with increasing BMI and is about 70% higher in people who are extremely obese than in people of normal weight (Fig. 2.6.4) [28].
Estimates of the percentage of cancers that can be attributed to excess body weight suggest that overweight and obesity are sub- stantial causes of cancer in many developed countries [29], but the magnitude of attributable risk has varied depending on the prevalence of obesity and on other underlying assumptions. For example, esti- mates for the United Kingdom for 2007 suggested that 5% and 6% of all incident cancers in men and women, respectively, were attribut- able to overweight and obesity, with attributable risks of about 40% for cancers of the oesophagus (adeno- carcinoma), gall bladder (in women), and endometrium [30], whereas es- timates for the USA for 1999–2000 suggested that overweight and obe- sity were responsible for 4.2% and
Fig. 2.6.4. Relation of body mass index with risk of death from cancer of any type in men and women who had never smoked.
Fig. 2.6.5. Children compete in a sack race in Banda Aceh, Indonesia. Such encourage- ment of physical activity may reduce the likelihood of overweight and obesity in childhood.
Chapter 2.6 • Diet, obesity, and physical activity 131
2 E
T IO
LO G
Y C
H A
P T
E R
2 .6
14.3% of cancer mortality in men and women, respectively, rising to 14.2% and 19.8%, respectively, in men and women who had never smoked [31]. These proportions are predicted to grow in most countries due to the increasing prevalence of overweight and obesity.
Additional research is required on the possible associations of overweight and obesity with the risk of less common types of can- cer. There is also a need to explore further the importance of obesity at different ages with lifetime risk of cancer, and to better understand
whether the distribution of body fat is important.
Mechanisms The mechanisms through which obesity increases cancer risk are only partially understood but ap- pear to vary between different cancer sites. For adenocarcinoma of the oesophagus, the increase in cancer risk probably involves an in- crease in the prevalence of chronic acid reflux from the stomach into the oesophagus, which damages the oesophageal epithelium. For breast and endometrial cancers, the increased risk with obesity in
postmenopausal women is prob- ably caused by the increase in cir- culating estradiol, due to increased formation of estrogens from precur- sor hormones in the adipose tissue; obesity also increases the risk of en- dometrial cancer in premenopausal women, and this may be due to anovulation and therefore reduced production of progesterone in obese premenopausal women. For colon cancer in men, kidney cancer, and other cancers, the mechanisms by which obesity increases risk are less clear but may involve increases in insulin and other hormonal changes in obesity.
Diet and the gut microbiome Johanna W. Lampe
The human gut is host to a large and diverse community of microbes that have physiological effects and carry out metabolic functions that can influence host health. Diet af- fects the amount and types of mi- crobes present in the gut, and, in turn, actions of the gut microbiota on dietary constituents may shape cancer susceptibility. Bacteria me- tabolize xenobiotics, both potentially beneficial (e.g. phytochemicals) and detrimental (e.g. carcinogens), and harvest otherwise inaccessible nutri- ents and/or sources of energy from the diet.
Gut microbes carry out unique metabolic reactions that the host cannot. Metagenomic studies of the gut microbiome (i.e. the collec- tive genomic content of a microbial community) have revealed the var- ied anaerobic reactions involved in fermentation and other processes. Enzymes specific to bacteria, and in some cases specific classes of bacteria, catalyse a range of meta- bolic reactions, such as hydrolysis of glycosides, amides, and esters, as well as reduction, ring cleavage, demethylation, and dehydroxylation. Plant polysaccharides (e.g. dietary fibre and resistant starch) are ma- jor dietary factors that mould the structure of the human gut microbial
community [1]. Gut bacterial metabo- lism of non-digestible carbohydrates produces fermentation end-prod- ucts, such as short-chain fatty ac- ids, that act as signalling molecules, serve as fuel to gut epithelial cells (butyrate) and peripheral tissues (ac- etate and propionate), and modulate inflammation. Hydrolysis of plant gly- cosides typically results in metabo- lites that are more biologically active than the parent compounds. Further bacterial transformation of aglycones can produce lower-molecular-weight compounds that systemically may be more or less active. High inter-indi- vidual variation in circulating concen- trations of phytochemicals and their metabolites is, in part, a reflection of variation in gut microbial activity [2]. Some microbial metabolites of dietary constituents are associated with increased risk of tumorigenesis. Ingested nitrate can be reduced to nitrite by bacterial nitrate reductase and produce N-nitroso compounds, which can form DNA adducts. Hydrogen sulfide, produced from sulfur-containing amino acids and inorganic sulfur sources by sulfate- reducing bacteria, has cytotoxic and genotoxic effects.
Gut microbes also promote the harvest of energy from the diet and contribute to fat deposition [3]. Thus,
the gut microbiome may influence adiposity and adiposity-associated inflammation, and therefore, indi- rectly, cancers for which excess body fat is a risk factor. Studies in animal models suggest that gut microbes play an important role in energy regulation and adiposity. In humans, the gut microbial commu- nity is altered in obese individuals and can change with weight loss; however, whether the structure and function of the gut microbial com- munity causes obesity in humans is unclear.
Understanding the complex and dynamic interaction between the gut microbiome and host diet may help elucidate mechanisms of carcinogenesis and guide future cancer prevention strategies.
References 1. Flint HJ (2012). Nutr Rev, 70 Suppl 1:S10–
S13. http://dx.doi.org/10.1111/j.1753-4887. 2012.00499.x PMID:22861801
2. Possemiers S et al. (2007). FEMS Microbiol Ecol, 61:372–383. http://dx.doi. org/10.1111/j.1574- 6941.2007.00330.x PMID:17506823
3. Tremaroli V, Bäckhed F (2012). Nature, 489:242–249. http://dx.doi.org/10.1038/ nature11552 PMID:22972297
132
Reversing the effect of obesity on cancer risk The clear evidence that overweight and obesity increase the risk of several types of cancer implies that weight loss should, to some extent, reverse this effect, and there is some direct evidence to support this un- derstanding. Observational cohort studies and randomized controlled trials of both dietary interventions and bariatric surgery have shown reductions in cancer incidence oc- curring within a few years after in- tentional weight loss [32].
Obesity and cancer survival Obesity may affect survival in can- cer patients [33]. Several studies of breast cancer patients have shown that obesity is correlated with a
References 1. World Cancer Research Fund/American
Institute for Cancer Research (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research.
2. Liu L, Zhuang W, Wang RQ et al. (2011). Is dietary fat associated with the risk of colorectal cancer? A meta-analysis of 13 prospective cohort studies. Eur J Nutr, 50:173–184. http://dx.doi.org/10.1007/s00394- 010-0128-5 PMID:20697723
3. Crowe FL, Allen NE, Appleby PN et al. (2008). Fatty acid composition of plasma phospholipids and risk of prostate cancer in a case-control analysis nested within the European Prospective Investigation into Cancer and Nutrition. Am J Clin Nutr, 88:1353–1363. PMID:18996872
4. Willett WC (2013). Nutritional Epidemiology, 3rd ed. New York: Oxford University Press.
5. Linos E, Willett WC, Cho E et al. (2008). Red meat consumption during adoles- cence among premenopausal women and risk of breast cancer. Cancer Epidemiol Biomarkers Prev, 17:2146–2151. http:// dx.doi.org/10.1158/1055 -9965.EPI- 08 - 0037 PMID:18669582
6. Lee SA, Shu XO, Li H et al. (2009). Adolescent and adult soy food intake and breast cancer risk: results from the Shanghai Women’s Health Study. Am J Clin Nutr, 89:1920–1926. http:// dx.doi.org /10. 3 9 4 5 /ajc n. 20 0 8 . 273 61 PMID:19403632
7. Park Y, Hunter DJ, Spiegelman D et al. (2005). Dietary fiber intake and risk of colorectal cancer: a pooled analysis of prospective cohort studies. JAMA, 294:2849–2857. http://dx.doi.org/10.1001/ jama.294.22.2849 PMID:16352792
8. Murphy N, Norat T, Ferrari P et al. (2012). Dietary fibre intake and risks of cancers of the colon and rectum in the European pro- spective investigation into cancer and nu- trition (EPIC). PLoS One, 7:e39361. http:// dx.doi.org/10.1371/journal.pone.0039361 PMID:22761771
9. Aune D, Chan DS, Lau R et al. (2011). Dietary fibre, whole grains, and risk of colorectal cancer: systematic review and dose-response meta-analysis of prospec- tive studies. BMJ, 343:d6617. http://dx.doi. org/10.1136/bmj.d6617 PMID:22074852
10. Aune D, Chan DS, Greenwood DC et al. (2012). Dietary fiber and breast cancer risk: a systematic review and meta-anal- ysis of prospective studies. Ann Oncol, 23:1394–1402. http://dx.doi.org/10.1093/ annonc/mdr589 PMID:22234738
poorer prognosis; this relationship is not well understood but may in- volve later diagnosis and also higher endogenous hormone levels in obese patients. Further research is needed on the effects of obesity on survival for breast cancer and for other types of cancer, including careful examination of relationships with stage at diagnosis and details of treatment.
Physical activity Physical activity is difficult to measure in epidemiological studies, and under- standing its relationship with cancer risk is complicated because it is gener- ally inversely related to obesity and is also correlated with other factors that can affect cancer risk. Considering all the available evidence, it is likely that
the most important impact of physical activity in relation to cancer preven- tion is through its contribution to protecting against weight gain and obesity. Thus, physical activity can contribute to reduction of risk of all types of cancer for which obesity in- creases risk. There is also evidence that physical activity reduces the risk of colon cancer and breast cancer and improves survival [34] indepen- dently of its effects on BMI, perhaps by acting through hormonal changes [1,27]. The type and amount of physi- cal activity required to reduce cancer risk remains uncertain, but the gen- eral consensus among researchers is that activity should be of at least moderate intensity and should aver- age at least 1 hour per day.
Chapter 2.6 • Diet, obesity, and physical activity 133
2 E
T IO
LO G
Y C
H A
P T
E R
2 .6
11. Ferrari P, Rinaldi S, Jenab M et al. (2013). Dietary fiber intake and risk of hormonal receptor-defined breast cancer in the European Prospective Investigation into Cancer and Nutrition study. Am J Clin Nutr, 97:344–353. http://dx.doi.org/10.3945/aj cn.112.034025 PMID:23269820
12. Cho E, Smith-Warner SA, Spiegelman D et al. (2004). Dairy foods, calcium, and colorectal cancer: a pooled analysis of 10 cohort studies. J Natl Cancer Inst, 96:1015–1022. http://dx.doi.org/10.1093/ jnci/djh185 PMID:15240785
13. Neuhouser ML, Wassertheil-Smoller S, Thomson C et al. (2009). Multivitamin use and risk of cancer and cardiovascular disease in the Women’s Health Initiative cohorts. Arch Intern Med, 169:294–304. http://dx.doi.org/10.1001/archinternmed. 2008.540 PMID:19204221
14. Jenab M, Bueno-de-Mesquita HB, Ferrari P et al. (2010). Association between pre- diagnostic circulating vitamin D concen- tration and risk of colorectal cancer in European populations: a nested case-con- trol study. BMJ, 340:b5500. http://dx.doi. org/10.1136/bmj.b5500 PMID:20093284
15. Fedirko V, Riboli E, Tjønneland A et al. (2012). Prediagnostic 25-hydroxyvitamin D, VDR and CASR polymorphisms, and survival in patients with colorectal cancer in western European populations. Cancer Epidemiol Biomarkers Prev, 21:582–593. http://dx.doi.org/10.1158/1055-9965.EPI- 11-1065 PMID:22278364
16. Kim DH, Smith-Warner SA, Spiegelman D et al. (2010). Pooled analyses of 13 pro- spective cohort studies on folate intake and colon cancer. Cancer Causes Control, 21:1919–1930. http://dx.doi.org/10.1007/ s10552-010-9620-8 PMID:20820900
17. Giovannucci E (2002). Modifiable risk factors for colon cancer. Gastroenterol Clin North Am, 31:925–943. http://dx.doi. org/10.1016/S0889 - 8553(02) 0 0 057-2 PMID:12489270
18. Cole BF, Baron JA, Sandler RS et al.; Polyp Prevention Study Group (2007). Folic acid for the prevention of colorectal adeno- mas: a randomized clinical trial. JAMA, 297:2351–2359. http://dx.doi.org/10.1001/ jama.297.21.2351 PMID:17551129
19. Qiao YL, Dawsey SM, Kamangar F et al. (2009). Total and cancer mortality after sup- plementation with vitamins and minerals: follow-up of the Linxian General Population Nutrition Intervention Trial. J Natl Cancer Inst, 101:507–518. http://dx.doi.org/10.1093/ jnci/djp037 PMID:19318634
20. Blot WJ, Li JY, Taylor PR et al. (1993). Nutrition intervention trials in Linxian, China: supplementation with specific vi- tamin/mineral combinations, cancer inci- dence, and disease-specific mortality in the general population. J Natl Cancer Inst, 85:1483–1492. http://dx.doi.org/10.1093/ jnci/85.18.1483 PMID:8360931
21. Gaziano JM, Sesso HD, Christen WG et al. (2012). Multivitamins in the prevention of cancer in men: the Physicians’ Health Study II randomized controlled trial. JAMA, 308:1871–1880. http://dx.doi.org/10.1001/ jama.2012.14641 PMID:23162860
22. Sacks FM, Bray GA, Carey VJ et al. (2009). Comparison of weight-loss diets with dif- ferent compositions of fat, protein, and car- bohydrates. N Engl J Med, 360:859–873. http://dx.doi.org/10.1056/NEJMoa0804748 PMID:19246357
23. Larsen TM, Dalskov SM, van Baak M et al.; Diet, Obesity, and Genes (Diogenes) Project (2010). Diets with high or low protein content and glycemic index for weight-loss maintenance. N Engl J Med, 363:2102–2113. http://dx.doi.org/10.1056/ NEJMoa1007137 PMID:21105792
24. Schwarzfuchs D, Golan R, Shai I (2012). Four-year follow-up after two-year dietary interventions. N Engl J Med, 367:1373–1374. http://dx.doi.org/10.1056/NEJMc1204792 PMID:23034044
25. Malik VS, Hu FB (2011). Sugar-sweetened beverages and health: where does the evidence stand? Am J Clin Nutr, 94:1161–1162. http://dx.doi.org/10.3945/ ajcn.111.025676 PMID:21993436
26. Mozaffarian D, Hao T, Rimm EB et al. (2011). Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med, 364:2392–2404. http:// dx.do i.o r g /10.10 5 6 / N E J M oa10142 9 6 PMID:21696306
27. IARC (2002). IARC Handbooks of Cancer Prevention, Vol. 6: Weight Control and Physical Activity. Lyon: IARC.
28. Renehan AG, Tyson M, Egger M et al. (2008). Body-mass index and incidence of cancer: a systematic review and meta- analysis of prospective observational stud- ies. Lancet, 371:569–578. http://dx.doi. org/10.1016/S0140 - 6736(08) 60269 -X PMID:18280327
29. Berrington de Gonzalez A, Hartge P, Cerhan JR et al. (2010). Body-mass index and mortality among 1.46 million white adults. N Engl J Med, 363:2211–2219. http://dx.doi.org/10.1056/NEJMoa1000367 PMID:21121834
30. Key TJ, Spencer EA, Reeves GK (2010). Obesity and cancer risk. Proc Nutr Soc, 69:86–90. http://dx.doi.org/10.1017/S00296 65109991698 PMID:19954565
31. Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ (2003). Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. N Engl J Med, 348:1625–1638. http://dx.doi.org/10.1056/ NEJMoa021423 PMID:12711737
32. Byers T, Sedjo RL (2011). Does intentional weight loss reduce cancer risk? Diabetes Obes Metab, 13:1063–1072. http://dx.doi. org/10.1111/j.1463 -1326.2011.0146 4.x PMID:21733057
33. Parekh N, Chandran U, Bandera EV (2012). Obesity in cancer survival. Annu Rev Nutr, 32:311–342. http://dx.doi.org/10.1146/an- nurev-nutr-071811-150713 PMID:22540252
34. Ballard-Barbash R, Friedenreich CM, Courneya KS et al. (2012). Physical activity, biomarkers, and disease outcomes in can- cer survivors: a systematic review. J Natl Cancer Inst, 104:815–840. http://dx.doi. org/10.1093/jnci/djs207 PMID:22570317
134
After the discovery that ciga- rette smoking is a major cause of cancer, and the development of modern epidemiological and toxi- cological methods, much more systematic and widespread efforts were undertaken to determine the causes of cancer, and many more causes were identified, in both
Summary • To date, 32 occupational agents
as well as 11 exposure circum- stances are identified as car- cinogenic to humans, and an additional 27 agents and 6 ex- posure circumstances primarily relevant to occupational expo- sure are probably carcinogenic to humans.
• Workplace exposure to several well-recognized carcinogens, such as asbestos, polycyclic aro- matic hydrocarbons, heavy met- als, diesel engine emissions, and silica, is still widespread.
• The burden of occupational can- cer among exposed groups may be substantial.
• Prevention of occupational can- cer is feasible and has taken place in industrialized countries during recent decades.
• Little information is available on occupational cancer risk in low- income countries, but it can be reasonably expected to become a large problem.
From the late 18th century until the early 20th century, remarkable num- bers of cases of scrotal cancer were
2.7 2 ETIOLOGY
Occupation Jack Siemiatycki Dana Loomis (reviewer)
Lesley Rushton (reviewer)
Fig. 2.7.1. A head nurse at a clinic out- side Accra, Ghana, withdraws blood from a woman for an HIV test. In discharging their duties, nurses in a variety of clini- cal circumstances may be exposed to biological and chemical agents identified as, or suspected to be, carcinogenic.
reported among chimney sweeps, coal tar and paraffin workers, shale oil workers, and mule spinners in the cotton textile industry. Remarkable numbers of lung cancer cases were reported among metal miners, and of bladder cancer cases among workers in a plant that produced dyestuffs from coal tar. During the first half of the 20th century, there were additional reports of cancer clusters among other identifiable occupational groups. Unexpectedly high numbers of occurrences of respiratory cancer were evident in such diverse occupational settings as nickel refineries, coal carboni- zation processes, chromate manu- facture, manufacture of sheep dip containing inorganic arsenicals, and manufacture of asbestos products [1]. Each of these discoveries was typically based on astute observa- tion of particular cases by a clinician, followed up with rather primitive ret- rospective cohort studies. In most of such instances of increased risk, the relevant information concerned a particular occupation or industry, with little or no information allow- ing risk to be attributed to particular chemicals. These high-risk occupa- tions constituted virtually the only known causes of human cancer until the discovery in the 1950s of the cancer-causing effects of ciga- rette smoking.
Chapter 2.7 • Occupation 135
2 E
T IO
LO G
Y C
H A
P T
E R
2 .7
occupational and non-occupational settings. Even today, however, oc- cupational carcinogens represent a large fraction of all known human car- cinogens. Although the discovery of occupational carcinogens provides an immediate means for preventing occu- pational cancer, the potential benefit of such discoveries goes beyond the factory walls since most occupation- al carcinogens are also found in the general environment and in consumer products, sometimes at concentra- tions as high as those encountered in the workplace. Examples include polycyclic aromatic hydrocarbons in charcoal-broiled foods, 2-naphthyl- amine in tobacco smoke, asbestos fibres in poorly maintained buildings, and atmospheric pollution by diesel engine emissions, benzene, and poly- chlorinated biphenyls. Furthermore, research on occupational carcino- gens can sometimes provide useful information for assessing risks in- curred at levels found in the general environment.
Specifying occupational carcinogens A large volume of epidemiologi- cal and experimental data concern cancer risks in different workplaces. This chapter includes a tabular sum- mary of current knowledge on oc- cupational carcinogens, the occupa- tions and industries in which they are found, and their target organs [2].
Although seemingly simple, drawing up an unambiguous list of occupational carcinogens is difficult. The first source of ambiguity is the
definition of an “occupational carcin- ogen”. Exposures to most occupa- tional carcinogens also occur in the general environment and/or in the course of using consumer products, and reciprocally, most environmen- tal exposures and those associated with using certain consumer prod- ucts, including medications, foods, and others, also occur in some occu- pational context. The distinction be- tween occupational and non-occu- pational exposures can be arbitrary. For instance, whereas exposures to tobacco smoke, sunlight, and immu- nosuppressive medications are gen- erally not identified as occupational exposures, there are people whose occupation results in them being in contact with these agents to a degree that would not otherwise occur. Also, whereas asbestos, benzene, diesel engine emissions, and radon gas are considered to be occupational car- cinogens, exposure to these agents is also experienced by the general population, and indeed many more people are probably exposed to these substances in the course of day-to-day life than are exposed at work. There is no simple criterion to distinguish “occupational” carcino- gens from “non-occupational” car- cinogens. Given this complexity, the following operational procedure is adopted: a carcinogen is considered to be “occupational” if the primary human exposure circumstance to the agent is in the workplace and/or if the main epidemiological studies that led to the identification of an ele- vated risk of cancer were undertaken among workers, and if notable num- bers of workers are exposed to the agent. Even this operational defini- tion is somewhat vague and requires judgement in its implementation.
Another source of ambiguity de- rives from the nature of the occupa- tional circumstances that have been investigated. For instance, there is evidence that exposure to soot may cause skin tumours and that expo- sure to 2-naphthylamine may cause bladder tumours, and it is likely that the historical observations of excess risks of scrotal cancer in chimney sweeps and of bladder cancer in dye
factory workers were respectively related to these two carcinogens. If there is persuasive evidence that a given chemical or mixture is carcino- genic – soot and 2-naphthylamine, in the current example – then this may be considered a relatively estab- lished proposition, requiring consid- eration of the level of exposure but little else. However, the knowledge that an occupation or industry is as- sociated with excess cancer risk is rooted in the specific time and place at which the research took place. In some instances, an occupationally characterized group may be shown to experience excess cancer risk but the causative agent is unknown, or at least unproven; examples are lung cancer among painters and bladder cancer among workers in the alumin- ium industry. An occupation does not in itself confer a carcinogenic risk; it is the exposures or conditions of work that may confer a risk. Thus, the statement that a given occupa- tion involves a carcinogenic risk is potentially misleading and should be considered in light of the different exposure circumstances that may be associated with a given occupation in different times or places.
IARC Monographs IARC Monographs provide authori- tative information for compiling a list of occupational carcinogens [3]. The objective of the Monographs Programme, which has been operating
Fig. 2.7.2. A mine worker in Burdwan, India. Workers engaged in mining a range of products experience increased risk of lung cancer.
Fig. 2.7.3. A man in Uruguay wears a protective mask while spray-painting a car. Work as a painter is linked to in- creased risk of lung cancer and bladder cancer.
136
Evaluating carcinogens: dioxins and dioxin-like substances Kurt Straif
Since 1997 [1], 2,3,7,8-tetrachlo- rodibenzo-para-dioxin (TCDD) has been classified by IARC as Group 1, carcinogenic to humans, based on limited evidence of carcinogenicity in humans, sufficient evidence in rodents, and strong evidence in hu- mans and animals for a mechanism via initial binding to the aryl hydro- carbon receptor (AhR), which leads to changes in gene expression, cell replication, and apoptosis. Thus, TCDD was one of the first agents to be upgraded to Group 1 on the basis of mechanistic evidence (Table B2.7.1).
There is now sufficient evidence from epidemiological studies for all cancers combined, making TCDD the first agent to be classified as Group 1 initially based on strong mechanistic data and later con- firmed by increased cancer risk in humans [2]. This highlights the ability of mechanistic information to
provide robust evidence of carcino- genicity. Sufficient evidence for an increased risk for all cancers com- bined is mainly derived from heavily exposed occupational cohorts. In addition, there is limited evidence for an increased risk of lung cancer, soft-tissue sarcoma, and non-Hodg- kin lymphoma.
Like TCDD, 2,3,4,7,8-penta- chlorodibenzofuran and 3,3′,4,4′,5- pentachlorobiphenyl (PCB 126) are complete carcinogens in experi- mental animals, and there is strong evidence that they act through the same AhR-mediated mechanism. Therefore, in the context of the re-evaluation of TCDD, these two chemicals were classified as Group 1. Recognizing the complexity of the mechanistic evaluation, the IARC Monographs Volume 100F Working Group decided to make evaluations only for these two in- dicator chemicals, and suggested
that future Monographs meetings be focused on other dioxin-like compounds.
PCB congeners can be catego- rized by their degree of chlorination, substitution pattern, and binding affinity to receptors. Twelve con- geners with high affinity for AhR are referred to as dioxin-like PCBs. PCBs are readily absorbed and dis- tributed in the body, and accumulate in adipose tissue. On the basis of sufficient evidence of carcinogenic- ity in experimental animals and hu- mans (based on an increased risk of melanoma), PCBs were classified as Group 1 [3]. In addition, dioxin-like PCBs were also classified as Group 1 on the basis of extensive evidence of an AhR-mediated mechanism of carcinogenesis that is identical to that of TCDD, and sufficient evidence of carcinogenicity in experimental animals. It was noted that the carcinogenicity of PCBs cannot
Table B2.7.1. Evolution of evaluations of 2,3,7,8-tetrachlorodibenzo-para-dioxin (TCDD) and other dioxin-like compounds
IARC Monographs volume (V) or supplement (S) (year)
Agent Degree of evidence of carcinogenicity Overall evaluation
of carcinogenicity to humansaIn humans In animals
V7 (1974), V18 (1978), S1 (1979), S4 (1982)
Polychlorinated biphenyls Inadequate Sufficient Group 2B
V15 (1977), S1 (1979) TCDD No overall evaluation V18 (1978) Polybrominated biphenyls No overall evaluation S4 (1982), S7 (1987) TCDD Inadequate Sufficient Group 2B V41 (1986), S7 (1987) Polybrominated biphenyls Inadequate Sufficient Group 2B S7 (1987) Chlorinated dibenzodioxins
(other than TCDD) Inadequate Inadequate Group 3
S7 (1987) Polychlorinated biphenyls Limited Sufficient Group 2A V69 (1997) TCDD Limited Sufficient Group 1b
V69 (1997) Polychlorinated dibenzo-para- dioxins (other than TCDD)
No data Inadequate or limited
Group 3
V69 (1997) Polychlorinated dibenzofurans Inadequate Inadequate or limited
Group 3
V100F (2012) TCDD Sufficient Sufficient Group 1 V100F (2012) 3,3′,4,4′,5-Pentachlorobiphenyl
(PCB 126) Sufficient Group 1b
V107 (2014) Polychlorinated biphenyls Sufficient Sufficient Group 1 V107 (2014) Dioxin-like polychlorinated
biphenylsc Sufficient Group 1b
V107 (2014) Polybrominated biphenyls Inadequate Sufficient Group 2Ab
a Group 1, carcinogenic to humans; Group 2A, probably carcinogenic to humans; Group 2B, possibly carcinogenic to humans; Group 3, not classifiable as to its carcinogenicity to humans; Group 4, probably not carcinogenic to humans. b Upgraded on the basis of mechanistic evidence. c With a toxic equivalency factor (TEF) according to WHO (PCBs 77, 81, 105, 114, 118, 123, 126, 156, 157, 167, 169, 189).
Chapter 2.7 • Occupation 137
2 E
T IO
LO G
Y C
H A
P T
E R
2 .7since 1971, is to publish critical
reviews of epidemiological and ex- perimental data on carcinogenicity for chemicals, groups of chemicals, industrial processes, other complex mixtures, physical agents, and bio- logical agents to which humans are known to be exposed, and to evaluate data indicative of carcinogenicity.
Agents are selected for evalua- tion on the basis of two criteria: hu- mans are exposed to the agent, and there is reason to suspect that the agent may be carcinogenic. Direct evidence of carcinogenicity of an agent can be derived from epidemio- logical studies or from experimental studies of animals (usually rodents). Account may also be taken of chem- ical structure–activity relationships and data indicative of mechanism, including absorption and metabo- lism of the agent and physiological change induced, together with mu- tagenic, toxic, and other effects ex- hibited by the agent. Expert Working Groups are convened to evaluate all relevant data.
Over the past 40 years, more than 100 meetings have been con- vened by the IARC Monographs Programme and almost 1000 agents have been evaluated. Exposure to many such chemicals or mixtures occurs occupationally.
Occupational agents or exposure circumstances evaluated as carcinogenic or probably carcinogenic Table 2.7.1 lists occupational agents, occupations, and industries that have been classified as Group 1, carcino- genic to humans. The table explicitly distinguishes 32 chemical or physi- cal agents from 11 occupations and industries that involve an increased
risk of cancer but for which the re- sponsible agent has not been speci- fied. The overall burden of cancer caused by a given agent is a func- tion of several factors, including the prevalence of the exposure, the type or types of cancer involved, and the relative risk of cancer induced as de- termined by each known exposure circumstance. Among the carcino- gens listed in Table 2.7.1 that may induce the largest excess numbers of cases are asbestos, diesel engine emissions, silica, solar radiation, and second-hand tobacco smoke [4]. Some of the carcinogens listed occur naturally, such as wood dust or solar radiation, whereas some are man-made, such as 1,3-butadiene or vinyl chloride. Some are single chemical compounds, such as ben- zene or trichloroethylene; others are families of compounds that include some carcinogens, and still others are mixtures of varying chemical
composition, of which diesel engine emissions and mineral oils are ex- amples. Most known human carcino- gens have been established to induce only one or a few different types of cancer.
Among the occupations and in- dustries shown in Table 2.7.1, most are industries in which the number of workers is relatively small, and so the population burden of any risk attribut- able to these industries would be lim- ited, at least in developed countries. But one occupational group – paint- ers – stands out as an occupation that is widespread on a population basis, and for which the agent or agents re- sponsible for the excess risk of lung and bladder cancer have not been identified. Aromatic amines such as benzidine and 2-naphthylamine may be responsible for some of the excess bladder cancer risk, but the cause of excess lung cancer risk is not so readily suggested.
be solely attributed to the carcino- genicity of the dioxin-like PCBs. Interestingly, particularly low-chlori- nated PCBs also seem to act via a gen- otoxic mechanism. Polybrominated biphenyls (PBBs) were classified as Group 2A, probably carcinogenic to humans, taking into account
evidence of an AhR-mediated mechanism of carcinogenicity.
These evaluations refer to can- cer hazard identification; the burden of cancer attributable to environ- mental exposures, for example via TCDD-contaminated food, is still controversial.
References 1. IARC (1997). IARC Monogr Eval Carcinog
Risks Hum, 69:1–631. PMID:9379504
2. IARC (2012). IARC Monogr Eval Carcinog Risks Hum, 100F:1–599. PMID:23189753
3. Lauby-Secretan B et al. (2013). Lancet Oncol, 14:287–288. http://dx.doi.org/10.1016/S1470- 2045(13)70104-9 PMID:23499544
Fig. 2.7.4. A man works in a toxic environment at a tannery in the densely populated area of Hazaribagh in Dhaka, Bangladesh. Occupational exposure to chromium compounds, specifically including work in chromate production, causes increased lung cancer risk.
138
Table 2.7.1. Occupational exposures, occupations, industries, and occupational circumstances classified as definite carcino- genic exposures (Group 1) by the IARC Monographs, Volumes 1–106
Agent, occupation, or industry Cancer site/Cancer Main industry or use
Chemical or physical agent Acid mists, strong inorganic Larynx Chemical 4-Aminobiphenyl Bladder Rubber Arsenic and inorganic arsenic compounds
Lung, skin, bladder Glass, metals, pesticides
Asbestos (all forms) Larynx, lung, mesothelioma, ovary Insulation, construction, renovation Benzene Leukaemia Starter and intermediate in chemical
production, solvent Benzidine Bladder Pigments Benzo[a]pyrene Lung, skin (suspected) Coal liquefaction and gasification, coke
production, coke ovens, coal-tar distillation, roofing, paving, aluminium production
Beryllium and beryllium compounds Lung Aerospace, metals Bis(chloromethyl)ether; chloromethyl methyl ether
Lung Chemical
1,3-Butadiene Leukaemia and/or lymphoma Plastics, rubber Cadmium and cadmium compounds Lung Pigments, batteries Chromium (VI) compounds Lung Metal plating, pigments Coal-tar pitch Lung, skin Construction, electrodes Diesel engine exhaust Lung Transportation, mining Ethylene oxide – Chemical, sterilizing agent Formaldehyde Nasopharynx, leukaemia Plastics, textiles Ionizing radiation (including radon-222 progeny)
Thyroid, leukaemia, salivary gland, lung, bone, oesophagus, stomach, colon, rectum, skin, breast, kidney, bladder, brain
Radiology, nuclear industry, underground mining
Leather dust Nasal cavity Shoe manufacture and repair 4,4′-Methylenebis(2-chloroaniline) (MOCA)
– Rubber
Mineral oils, untreated or mildly treated Skin Lubricant 2-Naphthylamine Bladder Pigments Nickel compounds Nasal cavity, lung Metal alloy Shale oils Skin Lubricant, fuel Silica dust, crystalline, in the form of quartz or cristobalite
Lung Construction, mining
Solar radiation Skin Outdoor work Soot Lung, skin Chimney sweeps, masons, firefighters 2,3,7,8-Tetrachlorodibenzo-para-dioxin (TCDD)
– Chemical
Tobacco smoke, second-hand Lung Bars, restaurants, offices ortho-Toluidine Bladder Pigments Trichloroethylene Kidney Solvent, dry cleaning Vinyl chloride Liver Plastics Wood dust Nasal cavity Wood
Occupation or industry, without specification of the responsible agent Aluminium production Lung, bladder – Auramine production Bladder – Coal gasification Lung – Coal-tar distillation Skin – Coke production Lung – Haematite mining (underground) Lung – Iron and steel founding Lung – Isopropyl alcohol manufacture using strong acids
Nasal cavity –
Magenta production Bladder – Painter Bladder, lung, mesothelioma – Rubber manufacture Stomach, lung, bladder, leukaemia –
Chapter 2.7 • Occupation 139
2 E
T IO
LO G
Y C
H A
P T
E R
2 .7
Table 2.7.2. Occupational exposures, occupations, industries, and occupational circumstances classified as probable carcino- genic exposures (Group 2A) by the IARC Monographs, Volumes 1–106
Agent, occupation, or industry Suspected target organ Main industry or use
Chemical or physical agent Acrylamide – Plastics Bitumens (combustion products during roofing)
Lung Roofing
Captafol – Pesticide α-Chlorinated toluenes (benzal chloride, benzotrichloride, benzyl chloride) and benzoyl chloride (combined exposures)
– Pigments, chemicals
4-Chloro-ortho-toluidine Bladder Pigments, textiles Cobalt metal with tungsten carbide Lung Hard-metal production Creosotes Skin Wood Diethyl sulfate – Chemical Dimethylcarbamoyl chloride – Chemical 1,2-Dimethylhydrazine – Research Dimethyl sulfate – Chemical Epichlorohydrin – Plastics Ethylene dibromide – Fumigant Glycidol – Pharmaceutical industry Indium phosphide – Semiconductors Lead compounds, inorganic Lung, stomach Metals, pigments Methyl methanesulfonate – Chemical 2-Nitrotoluene – Production of dyes Non-arsenical insecticides – Agriculture Polychlorinated biphenyls – Electrical components Polycyclic aromatic hydrocarbons (several apart from benzo[a]pyrene)
Lung, skin Coal liquefaction and gasification, coke production, coke ovens, coal-tar distillation, roofing, paving, aluminium production
Styrene-7,8-oxide – Plastics Tetrachloroethylene (perchloroethylene)
– Solvent
1,2,3-Trichloropropane – Solvent Tris(2,3-dibromopropyl) phosphate – Plastics, textiles Vinyl bromide – Plastics, textiles Vinyl fluoride – Chemical
Occupation or industry, without specification of the responsible agent Art glass, glass containers, and pressed ware (manufacture of)
Lung, stomach –
Carbon electrode manufacture Lung – Food frying at high temperature – – Hairdressers or barbers Bladder, lung – Petroleum refining – –
Occupational circumstance, without specification of the responsible agent Shiftwork that involves circadian disruption
Breast Nursing, several others
Table 2.7.2 lists occupational agents, occupations, and indus- tries that have been classified as Group 2A, probably carcino- genic to humans. The table explicitly distinguishes 27 chemical or physical agents from 5 occupations and industries that have been found to present a probable risk but for
which a causative agent has not been identified, and the other sin- gular occupational circumstance – shiftwork. Most agents in Table 2.7.2 are unequivocally carcinogenic in experimental animals, with little or no epidemiological evidence to confirm or contradict the evidence in animals. For a few of the agents,
including lead compounds and creo- sotes, there is a reasonable body of epidemiological evidence. However, the evidence taken together provides limited evidence of carcinogenicity to humans by IARC Monographs cri- teria, i.e. because bias, confound- ing, or chance cannot be excluded as contributing to the association
140
evident in epidemiological studies, or because different studies provide conflicting results.
Polycyclic aromatic hydrocar- bons pose a particular challenge in the identification of occupational carcinogens. This class of chemicals includes several potent experimen- tal carcinogens, such as benzo[a] pyrene, benz[a]anthracene, and dibenz[a,h]anthracene. However, hu- mans are always exposed to mixtures of polycyclic aromatic hydrocarbons; several such mixtures are indicated in Tables 2.7.1 and 2.7.2, including coal tars, soots, and creosotes. Because of the difficulty of isolat- ing the impact of specific polycyclic aromatic hydrocarbons in exposure assessment, it is difficult to evaluate human cancer risks associated with individual members of this class of chemicals. Only for benzo[a]pyrene has the evidence warranted a Group 1 evaluation, based on mechanis- tic data taken together with other available evidence, but there are probably many more individual poly- cyclic aromatic hydrocarbons that are carcinogenic to humans.
In the past, epidemiological re- search of occupational risk factors has largely focused on occupational exposures associated with “dirty”
industrial environments. In recent decades, however, occupational hygiene in many industries has im- proved or different technology has been adopted such that the histori- cal circumstances no longer apply, at least in developed countries. Increasing attention is now being paid to nonchemical agents in the work environment. Physical agents such as solar radiation and electromag- netic fields have been investigated, but behavioural and ergonomic char- acteristics of particular occupations, such as physical activity (or seden- tary behaviour) are now also recog- nized as occupational cancer risk fac- tors. Together with such factors may be included exposure to second-hand tobacco smoke at work. For almost all these risk factors, the distinction between occupational and non-occu- pational exposure is becoming more blurred. Although it is not of critical importance to maintain a clear dis- tinction between occupational and non-occupational factors, it does fa- cilitate communication and regulation to make this distinction, and the dis- tinction may be critically relevant to controlling the exposure in question.
Industries and occupations are constantly evolving. Even if we knew all there was to know about the cancer risks in today’s occupational environments – which we do not – continuing to monitor cancer risks in occupational settings would remain an important activity because oc- cupational exposure circumstances change over time and novel expo- sures may be introduced; a recent ex- ample is exposure to nanoparticles.
Although the lists of occupa- tional carcinogens and associated exposures shown in Tables 2.7.1 and 2.7.2 are long, they are not complete. There are likely many more occu- pational carcinogens that have not been discovered or properly docu- mented. For many, if not most, occu- pational circumstances, there is no relevant epidemiological evidence about carcinogenic risk. One of the foremost problems in occupational epidemiology is to reveal as-yet- unrecognized carcinogens and car- cinogenic risks.
Since the revolution in genetic research methods, there has been a shift in research resources on oc- cupational cancer, from an attempt to assess the main effects of occu- pations and occupational exposures to an attempt to assess genetic risk factors or, at best, gene–environ- ment interactions. This is an inter- esting and worthwhile pursuit, but it has not yet led to a proportionate increase in knowledge of new car- cinogens. It remains the case that almost all the knowledge that has accrued about occupational risk factors has been gained without re- course to genetic data.
Estimates of the burden of occupational carcinogens Over the years, there have been mul- tiple attempts, sometimes accompa- nied by controversy, to estimate what proportion of cancer cases are at- tributable to occupation. Estimating attributable fractions is feasible when the exposure factor is well defined and there is a body of evidence to support estimates of the magnitude of risk associated with the risk factor and the prevalence of exposure to the risk factor in the population, such as is the case for cigarette smoking.
Fig. 2.7.5. Asbestos fibres are the fine, elongated crystals that bind together to give asbestos its strength and durabil- ity. These fibres are naturally resistant to heat and electricity – the primary rea- son why asbestos was incorporated into thousands of commercial products like insulation and roofing materials. Among asbestos workers there is increased inci- dence of lung, larynx, and ovarian cancer and mesothelioma.
Fig. 2.7.6. An asbestos abatement work- er uses appropriate personal protective clothing and equipment in the course of removing asbestos. Despite bans on the use of asbestos, the material may often be encountered in the course of demoli- tion or rebuilding.
Chapter 2.7 • Occupation 141
2 E
T IO
LO G
Y C
H A
P T
E R
2 .7
None of these conditions exist for the generic class of occupational exposures. Apart from the difficulty of establishing an unequivocal list of occupational carcinogens, there are many obstacles to deriving reliable estimates of attributable fractions
for this class of carcinogens. These include the incompleteness of lists of occupational carcinogens, mea- gre information on quantitative rela- tive risks associated with exposure to known carcinogens, and scant information on the prevalence of
exposure, both historically and quantitatively, among workers. Furthermore, as with any estimates of attributable fractions, the results are rooted in the time and place at which estimates of exposure preva- lence are estimated. Still, by way of
Preventing occupational cancer: successes and failures Lesley Rushton
Approaches to preventing workplace exposures to occupational carcino- gens and reduction of occupational cancer include eliminating the pro- duction or use of carcinogens and controlling exposure to below a mini- mal risk exposure level, for exam- ple an occupational exposure limit (OEL). Examples of potential control measures are given in Table B2.7.2. All routes of exposure need to be considered, and control measures should not themselves cause addi- tional hazards.
Current occupational cancer is due to past exposures, sometimes over many years, and because of the long latency of many cancers, these (often high) exposures will contrib- ute to substantial occupational can- cers 20–30 years in the future [1]. Exposure to many established occu- pational carcinogens has gradually decreased over time in high-income countries due to an overall decline in heavy manufacturing industry and to regulatory activity and resulting risk reduction controls [2,3]. Some ex- posures will rapidly disappear in the future – for example, second-hand smoke, as the willingness of society to accept a ban on workplace smok- ing increases.
There has been mixed success in reducing other important expo- sures, such as asbestos, a major contributor to the total occupational cancer burden. Increasing calls for a total ban on asbestos production and use contrast with continuing mining and widespread use of prod- ucts containing chrysotile asbestos (an IARC Group 1 carcinogen), par- ticularly in developing countries.
In some circumstances, it can be demonstrated that certain risk re- duction strategies, such as improve- ment of compliance with a current OEL (e.g. for silica exposure) and targeting small- and medium-sized industries, would be more effective than other measures, such as OEL reduction [1]. Similar strategies may be relevant in many rapidly industri- alizing countries.
Reduction of high levels of expo- sure to occupational carcinogens re- mains an urgent priority in low- and middle-income countries, where the newest technology may be un- available, there is little or no regula- tion, and many may be employed in small workplaces, including children. Appropriate adaptation of effective regulation and control measures to different circumstances is required. However, the large numbers of
workers who continue to be exposed to low levels of occupational carcino- gens and who will contribute to the total occupational cancer burden must not be forgotten. Female work- ers are particularly affected; many are employed in service industries where low levels of many carcino- gens, including asbestos, occur. Concurrent exposure to multiple car- cinogens is of concern. Protection measures for one carcinogen may also simultaneously reduce others – for example, measures to reduce general dust; these will also poten- tially reduce non-malignant occupa- tional disease, such as respiratory ill health. A concerted international effort is required to prioritize strate- gies to reduce occupational cancer in all workplaces where exposure to known carcinogens is still occurring.
References 1. Hutchings SJ et al. (2012). Cancer Prev
Res (Phila), 5:1213–1222. http://dx.doi. org/10.1158/1940-6207.CAPR-12-0070 PMID:22961776
2. Cherrie JW (2009). Occup Med (Lond), 59:96–100. http://dx.doi.org/10.1093/occ med/kqn172 PMID:19233829
3. Symanski E et al. (1998). Occup Environ Med, 55:310–316. http://dx.doi.org/10.1136/ oem.55.5.310 PMID:9764108
Table B2.7.2. Measures to control workplace exposures to occupational carcinogens
Reduction method Examples of good practice
Worker education Provide information and training on all workplace carcinogens and the use of appropriate control methods. Use information media (e.g. posters, leaflets, data sheets) imaginatively and strategically.
Safer occupational practices Design and operate effective and reliable processes and activities to minimize exposure. Provide safe storage, handling, transportation, and disposal facilities. Use suitable personal protective equipment.
Surveillance Design and use procedures for regular measuring and monitoring of carcinogens in the workplace, including accidental high exposures.
142
illustration, recent estimates have been in the range of 4–8% of all cancers attributable to occupation- al cancer risk factors in developed countries [5,6,7], with higher fractions for lung cancer than for other cancers [6]. Estimates for other areas of the world are more speculative [8].
In developed countries, the hey- day of dirty, smoky industrial work- places was the first half of the 20th century. As a result of many social, economic, and technological forces in the past 50 years, in developed countries there have been declines both in the numbers of workers in- volved in “dirty” blue-collar work and in the concentration levels of pollutants in typical workplaces. In developing countries, however, the picture is reversed. There, indus- trial activities have undergone rapid growth, often in the absence of mean- ingful occupational hygiene control measures. Nor is there an infrastruc- ture to conduct research into possi- ble hazards in those environments. If the occupational environment in developing countries continues to be poorly regulated, a significant increase in occupational cancer
there may be expected in the com- ing decades.
Prevention The designation of an agent as carci- nogenic is an important public health statement, as well as a scientific one. Such a designation usually has im- plications for engineering and/or in- dustrial hygiene measures to reduce or eliminate occupational exposure to the agent. There are many pos- sible ways in which exposure to an agent may be prevented or at least reduced, depending on the agent and the nature of the industrial or occu- pational process in which the agent is used or produced (see “Preventing occupational cancer: successes and failures”). Sometimes one or another of the following classes of measures can be adopted: substitution of one type of raw material for another, modification or mechanization of an industrial process, and/or improved ventilation and/or the use of personal protective equipment. In addition to protecting workers, the designation of a carcinogen may trigger the need to exercise better control of emissions of that carcinogen into the atmo-
sphere, water, or soil, or changes to policies for marketing consumer and other products.
All of these policies and proce- dures would be facilitated through precise and reliable data on the magnitude of risks associated with different agents, and on the nature of dose–response relationships. Unfortunately, such information is not always available, or not in a form that facilitates intervention. Even the qualitative assessment of carcino- genicity is often rendered difficult by the fragility of data on exposure histories of workers. For quantitative risk estimation, the data sources are even more fragile, especially as the statistical modelling often requires an estimate of the exposure situation in the remote past. Thus, even if certain workers are known to have experi- enced an excess risk of cancer in the past, it is challenging to derive reli- able estimates as to whether current workers, known to be exposed to dif- ferent, often lower, levels of specified agents, would also be at risk. Equally challenging is the goal of determining a level of exposure such that carcino- genic risk is virtually eliminated.
References
1. Siemiatycki J, Richardson L, Boffetta P (2006). Occupation. In: Schottenfeld D, Fraumeni JF, Jr., eds. Cancer Epidemiology and Prevention, 3rd ed. New York: Oxford University Press, pp. 322–354.
2. Siemiatycki J, Richardson L, Straif K et al. (2004). Listing occupational carcino- gens. Environ Health Perspect, 112:1447– 1459. http://dx.doi.org/10.1289/ehp.7047 PMID:15531427
3. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. http:// monographs.iarc.fr/
4. Rushton L, Bagga S, Bevan R et al. (2010). Occupation and cancer in Britain. Br J Cancer, 102:1428–1437. http://dx.doi.org/10.1038/sj.bjc.6605637 PMID:20424618
5. Nurminen M, Karjalainen A (2001). Epidemiologic estimate of the proportion of fatalities related to occupational factors in Finland. Scand J Work Environ Health, 27:161–213. http://dx.doi.org/10.5271/sjweh. 605 PMID:11444413
6. Steenland K, Burnett C, Lalich N et al. (2003). Dying for work: the magnitude of US mortality from selected causes of death associated with occupation. Am J Ind Med, 43:461– 482. http://dx.doi.org/10.1002/ajim. 10216 PMID:12704620
7. Rushton L, Hutchings SJ, Fortunato L et al. (2012). Occupational cancer burden in Great Britain. Br J Cancer, 107 Suppl 1:S3– S7. http://dx.doi.org/10.1038/bjc.2012.112 PMID:22710676
8. Driscoll T, Nelson DI, Steenland K et al. (2005). The global burden of disease due to occupational carcinogens. Am J Ind Med, 48:419– 431. http://dx.doi.org/ 10.1002/ajim.20209 PMID:16299703
Chapter 2.8 • Radiation: ionizing, ultraviolet, and electromagnetic 143
2 E
T IO
LO G
Y C
H A
P T
E R
2 .8
Summary
• Exposure to all types of ionizing radiation, from both natural and man-made sources, increases the risk of various types of ma- lignancy; the risk is higher if the exposure occurs early in life.
• Cancer incidence rates among irradiated patients, workers in nuclear plants, and others un- derpin current consensus on the absence of a threshold for the induction of cancers by radia- tion and presumption of a linear dose–response relationship.
2.8 2 ETIOLOGY
Radiation: ionizing, ultraviolet, and electromagnetic Ausrele Kesminiene Joachim Schüz
Bruce K. Armstrong (reviewer) Sarah C. Darby (reviewer)
Fig. 2.8.1. The electromagnetic spectrum. The diagram shows several important divisions based on the properties and applica- tions of the different frequencies: pale blue, extremely low-frequency; orange, radiofrequency; green, microwave; red, infrared; dark blue, ultraviolet; yellow, X-ray.
• Exposure to ultraviolet radiation – both from the sun and from tanning devices – is established to cause all types of skin can- cers, including melanoma. Public health campaigns discourage de- liberate sun exposure and speci- fy options for sun protection.
• Associations between extremely low-frequency magnetic fields and cancer are restricted to in- creased risk of childhood leukae- mia, where a causal relationship has not been recognized.
• Associations between heavy use of mobile phones and cer- tain brain cancers have been
observed, but causal interpreta- tion is controversial; more data are needed, particularly on long- er-term use of mobile phones.
Natural and man-made sources gen- erate radiant energy in the form of electromagnetic waves. These are characterized by their wavelength, frequency, or photon energy. The electromagnetic spectrum extends from static (non-alternating) electric and magnetic fields to low-frequency electric and magnetic fields (low en- ergy; long wavelength), intermediate and radiofrequency electromagnetic fields, microwaves, optical radiation
144
(infrared, visible light, ultraviolet ra- diation), and ionizing radiation (high energy; very short wavelength) (Fig. 2.8.1). The interaction of vari- ous types of radiation with biological systems is generally well understood on a cellular level and, when estab- lished, is determined by the intensity of the radiation, the related energy, and the energy absorbed by the ex- posed tissue. Ionizing radiation and ultraviolet radiation are known to cause cancer, but questions remain about dose-related effects, suscep- tibility, and mechanisms.
Ionizing radiation By definition, ionizing radiation is sufficiently energetic to remove oth- erwise tightly bound electrons from atoms. Such radiation can be in the form of electromagnetic rays, such as X-rays or γ-rays, or in the form of subatomic or related particles, such as protons or neutrons, as well as α-particles and β-particles. Human exposure to environmental ionizing radiation is unavoidable [1].
Sources and exposures Natural radiation For the vast majority of people, expo- sure from natural sources accounts for most of the total annual dose of ionizing radiation (Fig. 2.8.2). The two major sources of natural radiation are cosmic rays and ra- dionuclides originating from the Earth’s crust, radiation from which is referred to as terrestrial radiation. Radionuclides are everywhere – in the ground, rocks, building materials, and drinking-water, as well as in the human body itself (sometimes desig- nated as internal sources). Inhalation of radon gas, which arises from the decay of radium-226, is the leading source of natural ionizing radiation exposure in humans, and this source is responsible for nearly half of the average annual dose.
Man-made sources The greatest contribution to ionizing radiation exposure from man-made sources is medical radiation. Other
man-made sources of exposure include consumer products, fall- out from weapons testing, nuclear accidents (as exemplified by the Chernobyl disaster), routine releas- es from nuclear installations, and occupational exposure. Exposures in the course of medical care arise from certain diagnostic procedures, such as radiography and computed tomography (CT), or as a conse- quence of treatment, most common- ly during radiotherapy for cancer. In high-income countries, medical radi- ation exposure is increasing rapidly due to expansion in use of CT, angi- ography, and radiographically con- trolled interventional procedures. In some countries, average exposure from medical radiation now exceeds that from natural sources.
Cancer causation Ionizing radiation is one of the most intensely studied carcinogens [1]. Evidence that ionizing radia- tion can cause human cancer has come from the follow-up of patients therapeutically irradiated for cancer
and for non-malignant conditions including ankylosing spondylitis and tinea capitis. This evidence is complemented by epidemiological studies involving people exposed to ionizing radiation as a result of nuclear accidents and atmospheric nuclear test sites and, most nota- bly, thermonuclear warfare in the Japanese cities of Hiroshima and Nagasaki. Lately, evidence on the carcinogenic effects of ionizing ra- diation has emerged from studies on populations with protracted low- dose radiation exposures in occu- pational settings and also from pa- tients who have undergone medical diagnostic procedures. These data are matched by tumorigenesis evi- dent in animal bioassays.
A variety of mechanistic investi- gations have been made to evaluate the effects of different types of ion- izing radiation, centred on the effects of variation in dose and exposure pattern, and with reference to cel- lular and molecular end-points. The energy-deposition characteristics of all sources of ionizing radiation cover
Fig. 2.8.2. Annual per capita effective dose of ionizing radiation to the global population (1997–2007) due to all sources, in millisievert (mSv). Per caput, per unit of population.
Chapter 2.8 • Radiation: ionizing, ultraviolet, and electromagnetic 145
2 E
T IO
LO G
Y C
H A
P T
E R
2 .8
a wide variety of molecular damage, specifically indicating DNA damage as the biological determinant of out- come from exposure [2,3]. However, only a small fraction of such changes result in malignant transformation. The processes through which cells harbouring DNA damage may give rise to neoplasia are not fully under- stood, although a basis for mutation is evident. However, demonstration of epigenetic effects and the role played by genomic instability and effects evident in surrounding cells not subject to radiation damage, so- called bystander effects, all attest to biological complexity. In addition, host factors such as age, sex, im- mune status, and genetic variations in specific genes may determine susceptibility.
X-radiation and γ-radiation The study of the survivors of the ther- monuclear bombings in Hiroshima and Nagasaki holds an important place in radiation epidemiology. The survivors were exposed primarily to γ-radiation, although there was also a neutron contribution. Leukaemia was the first cancer consequent upon the radiation exposure in this cohort [4] and exhibited the highest
relative risk – more than 5-fold – of any cancer. Analyses of leukaemia mortality during 1950–2000 estab- lished that the excess risk was larg- est among those exposed at younger ages, but the risk decreased more rapidly than for those exposed later in life [5]. Approximately 11% of the solid cancer cases among survivors of detonations who were exposed to 0.005 Gy or more are thought to have been caused by the radiation. Significant radiation-associated in- creases in incidence occurred for multiple cancers, including those of the oral cavity, oesophagus, stom- ach, colon, lung, breast, bladder, nervous system, and thyroid, as well as non-melanoma skin cancer [6].
In a large IARC study of nuclear industry workers from 15 countries exposed to protracted low-dose radiation, a statistically significant positive dose–response associa- tion was evident for lung cancer mortality; no other specific cancer category demonstrated a statistical- ly significant dose–response trend [7]. A study of the United Kingdom National Registry for Radiation Workers, which included many workers from the 15-country study but with a longer period of obser-
vation, found a positive association between radiation dose and mortali- ty due to leukaemia excluding chron- ic lymphocytic leukaemia, and also between radiation dose and mortal- ity due to all malignant neoplasms excluding leukaemia. Similar results were recorded in analyses of inci- dence of leukaemia and all malig- nant neoplasms [8].
Excesses of childhood leukae- mia incidence in populations near nuclear installations in the United Kingdom, Germany, France, and other countries have been recorded [9], but in the absence of information on other major causes of childhood leukaemia, which might confound possible exposure to ionizing radia- tion in these environments, the cor- rect interpretation of these findings is problematic [10].
Recent increases in the rates of CT use, which account for rela- tively high radiation doses in a broad cross-section of the community compared with doses attributable to conventional radiography, have raised health concerns despite the substantial immediate benefit of such scans to the individual patient when clinically indicated. In 2012, the first results were published from a historical cohort study of more than 175 000 patients without previous cancer diagnoses who were exam- ined with CT scans in Great Britain [11]. The study reported a positive association between radiation dose from CT scans and leukaemia and brain tumours. Potential increases in future cancer risk, attributable to the rapid expansion in CT use, have encouraged the design of new CT scanners with dose-reduction op- tions, as well as increased aware- ness among medical practitioners of the need to limit use of CT scans to situations involving a clear medi- cal benefit and to optimize CT radia- tion doses, particularly for paediat- ric patients. Optimization involves keeping radiation exposure as low as reasonably achievable for every examination. Various modality- and procedure-specific techniques are available, although they are not al- ways used.
Fig. 2.8.3. A technician at the University of Washington Medical Center, in Seattle, prepares a patient for a positron emission tomography-computed tomography (PET-CT) scan.
146
An IARC announcement that made waves Robert A. Baan
In the early evening of Tuesday 31 May 2011, a Working Group of experts invited by the IARC Monographs Programme concluded what had become widely known as the “mobile phone radiation meeting”. Although the Working Group decid- ed on a broad evaluation of “radio- frequency electromagnetic fields” – placing this agent in Group 2B, possibly carcinogenic to humans – the post-meeting media attention was focused almost exclusively on two sets of epidemiological data that showed a slight increase in the risk for glioma – a malignant type of brain cancer – among heavy users of wireless and cordless telephones. With 6 billion such users across the globe, never in the history of the IARC Monographs had an evalua- tion touched so many.
Indeed, the highest human ex- posure to radiofrequency radiation comes from a functioning mobile phone held against the ear during a voice call. Holding a mobile phone against the head can result in rela- tively high exposures in the brain, de- pending on the position of the phone
and its antenna, and the quality of the link with the base station: the better the connection, the lower the energy output from the phone. It is of interest to note that modern, third-generation (3G) phones emit, on average, about 1% of the energy emitted by the GSM (2G) phones that were in use when the above-mentioned epidemiological studies were conducted.
Also considered by the Working Group – but not mentioned in the media – were single-study findings of leukaemia or lymphoma among workers in plastic sealing and in ra- dar maintenance, where exposures to radiofrequency fields are relatively high as well. In addition, studies in experimental animals showed ef- fects with a variety of exposure sce- narios and frequency ranges, which led the Working Group to decide not to limit the evaluation to the two nar- row frequency bands used for wire- less communication, but to define the agent as “radiofrequency electromag- netic fields”. This being so, the hazard evaluation also encompasses expo- sures associated with, for example, base-station antennae, local area
networks (Wi-Fi), radio or television broadcast antennae, and smart me- ters. Because exposures in these cases are 3–4 orders of magnitude lower than those from mobile phones, the ensuing risk from these sources is likely to be much less. Nonetheless, this latter point has remained a major issue in the continuing debate since the IARC evaluation was announced.
Prospective epidemiological stud- ies – with accurate exposure assess- ment – still in progress now focus on possible effects of mobile phone use, also among youngsters. A large cancer bioassay in rodents is be- ing conducted by the United States National Toxicology Program. Also, the search for a plausible mechanism continues, seeking complementary in- formation to that from the epidemio- logical findings.
References
Baan R et al. (2011). Lancet Oncol, 12:624– 626. http://dx.doi.org/10.1016/S1470-2045 (11)70147-4 PMID:21845765
IARC (2013). IARC Monogr Eval Carcinog Risks Hum, 102:1–460.
Internalized radionuclides Internalized radionuclides that emit α-particles and β-particles are also carcinogenic to humans. For most people, exposure to ionizing radia- tion from inhaled and tissue-depos- ited radionuclides is mainly from naturally occurring radon-222. The epidemiological evidence that radon is a cause of lung cancer in humans is derived largely from cohort stud- ies of underground miners who were exposed to high levels of radon in the past [12]. Evidence has subsequently emerged from case–control studies in North America, Europe [13], and China [14] that environmental radon exposure in buildings, and especially in homes, acts as a cause of lung cancer in the general population. Until recently, naturally occurring ra- diation was perceived as unalterable.
However, it has now been shown that low radon concentrations can usually be achieved very cheaply if appro- priate radon control measures are installed in new buildings. Several national programmes and interna- tional recommendations, including a WHO Handbook [15], were de- signed to provide recommendations on reducing cancer risk from radon, together with policy options.
Another source of exposure to internalized radionuclides for rel- evant workers and for the general community has been the occurrence of accidents and other releases from nuclear power plants. The largest nuclear accident in history occurred on 26 April 1986 at the Chernobyl nuclear plant in northern Ukraine. The Chernobyl accident resulted in a large release of radionuclides –
of which the most important were iodine-131 and caesium-137 – that were deposited over a very wide area, particularly in Belarus, the western part of the Russian Federation, and Ukraine. In a large case–control study [16], exposure to iodine-131 in childhood was as- sociated with an increased risk of thyroid cancer: for a dose of 1 Gy, the estimated odds ratio of thyroid cancer varied from 5.5 (95% confi- dence interval [CI], 3.1–9.5) to 8.4 (95% CI, 4.1–17.3), depending on the risk model.
Ultraviolet radiation Sources and exposures Solar radiation is the main source of human exposure to ultraviolet radia- tion. In addition to this natural source,
Chapter 2.8 • Radiation: ionizing, ultraviolet, and electromagnetic 147
2 E
T IO
LO G
Y C
H A
P T
E R
2 .8
tanning lamps and beds are a com- mon source of exposure. Ultraviolet radiation is conventionally classified into three types: UVA (wavelengths of 315–400 nm), UVB (280–315 nm), and UVC (100–280 nm). An intact ozone layer in the atmosphere ab- sorbs almost all UVC, as well as ap- proximately 90% of UVB. Because UVA radiation is less blocked by the ozone layer, the radiation reaching the Earth’s surface is largely com- posed of UVA (95%), with just 5% UVB. The level of solar ultraviolet radiation exposure at the Earth’s sur- face varies with latitude, altitude, time of day, time of year, cloud cover, other atmospheric factors (specifically in- cluding air pollution), and reflection from nearby surfaces (e.g. snow, water). Tanning lamps and beds emit mainly UVA, with less than 5% UVB. Powerful tanning equipment may be a source 10–15 times as intense as midday sunlight on the Mediterranean Sea [17].
The most common acute skin reaction induced by exposure to ul- traviolet radiation is erythema (skin reddening), an inflammatory process in the skin, commonly called sun- burn; tanning is another response. Tanning of the skin is predominantly triggered by DNA damage induced by ultraviolet radiation. UVB is far more efficient than UVA in inducing a deep, persistent tan and is 1000 times as potent as UVA in induc- ing sunburn. Individual reactions to ultraviolet radiation depend highly on skin type, as classified by indi- viduals’ susceptibility to sunburn and tanning. Generally, people with fair skin are more susceptible to sun- burn and are less likely to develop a tan than people with olive or darker skin. A tan provides some protection against acute effects, and probably chronic effects, of sun exposure. Exposure to ultraviolet radiation also involves beneficial health ef- fects. Photosynthesis mediated by sunlight is the most common source of vitamin D; however, it appears that rather modest amounts of sun expo- sure are needed to increase vitamin D levels sufficiently [18].
Cancer causation Exposure to sunlight has been shown to be the main cause of skin cancer, including cutaneous malig- nant melanoma, basal cell carcino- ma, and squamous cell carcinoma [19]. Epidemiological evidence has established that sunbed use also increases skin cancer risk [17]. In 2009, radiation from tanning devices was classified as carcinogenic to hu- mans (Group 1) [1], the same as the earlier evaluation of solar radiation.
Although basal cell carcinoma and squamous cell carcinoma rep- resent the most frequent types of skin cancer and contribute to the ris- ing morbidity, cutaneous melanoma causes most skin cancer deaths because of its greater tendency to metastasize. Over the past 50 years, the incidence of all skin cancer types has steeply increased in Caucasian populations worldwide, with highest incidence rates where fair-skinned populations are exposed to intense ultraviolet radiation in countries such as Australia [20]. Individual risk varies widely, depending on en- vironment, behaviour, and genetic constitution. In addition to ultraviolet radiation, a sun-sensitive phenotype and other genetically determined characteristics influence risk of skin cancer. Melanoma at younger ages occurs mainly at body sites with in- termittent sun exposure, whereas in older people, who experience mela- noma more frequently on the head and neck, risk is thought be more closely related to chronic exposure [21]. Findings from seminal migrant studies indicate that childhood is a susceptible period for ultraviolet carcinogenesis. Risk of squamous cell carcinoma occurrence is related to the total cumulative lifetime solar exposure, whereas that for basal cell carcinoma is more complicated and is suggested to be more strongly as- sociated with intermittent exposure to ultraviolet radiation.
Prevention The first sun protection activities were initiated in Australia in the 1960s. In the early 1980s, the prima- ry prevention campaign “Slip! Slop!
Slap!” was introduced, advising peo- ple to slip on a shirt, slop on the sun- screen, and slap on a hat. In 1988, the more comprehensive and broad- based skin cancer prevention pro- gramme SunSmart was implement- ed. Comparable activities, including education programmes and public awareness campaigns, were imple- mented starting in the mid-1980s in the USA and in several European countries (Fig. 2.8.4). Several or- ganizations, including WHO, have developed recommendations about indoor tanning. Some jurisdictions have established laws banning provi- sion of commercial sunbed services to those under 18 years old.
Electric, magnetic, and electromagnetic fields Sources and exposures Ubiquitous exposure occurs to ex- tremely low-frequency electromag- netic fields due to power transmission and the use of electrical appliances, and to fields in the radiofrequen- cy range due to communication
Fig. 2.8.4. Danish Cancer Society and TrygFonden campaign for the preven- tion of skin cancer (2013): “A little shade does not hurt. Reduce your sun between 12 and 3 pm. Sunscreen is not always enough. Use the shade and reduce your risk of skin cancer.”
148
and broadcasting. Normal residential background exposure to extremely low-frequency magnetic fields is usually below 0.1 µT. A small frac- tion of households located very close to high-voltage power lines or other sources can have appreciably higher background exposures. Higher but short-term exposures occur when electrical devices are used and may also be experienced in particular cat- egories of work, such as by electri- cians and electrical engineers. For most people, the highest exposure to radiofrequency electromagnetic fields occurs when using mobile (cell) phones because the source of emission is held close to the head. Much lower levels of exposure arise from transmitters, but field strength may exceed 1 V/m even at points several kilometres from high-output television or radio broadcast trans- mitters. The number of sources con- tinues to increase with further use of the whole electromagnetic frequency spectrum.
Cancer risk Studies have been conducted in residential settings by investigating cancer risk in relation to the near- est overhead high-voltage power lines and the resulting magnetic fields, as well as in occupational settings that involve electrical work. Epidemiological studies have consistently recorded a positive association of extremely low-fre- quency magnetic fields with child- hood leukaemia, with an appar- ently 2-fold higher risk at average 24-hour exposure levels exceeding 0.3–0.4 µT (Fig. 2.8.5) [22–25]. However, a causal relationship has not been established due to the po- tential for bias and confounding in the observational studies and be- cause supporting evidence from experimental studies and mecha- nistic data are lacking [26]. If a causal association did exist, it is estimated that < 1– 4% of childhood leukaemia cases could be attribut- able to exposure to extremely low- frequency magnetic fields [27]. The 2001 IARC Monograph on extremely
low-frequency magnetic fields clas- sified them as possibly carcinogenic to humans (Group 2B); the evidence for other types of malignancy was evaluated to be inadequate. Other reviews came to similar conclusions later [22,28]. Recent studies have not shown an effect of exposure to extremely low-frequency magnetic fields on survival after childhood leukaemia [29].
In the 2001 IARC Monograph evaluation, extremely low-frequen- cy electric fields and static electric and magnetic fields were consid- ered not classifiable as to their car- cinogenicity to humans (Group 3). Since 2001, there have been few studies relevant to these evalua- tions and none suggest a basis for re-evaluation, as recently reflected by an expert panel of the European Commission [28].
Radiofrequency electromag- netic fields have been classified as possibly carcinogenic to hu- mans (Group 2B) (see “An IARC announcement that made waves”) (Fig. 2.8.6). Case–control studies on mobile phone use and cancer have reported increased risks of gli- oma and acoustic neuroma in heavy users of mobile phones [30]. A large Danish nationwide cohort study of mobile phone subscribers did not reveal any association with brain tu- mour risk. Such an increased risk was suggested from a series of in- terrelated case–control studies in 13 countries, Interphone, in which a 40% increased risk for glioma and also for acoustic neuroma was observed, restricted to the 10% of people who were the heaviest users of mobile phones. Several factors, including inaccuracy and evidence of bias in self-reported use, pre- vented causality being established by these studies [31]. Time trends in glioma incidence based on Nordic countries and the USA exclude any large increase in incidence attribut- able to mobile phone use, albeit with reference to a relatively short time from initiation of exposure. No asso- ciation was observed between mo- bile phone use and other cancers. Several studies on occupational
Fig. 2.8.5. Comparison of results from pooled analyses of epidemiological studies of residential exposure to extremely low-frequency magnetic fields and the risk of child- hood cancer: (A) childhood leukaemia [23]; (B,C) childhood leukaemia [24], excluding (B) and including (C) a study from Brazil; and (D) childhood brain tumours [25]. Pooled odds ratios and their 95% confidence intervals (vertical axis) are shown by increas- ing levels of exposure to extremely low-frequency magnetic fields (reference category, < 0.1 μT).
Chapter 2.8 • Radiation: ionizing, ultraviolet, and electromagnetic 149
2 E
T IO
LO G
Y C
H A
P T
E R
2 .8
exposure to radiofrequency elec- tromagnetic fields provide no con- sistent associations. With regard to environmental exposures from trans- mitters, including television, radio, and military transmissions as well as mobile phone networks, the evi- dence is inadequate due to lack of high-quality studies with accurate individual exposure assessment. Some large studies on childhood cancer and fields generated by high-output television and/or radio transmitters reported inconsistent or no associations [22]. Few data are available for electromagnetic fields in the intermediate frequency range [28].
Fig. 2.8.6. Mobile phone use has become very popular over the past few decades. Studies have been done mainly in relation to mobile phone use causing radiofrequency electromagnetic field exposure to the head.
150
References
1. IARC (2012). Radiation. IARC Monogr Eval Carcinog Risks Hum, 100D:1– 437. PMID:23189752
2. UN Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) (2000). Sources and Effects of Ionizing Radiation, Vol. I: Sources. New York: UN.
3. UN Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) (2000). Sources and Effects of Ionizing Radiation, Vol. II: Effects. New York: UN.
4. Folley JH, Borges W, Yamawaki T (1952). Incidence of leukemia in survivors of the atomic bomb in Hiroshima and Nagasaki, Japan. Am J Med, 13:311–321. http:// dx.doi.org/10.1016/0002-9343(52)90285-4 PMID:12985588
5. Preston DL, Pierce DA, Shimizu Y et al. (2004). Effect of recent changes in atomic bomb survivor dosimetry on cancer mor- tality risk estimates. Radiat Res, 162:377– 389. http://dx.doi.org/10.1667/RR3232 PMID:15447045
6. Preston DL, Ron E, Tokuoka S et al. (2007). Solid cancer incidence in atomic bomb survivors: 1958–1998. Radiat Res, 168: 1–64. http://dx.doi.org/10.1667/RR0763.1 PMID:17722996
7. Cardis E, Vrijheid M, Blettner M et al. (2007). The 15-country collaborative study of cancer risk among radiation workers in the nuclear industry: estimates of radiation- related cancer risks. Radiat Res, 167:396– 416. http://dx.doi.org/10.1667/RR0553.1 PMID:17388693
8. Muirhead CR, O’Hagan JA, Haylock RG et al. (2009). Mortality and cancer incidence following occupational radiation exposure: third analysis of the National Registry for Radiation Workers. Br J Cancer, 100: 206–212. http://dx.doi.org/10.1038/sj.bjc. 6604825 PMID:19127272
9. Laurier D, Jacob S, Bernier MO et al. (2008). Epidemiological studies of leukae- mia in children and young adults around nuclear facilities: a critical review. Radiat Prot Dosimetry, 132:182–190. http://dx.doi. org/10.1093/rpd/ncn262 PMID:18922823
10. Wakeford R (2013). The risk of childhood leukaemia following exposure to ionising radiation–a review. J Radiol Prot, 33:1–25. http://dx.doi.org/10.1088/0952-4746/33/1/1 PMID:23296257
11. Pearce MS, Salotti JA, Little MP et al. (2012). Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet, 380:499–505. http://dx.doi. org /10.1016/S0140 - 673 6(12) 6 0 815 - 0 PMID:22681860
12. BEIR VI (1999). Health Effects of Exposure to Radon. Washington, DC: National Academy Press.
13. Darby S, Hill D, Deo H et al. (2006). Residential radon and lung cancer – de- tailed results of a collaborative analysis of individual data on 7148 persons with lung cancer and 14,208 persons without lung cancer from 13 epidemiologic studies in Europe. Scand J Work Environ Health, 32 Suppl 1:1–83. PMID:16538937
14. Lubin JH, Wang ZY, Boice JD Jr et al. (2004). Risk of lung cancer and residential radon in China: pooled results of two studies. Int J Cancer, 109:132–137. http://dx.doi.org/ 10.1002/ijc.11683 PMID:14735479
15. WHO (2009). WHO Handbook on Indoor Radon: A Public Health Perspective. Geneva: WHO.
16. Cardis E, Kesminiene A, Ivanov V et al. (2005). Risk of thyroid cancer after expo- sure to 131I in childhood. J Natl Cancer Inst, 97:724–732. http://dx.doi.org/10.1093/jnci/ dji129 PMID:15900042
17. Boniol M, Autier P, Boyle P, Gandini S (2012). Cutaneous melanoma attributable to sunbed use: systematic review and meta- analysis. BMJ, 345:e4757. http://dx.doi.org/ 10.1136/bmj.e4757 PMID:22833605
18. Webb AR, Kift R, Durkin MT et al. (2010). The role of sunlight exposure in determin- ing the vitamin D status of the U.K. white adult population. Br J Dermatol, 163:1050– 1055. http://dx.doi.org/10.1111/j.1365-2133. 2010.09975.x PMID:20716215
19. MacKie RM, Hauschild A, Eggermont AM (2009). Epidemiology of invasive cutaneous melanoma. Ann Oncol, 20 Suppl 6:vi1–vi7. http://dx.doi.org/10.1093/annonc/mdp252 PMID:19617292
20. Erdmann F, Lortet-Tieulent J, Schüz J et al. (2013). International trends in the incidence of malignant melanoma 1953–2008 – are recent generations at higher or lower risk? Int J Cancer, 132:385–400. http://dx.doi. org/10.1002/ijc.27616 PMID:22532371
21. Whiteman DC, Watt P, Purdie DM et al. (2003). Melanocytic nevi, solar keratoses, and divergent pathways to cutaneous mel- anoma. J Natl Cancer Inst, 95:806–812. http://dx.doi.org/10.1093/jnci/95.11.806 PMID:12783935
22. Schüz J, Ahlbom A (2008). Exposure to electromagnetic fields and the risk of childhood leukaemia: a review. Radiat Prot Dosimetry, 132:202–211. http://dx.doi. org/10.1093/rpd/ncn270 PMID:18927133
23. Ahlbom A, Day N, Feychting M et al. (2000). A pooled analysis of magnetic fields and childhood leukaemia. Br J Cancer, 83: 692–698. http://dx.doi.org/10.1054/bjoc. 2000.1376 PMID:10944614
24. Kheifets L, Ahlbom A, Crespi CM et al. (2010). Pooled analysis of recent stud- ies on magnetic fields and childhood leukaemia. Br J Cancer, 103:1128–1135. http://dx.doi.org/10.1038/sj.bjc.6605838 PMID:20877339
25. Kheifets L, Ahlbom A, Crespi CM et al. (2010). A pooled analysis of extremely low- frequency magnetic fields and childhood brain tumors. Am J Epidemiol, 172:752– 761. http://dx.doi.org/10.1093/aje/kwq181 PMID:20696650
26. IARC (2002). Non-ionizing radiation, part 1: static and extremely low-frequency (ELF) electric and magnetic fields. IARC Monogr Eval Carcinog Risks Hum, 80:1–395. PMID:12071196
27. WHO (2007). Environmental Health Criteria 238: Extremely Low Frequency Fields. Geneva: WHO.
28. European Commission Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR) (2009). Health Effects of Exposure to EMF. Brussels: European Commission – Directorate-General for Health and Consumers.
29. Schüz J, Grell K, Kinsey S et al. (2012). Extremely low-frequency magnetic fields and survival from childhood acute lympho- blastic leukemia: an international follow-up study. Blood Cancer J, 2:e98. http://dx.doi. org/10.1038/bcj.2012.43 PMID:23262804
30. IARC (2013). Non-ionizing radiation, part 2: radiofrequency electromagnetic fields. IARC Monogr Eval Carcinog Risks Hum, 102:1–460.
31. INTERPHONE Study Group (2010). Brain tumour risk in relation to mobile telephone use: results of the INTERPHONE interna- tional case-control study. Int J Epidemiol, 39:675–694. http://dx.doi.org/10.1093/ije/ dyq079 PMID:20483835
Chapter 2.9 • Pollution of air, water, and soil 151
2 E
T IO
LO G
Y C
H A
P T
E R
2 .9
Summary
• Many known, probable, and possible carcinogens can be found in the environment, and all people carry traces of these pollutants in their bodies.
• Air pollution from vehicle emissions, power generation, household combustion of solid fuels, and a range of industries includes known carcinogens such as diesel emissions, poly- cyclic aromatic hydrocarbons, benzene, and 1,3-butadiene, together with inorganic carcin- ogens such as asbestos, arse- nic, and chromium compounds. Exposure to outdoor air pollu- tion in general, and specifically to particulate matter, causes lung cancer. Household com- bustion of solid fuels causes lung cancer and is associated with other cancers in many low- and middle-income countries.
• In drinking-water, inorganic ar- senic is a recognized carcino- gen. Other contaminants, such as disinfection by-products, or- ganic solvents, nitrates, nitrites, and some pesticides, may also contribute to an increased can- cer burden.
2.9 2 ETIOLOGY
Pollution of air, water, and soil Aaron J. Cohen Kenneth P. Cantor
Mark J. Nieuwenhuijsen (reviewer) José Rogelio Pérez Padilla (reviewer)
• There are wide disparities in ex- posure, and pollution levels can be particularly high in newly and rapidly industrializing countries where environmental monitoring and regulation are less exten- sive and rigorous.
• Pollution of air, water, and soil contributes to the world’s can- cer burden to differing degrees depending on the geographical setting.
• Exposure to important sources of environmental pollution in water and air can be reduced by regulatory action and techno- logical improvements, offering considerable promise for reduc- tion of the burden of cancer and other adverse health outcomes.
The impact of pollution on cancer development is often described with reference to “environmental factors”, which may lead to ambiguity. In a broad sense, environmental factors are implicated in the majority of hu- man cancers. Environmental factors may be understood to encompass everything that is not specifically ge- netic in origin. The term therefore includes many significant causes of cancer – such as tobacco smoking, alcohol consumption, and diet – that are also considered lifestyle or
behavioural factors. Evidence for the broad role of environmental factors comes from a variety of sources: from geographical variations in the distribution of the world cancer burden, from time trends showing increases or decreases in differ- ent forms of cancer, from studies of people migrating from one country to another, and from analytical stud- ies within populations that have both high and low exposure to selected environmental factors.
However, the term “environment” may also identify a prominent sub- set of factors that are distinguished from lifestyle factors because the individual has little, if any, control over exposure to these factors. As affecting all in the community, ex- posure to pollution occurs because of the chemical contamination of the air breathed, the water and food consumed, and the soil, sediment, surface waters, and groundwater that surround living space. Some, such as polycyclic aromatic hydro- carbons generated by the combus- tion of fossil and biomass fuels, are the direct result of human activity, while others, such as aflatoxins that contaminate foods, are generated by natural processes involving little or no human activity (see Chapter 2.11). Regardless, many carcinogens occur in the environment, and all
152
people carry traces of environmental pollutants in their bodies.
The cancer risks from environ- mental pollution are challenging to study. People are exposed to hun- dreds, if not thousands, of chemi- cals and other agents through their environment, and environmental exposure assessment, a neces- sary component of such studies, can be exceedingly complex. Some environmental pollutants are widely dispersed across the globe, while others are concentrated in small geographical areas near specific industrial sources. This results in wide disparities in the level of ex- posure to environmental pollutants. Some population groups may face high risks that do not have a no- ticeable impact on national cancer incidence statistics. IARC has identified environmental pollutants that are known to cause cancer in humans (Table 2.9.1) (http://mono graphs.iarc.fr/ENG/Classification/ index.php).
Asbestos Asbestos is one of the best charac- terized causes of human cancer in the workplace (see Chapter 2.7). The carcinogenic hazard associated with asbestos fibres has been recognized since the 1950s. Non-occupational exposure to asbestos may occur do- mestically and as a consequence of localized pollution. People who live with asbestos workers may be ex- posed to asbestos dust on clothes. The installation, degradation, remov- al, and repair of asbestos-containing products in the context of household maintenance represents another mode of residential exposure. Whole neighbourhoods may be exposed to asbestos as a result of local asbes- tos mining or manufacturing. Some parts of the world also experience asbestos exposure as a result of the erosion of asbestos or asbestiform rocks.
In common with occupational ex- posure, exposure to asbestos due to residential circumstances results in an increased risk of mesothelioma
[1]. Likewise, non-occupational ex- posure to asbestos may cause lung cancer, particularly among smokers [2]. A high incidence of mesothelio- ma as a consequence of neighbour- hood exposure is evident among inhabitants of villages in Turkey where houses and natural surround- ings contain the mineral erionite. All forms of asbestos are carcinogenic to humans [3], and stopping the use of all forms of asbestos is the most efficient way to eliminate asbestos- related diseases.
Outdoor air pollution Emissions from multiple sources, including motor vehicles, industrial processes, power generation, and the combustion of household solid fuel, pollute the ambient and indoor air in all populated regions of the globe [4]. The precise chemical and physical features of air pollution, which comprises hundreds of indi- vidual chemical constituents, vary around the world, reflecting differ- ences in the sources of pollution and
Table 2.9.1. Environmental pollutants classified by IARC as Group 1, carcinogenic to humans
Agent Cancer site/Cancer Main context of pollution for general population
Arsenic and inorganic arsenic compounds Lung, skin, bladder Water
Asbestos (all forms) Larynx, lung, mesothelioma, ovary Atmosphere
Benzene Acute non-lymphocytic leukaemia Atmosphere (engine emissions)
1,3-Butadiene Leukaemia, lymphoma Atmosphere
Chromium (VI) compounds Lung Water, soil
Diesel engine exhaust Lung Atmosphere (engine emissions)
Erionite Mesothelioma Atmosphere (particular regions only)
Ethylene oxidea Breast, lymphoid tumours Atmosphere (indoor)
Formaldehyde Nasopharynx, leukaemia Atmosphere (indoor, outdoor)
Household coal combustion (indoor emissions) Lung Atmosphere (indoor)
Outdoor air pollution Lung Atmosphere
Particulate matter in outdoor air pollution Lung Atmosphere
Polychlorinated biphenyls Skin Food, atmosphere (indoor)
Radon and its decay products Lung Atmosphere (indoor)
Silica dust, crystalline Lung Atmosphere
2,3,7,8-Tetrachlorodibenzo-para-dioxin (TCDD) All cancers combined Food, soil
Tobacco smoke, second-hand Lung Atmosphere (indoor)
Trichloroethylene Kidney Water, food
a Upgraded to Group 1 based on mechanistic data; limited evidence of carcinogenicity in humans.
Chapter 2.9 • Pollution of air, water, and soil 153
2 E
T IO
LO G
Y C
H A
P T
E R
2 .9
in meteorology, but these polluting mixtures contain specific chemicals or mixtures known to be carcinogen- ic to humans (Table 2.9.1) [5].
People are exposed to air pol- lution mixtures containing carcino- gens that may cause lung cancer or
cancers at other sites in a variety of settings, or micro-environments, both indoor and outdoor. A given individual’s total personal exposure to such mixtures or to specific car- cinogens is a function of the time spent in different micro-environments
characterized by a relatively homo- geneous concentration of the air pol- lution mixture or specific carcinogen. The actual dose of pollutants to the lung or other organs depends further on physiological characteristics such as pulmonary minute ventilation, metabolism, and rate of elimination. Typically, epidemiological studies rely solely on estimates of exposure to air pollution mixtures or specific carcino- gens in single micro-environments, such as place of residence, to esti- mate the risk of cancer [5].
Ambient air contains a variety of known human carcinogens, including polycyclic aromatic hydrocarbons, 1,3-butadiene, benzene, inorganic compounds such as arsenic and chromium, and radionuclides. Some, such as polycyclic aromatic hydrocar- bons, are present as part of complex mixtures such as diesel emissions, which are themselves carcinogenic [6]. Such mixtures often comprise carbon-based particles to which the organic compounds are adsorbed, sulfuric acid in particle form, and photochemical oxidants. Complex air pollution mixtures are character- ized for purposes of both regulatory policy and scientific research in terms of indicators, for example, the mass
Fig. 2.9.1. A traffic police officer observes exhaust pouring from an automobile in Salem, Tamil Nadu, India. Motor vehicles are a recognized source of air pollution worldwide.
Fig. 2.9.2. Estimated annual average atmospheric PM2.5 (particulate matter < 2.5 μm in diameter) concentrations (μg/m3), for 2005. An estimated 89% of the world’s population live in areas where the WHO air quality guideline (10 μg/m3 PM2.5 annual average) is exceeded.
154
Bisphenol A and cancer David Melzer and Tamara S. Galloway
The rising incidence of certain re- productive cancers since the 1970s has been attributed to multiple pos- sible factors, including exposure to endocrine disrupting agents in the environment. One widely studied compound is bisphenol A. This com- pound was first synthesized in the 1930s as an estrogen and is used as a loosely bound monomer in the production of polycarbonate plastic and in the epoxy resins lining food and beverage cans. More than 90% of people in the USA, Europe, and Asia are exposed to bisphenol A [1], probably through ingestion of bis- phenol A-contaminated food and liq- uids. There is growing evidence from epidemiological and laboratory stud- ies that exposure to this compound at levels found in the general popula- tion, about 0.2–20 ng/ml, may be as- sociated with adverse human health effects. Exposure to bisphenol A in utero or during childhood has also been associated with reproductive and developmental abnormalities [2], whereas in adults there is pro- spective evidence of its association with cardiovascular disease [3].
Whether exposure to bisphenol A is linked to cancer incidence is far from clear. Work in the 1980s by the United States National Toxicology Program based on laboratory studies with ro- dents found that the compound was not a robust carcinogen in the con- text of adult exposure and supported a recommended maximum daily dose of 50 µg/kg/day as safe for humans (www.epa.gov/iris/subst/0356.htm). Since then, hundreds of studies have shown that bisphenol A disrupts the hormone system and causes changes that mirror properties of cancer cells at concentrations within current limits of exposure, although this evidence
is contested [4,5]. Potential modes of action proposed for a carcinogenic effect include estrogenic endocrine disruption, promotion of tumorigenic progression, genotoxicity, and de- velopmental reprogramming that increases susceptibility to other car- cinogenic events [6].
Bisphenol A binds to and activates both types of estrogen receptor (ERα and ERβ), albeit with weaker affinity than estradiol itself [7]. Additional cel- lular targets have emerged that may shed further light on the potential link between this pollutant and cancer induction. For example, binding of bisphenol A to G protein-coupled receptors has been implicated in the proliferation of breast cancer cells and
recruitment of cancer-associated fi- broblasts [8]. Recent epidemiological studies have shown that exposure to bisphenol A is associated with enhanced expression of the orphan nuclear receptor estrogen-related re- ceptor α (ERRα) in white blood cells in vivo [9]. ERRα is upregulated in ad- vanced breast cancer cells and has been correlated with poor prognosis and the promotion of tumour growth [10].
References
1. Vandenberg LN et al. (2010). Environ Health Perspect, 118:1055–1070. http://dx.doi. org/10.1289/ehp.0901716 PMID:20338858
2. Diamanti-Kandarakis E et al. (2009). Endocr Rev, 30:293–342. http://dx.doi. org/10.1210/er.2009-0002 PMID:19502515
3. Melzer D et al. (2012). Circulation, 125:1482–1490. http://dx.doi.org/10. 1161/CIRCUL ATIONAHA .111.069153 PMID:22354940
4. Vandenberg LN et al. (2009). Endocr Rev, 30:75–95. http://dx.doi.org/10.1210/ er.2008-0021 PMID:19074586
5. Hengstler JG et al. (2011). Crit Rev Toxicol, 41:263–291. http://dx.doi.org/10.3109/104 08444.2011.558487 PMID:21438738
6. Keri RA et al. (2007). Reprod Toxicol, 24:240–252. http://dx.doi.org/10.1016/j. reprotox.2007.06.008 PMID:17706921
7. Delfosse V et al. (2012). Proc Natl Acad Sci U S A, 109:14930–14935. http:// dx.doi.org/10.1073/pnas.1203574109 PMID:22927406
8. Pupo M et al. (2012). Environ Health Perspect, 120:1177–1182. http://dx.doi. org/10.1289/ehp.1104526 PMID:22552965
9. Melzer D et al. (2011). Environ Health Perspect, 119:1788–1793. http://dx.doi. org/10.1289/ehp.1103809 PMID:21831745
10. Teng CT et al. (2011). In: Gunduz M et al., eds. Breast Cancer – Carcinogenesis, Cell Growth and Signalling Pathways. InTech. pp. 313–330.
concentration of fine particles (par- ticulate matter < 2.5 µm in diameter [PM2.5]) and ozone. The combustion of fossil fuels and biomass for power generation, cooking, and transpor- tation is the source of most of the
organic and inorganic compounds, acids, and oxidants, and contributes heavily to particulate air pollution in most populated areas [4,7].
Levels of ambient air pollutants, including PM2.5, have been declining
steadily in high-income countries over the past 20 years, although air pollution levels remain elevated in many urban areas and current levels are associated with a range of adverse health outcomes [4,7].
Fig. B2.9.1. Many food and liquid con- tainers, including baby bottles, are made of polycarbonate or have a lining that con- tains bisphenol A. Studies have shown that bisphenol A may increase cancer risk.
Chapter 2.9 • Pollution of air, water, and soil 155
2 E
T IO
LO G
Y C
H A
P T
E R
2 .9
Air quality has deteriorated, how- ever, in some middle-income coun- tries, including China and India, as a consequence of industrialization and an increased number of motor
vehicles. The uncontrolled burning of low-quality biomass fuels and coal contributes to poor air quality in low- and middle-income countries [8]. In 2005, estimated population-
weighted average annual levels of PM2.5 in East Asia were 4 times those in high-income North America, and 89% of the world’s population lived in areas where the WHO air quality guideline for exposure to PM2.5 was exceeded (Fig. 2.9.2) [7]. In the USA, hazardous air pollutants, or air tox- ics, are reported to exceed applicable reference concentrations, and have been estimated to increase cancer risk [9].
Exposure to air pollution mixtures has effects at the genetic and cellu- lar level that are considered likely to increase cancer risk. Urban air pollu- tion mixtures are mutagenic in in vivo assays and carcinogenic in rodents, these effects being largely mediated by combustion products [10]. Studies in human populations have identified markers of genetic damage and cell mutation associated with exposure to combustion emissions [11].
Long-term exposure to ambi- ent air pollution is associated with an increased risk of lung cancer. Populations in urban areas with el- evated ambient levels of polycyclic aromatic hydrocarbons and other
Fig. 2.9.3. The Niederaussem brown-coal power plant near Cologne, Germany. Coal- burning plants release carbon dioxide, fine particles, nitrogen oxide, and mercury, which can have a serious impact on public health. The coal ash produced by these plants contains contaminants such as arsenic, a known carcinogen.
Fig. 2.9.4. Proportion of the national population (%) estimated by WHO to use solid fuels for cooking, for 2010. Common solid fuels include wood, charcoal, dung, and crop residues, which are not only environmentally unsustainable but also harmful to the local communities. In homes without proper ventilation, exposure to indoor smoke poses serious health threats.
156
pollutants experience higher rates of lung cancer than rural popula- tions, independent of tobacco smok- ing. Higher rates of lung cancer are also associated with long-term residence near stationary air pollu- tion sources such as smelters, incin- erators, and power plants, and with residential proximity to busy roads [5,12]. Increased risk has been re- ported in studies of populations both in Europe and the USA and in countries in Asia [5,8,13]. Large co- hort studies of long-term exposure to PM2.5 in the USA and Europe that controlled for tobacco smok- ing report 6–36% increases in lung cancer associated with each 10 µg/m3 increase in PM2.5, an increment well within the range of exposure in North America and Europe, although the highest recorded levels occur in Saharan Africa, the Middle East, and Central Asia (Fig. 2.9.2). These stud- ies also report an increased risk of lung cancer in never-smokers [14,15]. Biomarkers of air pollution exposure and genetic effects have been used in some epidemiological studies to link ambient exposure with lung can- cer occurrence [16].
IARC recently reviewed the evidence summarized above and concluded that long-term exposure to outdoor air pollution, and specifi- cally particulate matter in outdoor air, causes lung cancer. Exposure to outdoor air pollution is also associ- ated with an increased risk of blad- der cancer [17]. Exposure to ambient PM2.5 has recently been estimated to have contributed to 3.2 million prema- ture deaths worldwide in 2010, due largely to cardiovascular disease, and 223 000 deaths from lung cancer. More than half of the lung cancer deaths attributable to ambient PM2.5 occurred in China and other countries in East Asia [18].
Comprehensive air quality man- agement programmes, using regu- latory approaches and adoption of cleaner technologies for transpor- tation, power generation, industrial production, and waste management have led to improvements in popula- tion health [19,20].
Air pollution by chlorofluoro- carbons is believed to be indirectly responsible for increases in skin cancer rates around the globe [21]. These chemicals, including carbon tetrachloride and methyl chloroform and other halons, are emitted from home air conditioners, foam cush- ions, and many other products. Chlorofluorocarbons are carried by winds into the stratosphere, where the action of strong solar radiation releases chlorine and bromine atoms that react with, and thereby eliminate, molecules of ozone. Depletion of the ozone layer is believed to be respon- sible for global increases in ultraviolet B (UVB) radiation (see Chapter 2.8).
Indoor air pollution Household air pollution Despite declines over the past 30 years in the percentage of the global population that burn solid fuels, either coal or biomass, for household cooking or heating, the number of people exposed has remained largely unchanged at 2.8 billion, due to population growth [22]. The proportion of the
population estimated by WHO to have been cooking with solid fuels in 2010 is highest in sub-Saharan Africa and South and East Asia (Fig. 2.9.4). Indoor burning of solid fuels in un- ventilated homes results in exposures to levels of fine particles (PM2.5) and other health-damaging air pollutants, such as carbon monoxide, that are many times ambient levels, and the components of this indoor smoke include known human carcinogens such as polycyclic aromatic hydrocar- bons, formaldehyde, and benzene. Household combustion of solid fuels also contributes to outdoor air pollu- tion in settings where solid fuels are widely used [18].
Exposure to household air pollu- tion from solid fuel use has recently been estimated to have contributed to 3.5 million premature deaths worldwide in 2010, due largely to cardiovascular disease and acute and chronic respiratory disease, and 126 000 deaths from lung cancer due to household coal burning [18]. Indoor burning of coal causes lung cancer in humans, and indoor burning of bio- mass may as well [23]. Exposure to emissions from indoor coal burning is
Fig. 2.9.5. On a cold day, a mother in Gatlang, a village in the Rasuwa district, Nepal, keeps her baby warm in front of a traditional wood stove. Indoor burning of solid fuels in unventilated homes results in exposures to levels of fine particles and other health- damaging air pollutants such as carbon monoxide.
Chapter 2.9 • Pollution of air, water, and soil 157
2 E
T IO
LO G
Y C
H A
P T
E R
2 .9
Pesticides and cancer Laura E. Beane Freeman
Pesticides encompass a large and diverse number of chemicals designed to kill pests, including weeds, insects, rodents, algae, and moulds, for agricultural, residential, and public health purposes. These chemicals make important contribu- tions to production and protection of agricultural commodities and control of insect disease vectors. They also present potential hazards to human health.
Unlike many other chemical agents, pesticides are designed for release into the environment, and exposure can occur occupationally, through environmental bystander exposure, and through ingestion of foods containing pesticides or pesticide residues. In 2007, the lat- est year for which estimates are available, 5.2 billion pounds (2.4 million tonnes) of pesticide active ingredient were applied in the world (http://w w w.epa.gov/opp00001/ pestsales/07pestsales/usage2007. htm#3_1). Despite widespread po- tential exposure, cancer risks as- sociated with long-term exposure to specific pesticides are generally not well characterized. Only one group of pesticides, inorganic arsenic com- pounds, and one pesticide contami- nant, the dioxin TCDD, are classified by IARC as Group 1, carcinogenic to humans. The fungicide captafol and the fumigant ethylene dibromide are classified as Group 2A, probably car- cinogenic to humans, as is occupa- tional exposure in the application of non-arsenical insecticides. However, there are several pesticides that are listed as Group 3, not classifiable as
to their carcinogenicity to humans, largely due to inadequate evidence in humans, and many more that have been linked with cancer that require further investigation (http://mono graphs.iarc.fr/ENG/Classification/ ClassificationsAlphaOrder.pdf).
A few areas where the evidence seems to be the strongest at this time are highlighted. Several or- ganochlorine and organophosphate insecticides have been linked with an increased risk of prostate cancer. This is notable since farmers, an oc- cupationally exposed group, have long been reported to have increased rates of prostate cancer [1]. Farmers also experience increased rates of lymphoma, and several reports have linked multiple pesticides to increased risk at this site, although confirmation is required [1]. Finally, a growing body of evidence suggests that parental exposure to pesticides, particularly of the mother during pregnancy, may increase the risk of leukaemia in chil- dren, although data on specific chemi- cals are largely lacking [2].
Developing accurate assessment of pesticide exposure is a major chal- lenge in epidemiological studies but is essential for identifying hazards associated with these chemicals. Many studies of pesticides and can- cer have focused on occupational exposures, where levels are likely to be highest, but it is also important for future research to assess effects of environmental exposures because of the widespread use of these chemi- cals. Future studies need to evaluate specific chemicals and to consider potential mechanisms of action to
support the biological plausibility of the epidemiological observations.
References
1. Blair A et al. (2009). J Agromedicine, 14:125– 131. http://dx.doi.org/10.1080/10599240902 779436 PMID:19437268
2. Van Maele-Fabry G et al. (2010). Cancer Causes Control, 21:787–809. http:// dx.doi.org/10.1007/s10552-010-9516-7 PMID:20467891
Websites Agents Classified by the IARC Monographs: h t t p: //m o n o g r a p h s . i a r c .f r/ E N G / C l a s s i f i cation/ClassificationsAlphaOrder.pdf
World and U.S. Pesticide Amount Used: http://w w w.epa.gov/opp0 0 0 01/pestsales/ 07pestsales/usage2007.htm#3_1
B2.9.2. A crop duster spraying pesti- cide in California.
associated with a 2-fold increase in lung cancer risk [24]. Women are the most heavily exposed, and nonsmok- ing women in some parts of China ex- perience very high lung cancer rates due to indoor coal burning. Indoor ex- posure to cooking-oil emissions from high-temperature frying may cause lung cancer. Indoor burning of solid
fuels has also been associated with increased rates of cancer of the up- per airways.
Improved ventilation of coal com- bustion emissions and the use of al- ternative types of stoves with lower emissions have been shown to re- duce lung cancer rates in exposed women [24,25]. The use of cleaner
fuels, such as compressed natural gas, is associated with reduced rates of lung cancer and other adverse health outcomes relative to the use of coal and biomass.
Involuntary smoking Tobacco smoke is an impor- tant source of indoor air pollution
158
worldwide, and involuntary smok- ing (exposure to second-hand or “environmental” tobacco smoke) is carcinogenic to humans (see Chapter 2.2).
Water pollution Disinfection by-products Access to clean water is one of the basic requirements of human health. Water quality is influenced by climate and weather, the geology of the soil, land use patterns, and discharges from agriculture and industry. The greatest and most immediate con- cern relates to infectious disease. Microbiological contamination of wa- ter is controlled by disinfection meth- ods based on chlorine, chloramine, or ozone. As a result of the interac- tion of chlorine with organic chemi- cals already present, drinking-water typically contains chlorination by- products, some of which may present a carcinogenic risk. Chloroform and other trihalomethanes are among the by-products most commonly detected, but not necessarily the most threatening to health. Studies of bladder cancer and some other cancers have suggested increased risk associated with consumption of chlorinated drinking-water [26]. Given the large number of people exposed to chlorination by-products, however, even a small increase in risk would implicate a substantial number of cases. It is desirable to reduce such by-products without reducing the effectiveness of disin- fection procedures.
Arsenic Arsenic causes cancers of the skin, lung, bladder, and possibly other organs [27]. The main source of en- vironmental exposure to arsenic for the general population is contami- nated drinking-water. High exposure to arsenic from drinking-water is found around the world, including ar- eas of Bangladesh, India, Mongolia, China, Argentina, Chile, Mexico, and the USA. In areas of high arsenic content, with concentrations typi- cally above 100 µg/l, there is strong evidence of an increased risk of
cancers of the bladder, skin, and lung after consumption of water with high arsenic contamination [27]. The data for other cancers, such as those of the liver and kidney, are more limited but are consistent and suggestive of a systemic effect. Given the many difficulties of conducting studies in areas with arsenic concentrations below 100 µg/l, the risks at these markedly lower levels are not clear. However, given the demonstrated risk at higher levels, the possibility of increased risk of different types of cancer at low exposures is plausible.
Estimates of the cancer burden associated with arsenic in drinking- water have been made for some regions of the world. In northern Chile, where most of the population was exposed to very high levels of arsenic in drinking-water from 1955 to 1970, arsenic might still account for 7% of all deaths among those aged 30 years and older, given the long latency of the effect [27]. In Bangladesh, tens of millions of people have been exposed to el- evated levels of arsenic in water, and an estimated 6% of all deaths, including cancer deaths, are due to these exposures [28]. West Bengal, India, which has a similar exposure
profile, may experience a similar burden of disease and death [29]. About 14 million people in Latin America are exposed to arsenic in drinking-water above the limit es- tablished by several national and international agencies [30].
Other drinking-water contaminants Several other groups of pollutants of drinking-water have been investigat- ed as possible sources of cancer risk in humans. These include organic compounds such as chlorinated sol- vents and pesticides from industrial, commercial, and agricultural activi- ties, and chemicals leached from waste sites. Organic pollutants that persist in the environment and accu- mulate in fish, such as polychlorinat- ed dibenzo-para-dioxin, polychlori- nated biphenyls, and organochlorine pesticides, are of particular concern, as is contamination by nitrates and nitrites, radionuclides, hormonally active compounds, and asbestos. For most pollutants, the epidemio- logical studies are inconclusive; however, an increased risk of stom- ach cancer has been repeatedly reported in areas with high nitrate
Fig. 2.9.6. Water pollution is usually identified with contaminants in water for domestic and agricultural use, but marine water may also be at risk. This oil spill occurred in the Gulf of Mexico in 2010 after an explosion on Deepwater Horizon, an offshore oil drilling unit.
Chapter 2.9 • Pollution of air, water, and soil 159
2 E
T IO
LO G
Y C
H A
P T
E R
2 .9
levels in drinking-water, and there is evidence of excess thyroid cancer [31].
Soil pollution A variety of toxic agents, including heavy metals, solvents, and persis- tent organic pollutants such as di- oxins, contaminate soil, sometimes at high concentrations [32]. In local regions, such pollution occurs by
substances produced as waste or as other consequences of a particu- lar mining or industrial process. The ranges of concentrations detected are specified for particular com- pounds in the corresponding IARC Monograph.
Soil contaminants present a car- cinogenic risk as a result of being vaporized and consequently inhaled, or being leached from the soil to con- taminate water supplies. Accordingly,
risk of cancer has generally been studied in relation to consequential air or water pollution, rather than with primary reference to particular levels of contaminants in soil. However, ac- cumulation of a variety of pollutants via the food-chain, and an estimate of any associated burden of disease, is being investigated through the WHO Foodborne Disease Burden Epidemiology Reference Group.
References
1. Boffetta P (2006). Human cancer from en- vironmental pollutants: the epidemiological evidence. Mutat Res, 608:157–162. http:// dx.doi.org/10.1016/j.mrgentox.2006.02.015 PMID:16843042
2. HEI Asbestos Literature Review Panel (1991). Asbestos in Public and Commercial Buildings: A Literature Review and Synthesis of Current Knowledge. Cambridge, MA: Health Effects Institute- Asbestos Research.
3. IARC (2012). Arsenic, metals, fibres, and dusts. IARC Monogr Eval Carcinog Risks Hum, 100C:1–499. PMID:23189751
4. WHO (2006). Air Quality Guidelines. Global Update 2005. Geneva: WHO. Available at http://www.euro.who.int/_ _data/assets/ pdf_file/0005/78638/E90038.pdf.
5. Samet JM, Cohen AJ (2006). Air pollu- tion. In: Schottenfeld D, Fraumeni JF, eds. Cancer Epidemiology and Prevention. New York: Oxford University Press, pp. 355–381.
6. IARC (2013). Diesel and gasoline engine exhausts and some nitroarenes. IARC Monogr Eval Carcinog Risks Hum, 105.
7. Brauer M, Amann M, Burnett RT et al. (2012). Exposure assessment for estimation of the global burden of dis- ease attributable to outdoor air pollu- tion. Environ Sci Technol, 46:652– 660. ht t p: //d x .d o i.o r g /10.10 21/e s 2 0 25752 PMID:22148428
8. HEI International Scientific Oversight Committee (2010). Outdoor Air Pollution and Health in the Developing Countries of Asia: A Comprehensive Review. HEI Special Report 18. Boston, MA: Health Effects Institute.
9. HEI Air Toxics Review Panel (2007). Mobile- Source Air Toxics: A Critical Review of the Literature on Exposure and Health Effects. HEI Special Report 16. Boston, MA: Health Effects Institute.
10. Claxton LD, Woodall GM Jr (2007). A review of the mutagenicity and rodent carcinogenic- ity of ambient air. Mutat Res, 636:36–94. http://dx.doi.org/10.1016/j.mrrev.2007.01.001 PMID:17451995
11. Lewtas J (2007). Air pollution combustion emissions: characterization of causative agents and mechanisms associated with cancer, reproductive, and cardiovascular effects. Mutat Res, 636:95–133. http:// dx.doi.org/10.1016/j.mrrev.2007.08.003 PMID:17951105
12. HEI Panel on the Health Effects of Traffic- Related Air Pollution (2010). Traffic- Related Air Pollution: A Critical Review of the Literature on Emissions, Exposure, and Health Effects. HEI Special Report 17. Boston, MA: Health Effects Institute.
13. Cao J, Yang C, Li J et al. (2011). Association between long-term exposure to outdoor air pollution and mortality in China: a cohort study. J Hazard Mater, 186:1594–1600. http:// dx.doi.org/10.1016/j.jhazmat.2010.12.036 PMID:21194838
14. Brunekreef B, Beelen R, Hoek G et al. (2009). Effects of Long-Term Exposure to Traffic- Related Air Pollution on Respiratory and Cardiovascular Mortality in the Netherlands: The NLCS-AIR Study. HEI Research Report 139. Boston, MA: Health Effects Institute.
15. Turner MC, Krewski D, Pope CA 3rd et al. (2011). Long-term ambient fine particu- late matter air pollution and lung cancer in a large cohort of never-smokers. Am J Respir Crit Care Med, 184:1374–1381. http:// dx.doi.org/10.1164/rccm.201106-1011OC PMID:21980033
16. Demetriou CA, Raaschou-Nielsen O, Loft S et al. (2012). Biomarkers of ambient air pol- lution and lung cancer: a systematic review. Occup Environ Med, 69:619–627. http:// dx.doi.org/10.1136/oemed-2011-100566 PMID:22773658
17. Loomis D, Grosse Y, Lauby-Secretan B et al. (2013). The carcinogenicity of outdoor air pollution. Lancet Oncol, 14:1262– 1263. http://dx.doi.org/10.1016/S1470-2045 (13)70487-X
18. Lim SS, Vos T, Flaxman AD et al. (2012). A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analy- sis for the Global Burden of Disease Study 2010. Lancet 380:2224–2260. http://dx.doi. org /10.1016/S0140 - 673 6(12) 6176 6 - 8 PMID:23245609
160
19. Pope CA 3rd, Ezzati M, Dockery DW (2009). Fine-particulate air pollution and life expec- tancy in the United States. N Engl J Med, 360:376–386. http://dx.doi.org/10.1056/ NEJMsa0805646 PMID:19164188
20. HEI Accountability Working Group (2003). Assessing Health Impact of Air Quality Regulations: Concepts and Methods for Accountability Research. HEI Communication 11. Boston, MA: Health Effects Institute. Available at http://pubs. healtheffects.org/view.php?id=153
21. IARC (2012). Radiation. IARC Monogr Eval Carcinog Risks Hum, 100D:1–437. PMID:23189752
22. Bonjour S, Adair-Rohani H, Wolf J et al. (2013). Solid fuel use for household cook- ing: country and regional estimates for 1980–2010. Environ Health Perspect, 121:784–790. http://dx.doi.org/10.1289/ehp. 1205987 PMID:23674502
23. Hosgood HD 3rd, Boffetta P, Greenland S et al. (2010). In-home coal and wood use and lung cancer risk: a pooled analysis of the International Lung Cancer Consortium. Environ Health Perspect, 118:1743–1747. http://dx.doi.org/10.1289/ehp.1002217 PMID:20846923
24. Hosgood HD 3rd, Wei H, Sapkota A et al. (2011). Household coal use and lung can- cer: systematic review and meta-analysis of case-control studies, with an emphasis on geographic variation. Int J Epidemiol, 40:719–728. http://dx.doi.org/10.1093/ije/ dyq259 PMID:21278196
25. Lan Q, Chapman RS, Schreinemachers DM et al. (2002). Household stove improvement and risk of lung cancer in Xuanwei, China. J Natl Cancer Inst, 94:826–835. http://dx.doi. org/10.1093/jnci/94.11.826 PMID:12048270
26. Richardson SD, Plewa MJ, Wagner ED et al. (2007). Occurrence, genotoxic- ity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: a review and roadmap for research. Mutat Res, 636:178–242. http:// dx.doi.org/10.1016/j.mrrev.2007.09.001 PMID:17980649
27. IARC (2004). Some drinking-water disinfect- ants and contaminants, including arsenic. IARC Monogr Eval Carcinog Risks Hum, 84:1–477. PMID:15645577
28. Flanagan SV, Johnston RB, Zheng Y (2012). Arsenic in tube well water in Bangladesh: health and economic impacts and implications for arsenic mitigation. Bull World Health Organ, 90:839–846. http://dx.doi.org/10.2471/BLT.11.101253 PMID:23226896
29. Chatterjee D, Halder D, Majumder S et al. (2010). Assessment of arsenic exposure from groundwater and rice in Bengal Delta Region, West Bengal, India. Water Res, 44:5803–5812. http://dx.doi.org/10.1016/j. watres.2010.04.007 PMID:20638702
30. Bundschuh J, Litter MI, Parvez F et al. (2012). One century of arsenic expo- sure in Latin America: a review of history and occurrence from 14 countries. Sci Total Environ, 429:2–35. http://dx.doi. org/10.1016/j.scitotenv.2011.06.024 PMID: 21959248
31. Ritter L, Solomon K, Sibley P et al. (2002). Sources, pathways, and relative risks of con- taminants in surface water and groundwater: a perspective prepared for the Walkerton inquiry. J Toxicol Environ Health A, 65:1–142. http:// dx.doi.org/10.1080/152873902753338572 PMID:11809004
32. Porta D, Milani S, Lassarino A et al. (2009). Systematic review of epidemiological studies on health effects associated with manage- ment of solid waste. Environ Health, 8:60. http://dx.doi.org/10.1186/1476-069X-8-60 PMID:20030820
Chapter 2.10 • Pharmaceutical drugs 161
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
0
Summary • Pharmaceutical drugs may have
the potential to induce or prevent cancer development.
• Antineoplastic agents used in cancer therapy can induce sec- ond cancers in patients who are apparently cured; this is most readily attributable to the geno- toxicity of these agents.
• Apar t from antineoplastic agents, drugs carcinogenic to humans include immunosup- pressants, hormonal agents, and phenacetin.
• Some drugs are implicated, but not established, as carcinogenic to humans by limited epidemio- logical data or by data restrict- ed to the results of bioassay in animals.
• Monitoring of possible carcino- genic risks associated with the use of pharmaceutical drugs is imperative.
• A few drugs have been ap- proved for cancer preventive therapy, and several others are being evaluated as preventive agents, including aromatase inhibitors, aspirin, statins, and metformin.
2.10 2 ETIOLOGY
Pharmaceutical drugs Søren Friis Jørgen H. Olsen
Margaret R. Karagas (reviewer)
Pharmaceutical drugs are chemi- cals that are developed and applied in medicine and dentistry for their ability to treat, prevent, or mitigate disease. Pharmaceutical drugs are invariably administered at levels that are biologically active, and thus may have side-effects in addition to their intended effects. A possible side-ef- fect is modification of the likelihood of malignant transformation, which may involve either a risk-enhancing or a risk-reducing effect.
Hormonal agents are well estab- lished as drugs with carcinogenic potential. These agents include oral contraceptives, hormone replace- ment therapy, and anti-estrogens, such as tamoxifen. Other carcino- genic drugs include antineoplastic agents that – particularly among those that are genotoxic – cause second cancers. For many such agents the benefit–risk ratio, de- termined by reference to patients achieving increased survival or re- lated outcome compared with those developing second cancers, is high and the continued use of these drugs is endorsed, although with intensive research to reduce dosage or de- velop safer and more effective re- placement products. In contrast, a decision to prescribe and use drugs that are established, or considered likely, to be carcinogenic is more
problematic in circumstances involv- ing non-life-threatening conditions.
In the most recent review involv- ing pharmaceuticals [1], the IARC Monographs Programme re-evaluat- ed 23 agents or combination thera- pies categorized as carcinogenic to humans (Group 1). These predomi- nantly include antineoplastic agents as well as a diverse range of drugs not used as cancer therapy.
Only a few drugs have been ap- proved for cancer preventive thera- py, including tamoxifen for women at risk of contralateral breast cancer.
Antineoplastic agents Antineoplastic agents include mul- tiple drugs that can induce second cancers in cancer patients who are apparently cured. Twelve such agents or combination therapies have been classified as carcinogenic to humans (Table 2.10.1) [1,2]. Of these, the ma- jority, including busulfan, chlorambu- cil, cyclophosphamide, melphalan, semustine (methyl-CCNU), thiotepa, and treosulfan, are alkylating agents that exhibit genotoxic activity, com- monly through alkylation of purine bases in DNA. Typically these agents cause acute myeloid leukaemia that often exhibits clonal loss of either chromosome 5 or 7, as distinct from acute myeloid leukaemia induced by topoisomerase II inhibitors, such as etoposide. Several such agents
162
Table 2.10.1. Antineoplastic drugs and other drugs evaluated by the IARC Monographs Programme
Group 1 agent Cancer on which sufficient evidence in humans is based Established mechanistic events
Antineoplastic agents Busulfan Acute myeloid leukaemia Genotoxicity (alkylating agent) Chlorambucil Acute myeloid leukaemia Genotoxicity (alkylating agent) Chlornaphazine Bladder Genotoxicity (alkylating agent, metabolism
to 2-naphthylamine derivatives) Cyclophosphamide Acute myeloid leukaemia, bladder Genotoxicity (metabolism to alkylating
agents) Etoposide (Group 2A in 2000) – Genotoxicity, translocations involving
MLL gene Etoposide in combination with cisplatin and bleomycin
Acute myeloid leukaemia Genotoxicity, translocations involving MLL gene (etoposide)
Melphalan Acute myeloid leukaemia Genotoxicity (alkylating agent) MOPPa combined chemotherapy Acute myeloid leukaemia, lung Genotoxicity Semustine (methyl-CCNU) Acute myeloid leukaemia Genotoxicity (alkylating agent) Thiotepa Leukaemia Genotoxicity (alkylating agent) Treosulfan Acute myeloid leukaemia Genotoxicity (alkylating agent)
Immunosuppressive agents Azathioprine Non-Hodgkin lymphoma, skin Genotoxicity, immunosuppression Ciclosporin Non-Hodgkin lymphoma, skin, multiple
other sites Immunosuppression
Other carcinogenic agents Analgesic mixtures containing phenacetin Renal pelvis, ureter (See phenacetin) Aristolochic acid (Group 2A in 2002) – Genotoxicity, DNA adducts in animals
are the same as those found in humans exposed to plants, A:T → T:A transversions in TP53, RAS activation
Methoxsalen plus ultraviolet A radiation Skin Genotoxicity after photo-activation Phenacetin (Group 2A in 1987) Renal pelvis, ureter Genotoxicity, cell proliferation Plants containing aristolochic acid Renal pelvis, ureter Genotoxicity, DNA adducts in humans,
A:T → T:A transversions in TP53 in human tumours
a Chlormethine (mechlorethamine), vincristine (Oncovin), procarbazine, and prednisone.
have since been supplemented or superseded by newer drugs [3]. It is therefore imperative that these new drugs and combinations thereof are monitored for possible carcino- genic effects. Typically, the antineo- plastic mechanisms of the replace- ment drugs are similar to those of the traditional drugs, and much in- sight can be gained from focusing on relevant mechanisms rather than the specific drugs. Since individual susceptibility is often a factor that emerges from relevant mechanistic analysis, this knowledge contributes to improved identification of patients who may be susceptible to drug- induced cancer [1].
Immunosuppressive agents Immunosuppressive agents, such as azathioprine and ciclosporin (also known as cyclosporin, cyclosporine,
or cyclosporin A), are carcinogenic to humans because of immuno- suppressive rather than genotoxic effects. Use of these agents may cause skin cancer and non-Hodgkin lymphoma; relevant evidence comes primarily from studies of patients who have undergone organ transplanta- tion (Table 2.10.1) [1].
Hormone therapy Female sex hormones, including estrogen-only and estrogen–pro- gestogen agents, are used widely to achieve various outcomes, including contraception and the amelioration of menopausal symptoms, uterine bleeding, and endometriosis [1]. In addition, these agents have been used widely for prevention of osteo- porosis and coronary heart disease among postmenopausal women, although the latter indication was
Fig. 2.10.1. A middle-aged woman sticks a hormone replacement therapy patch to her thigh. Hormonal agents are well estab- lished as drugs with carcinogenic potential.
Chapter 2.10 • Pharmaceutical drugs 163
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
0
Table 2.10.2. Hormonal agents assessed by the IARC Monographs Programme
Group 1 agent Cancer on which sufficient evidence in humans is based
Sites where cancer risk is reduced
Established mechanistic events
Other likely mechanistic events
Combined estrogen– progestogen meno- pausal therapy
Endometrium (risk decreases with number of days per month of progestogen use), breast
– Receptor-mediated events
Estrogen genotoxicity
Combined estrogen– progestogen oral contraceptives
Breast, cervix, liver Endometrium, ovary Receptor-mediated events
Estrogen genotoxicity, hormone-stimulated expression of human papillomavirus genes
Diethylstilbestrol Breast (exposure during pregnancy), vagina and cervix (exposure in utero) Limited evidence: testis (exposure in utero), endometrium
– Estrogen-receptor- mediated events (vagina, cervix), genotoxicity
Epigenetic programming
Estrogen-only meno- pausal therapy
Endometrium, ovary Limited evidence: breast
– Estrogen-receptor- mediated events
Genotoxicity
Tamoxifen Endometrium Breast Estrogen-receptor- mediated events, genotoxicity
–
abandoned in 2003 after publication of the Women’s Health Initiative trial, which could not replicate the risk re- duction for coronary heart disease reported in many observational stud- ies [4]. There is consistent evidence that female sex hormones may influ- ence the risk of cancers of the fe- male reproductive organs and breast (Table 2.10.2) [1]. The mechanisms of carcinogenesis of female sex hor- mones include estrogen-receptor- mediated responses and potentially direct genotoxic effects of the estro- genic hormones or their associated by-products (Table 2.10.2) [1].
Estrogen-only menopausal therapy Estrogen-only menopausal therapy refers to the use of estrogen without progestogen by perimenopausal and postmenopausal women, primar- ily for the treatment of menopausal symptoms [1]. After a substantial increase in the use of this regimen in the 1960s and early 1970s, the use declined after reports of a strong association between estrogen-only therapy and risk of endometrial cancer. Since then, estrogen-only menopausal agents have been pre- scribed predominantly to hysterec- tomized women. In the recent re- evaluation [1], IARC concluded that
estrogen-only menopausal therapy causes cancers of the endometrium and ovary (Table 2.10.2). Moreover, an increased risk of breast cancer was also noted.
Combined estrogen– progestogen menopausal therapy Combined estrogen–progestogen menopausal therapy was initially developed to avoid the excess oc- currence of endometrial cancer as- sociated with estrogen-only therapy [1]. Estrogen–progestogen agents were used heavily in the 1990s and the beginning of the 2000s. However, in 2003 the Women’s Health Initiative trial reported that use of combined es- trogen–progestogen agents among women aged 50–79 years at baseline was associated with a hazard ratio for coronary heart disease of 1.24 (95% confidence interval, 1.00–1.54) after a mean follow-up of 5.2 years [4]. This study has affected the prescribing of combined estrogen– progestogen menopausal therapy substantially, resulting in rapidly de- clining use of these agents, restricted indications, and reduced duration of therapy. Based on a comprehen- sive amount of evidence, IARC has concluded that long-term combined estrogen–progestogen menopausal
therapy causes cancers of the breast and endometrium (Table 2.10.2) [1]. The increased risk of endometrial cancer decreases with the number of days per month that progestogens are added to the regimen. IARC has also noted an inverse association be- tween the use of these agents and risk of colorectal cancer; however, this evidence is not conclusive. For several other sites, including cancers of the thyroid, lung, stomach, liver, urinary tract, pancreas, ovary, and cervix, as well as lymphoma, leu- kaemia, cutaneous melanoma, and central nervous system tumours, the current evidence does not point to an association with combined estrogen– progestogen menopausal therapy [1].
Fig. 2.10.2. Oral contraceptives are avail- able in estrogen–progestogen combina- tions as well as in progestogen-only pills.
164
Combined estrogen– progestogen contraceptives Numerous types of combined hor- monal contraceptives have been mar- keted since these agents first became available in the late 1950s [1]. With reference to the IARC Monographs Programme evaluation, there is sufficient evidence that combined estrogen–progestogen oral contra- ceptives cause cancers of the breast (notably among young women),
cervix, and liver (Table 2.10.2) [1]. IARC has also concluded that these agents are protective against can- cers of the endometrium and ovary (Table 2.10.2), and possibly also colo- rectal cancer, although the evidence is not conclusive for colorectal cancer [1]. For other sites, including cancers of the thyroid, lung, stomach, urinary tract, gall bladder, and pancreas, as well as lymphoma, cutaneous mela- noma, and central nervous system
tumours, the available literature does not support an association with com- bined estrogen–progestogen contra- ceptives [1].
Diethylstilbestrol This synthetic estrogen was first man- ufactured in 1938 [1]. Diethylstilbestrol was widely used, especially from the 1940s to the 1970s, to prevent po- tential miscarriages by stimulating
Breast implants and cancer Eric Lavigne
The demand for breast implants among women has been rapidly increasing since the first mammary prosthesis was inserted into a wom- an’s breast in the early 1960s [1]. Although breast implants are very popular for cosmetic purposes, con- siderable controversy remains about their long-term effects on cancer incidence. Concerns were initially raised about the possible carcino- genic effect of cosmetic breast im- plants, specifically for breast cancer [2]. However, the weight of epide- miological evidence based on large cohort studies with long follow-up periods showed no carcinogenic effect of breast augmentation on breast cancer [3–5].
Concerns about the carcino- genic effect of breast implants have also been characterized according to specific implant characteristics, such as the type of implant (saline- filled or silicone-gel-filled), implant placement (submuscular or subglan- dular), and implant envelope (with or without polyurethane coating). For the implant envelope, available labo- ratory studies show that 2,4-toluene- diamine, a biodegradation prod- uct of polyurethane, may promote the progression of already present mutated cells [6,7]. In addition, one epidemiological study showed an el- evation in breast cancer incidence for women with polyurethane-coated subglandular implants compared with women with subglandular im- plants without polyurethane coating [3]. However, this finding is based on
only one study with a small number of cases, and further investigations are required before this interpreta- tion can be used in clinical practice. No differences in breast cancer inci- dence have been reported between saline-filled and silicone-gel-filled implants [3,8]. In addition, one study found that women with subglandular implants had a reduced risk of breast cancer compared with women with submuscular implants [3].
Some studies have evaluated the relationship between breast implants and the incidence of can- cers at sites other than the breast. No conclusive findings can be re- ported about risk of other types of cancers [3,5]. Although there is con- cern about a possible link between breast implants and anaplastic large T-cell lymphomas of the breast [9], no epidemiological cohort studies have reported any evidence for an association [3,5,10].
Another concern is that breast implants may obscure the visuali- zation of breast tissue with mam- mography, which may impair the ability to identify breast cancer at an early stage, for which survival is generally more favourable. In fact, recent epidemiological evidence has suggested that breast cancers are diagnosed at more advanced stages among women with cosmet- ic breast implants and that breast cancer-specific survival is lower in these women [11]. However, further studies are warranted about breast cancer diagnosis and
prognosis among women with breast augmentation.
References 1. Love S, Lindsey K (2000). Variations in
development. Dr. Susan Love’s Breast Book, 3rd ed. Cambridge, MA: Perseus Publishing.
2. Deapen DM et al. (1986). Plast Reconstr Surg, 77:361–368. http://dx.doi.org/10. 10 97/ 0 0 0 0 6 5 3 4 -19 8 6 0 3 0 0 0 - 0 0 0 01 PMID:3952193
3. Pan SY et al. (2012). Int J Cancer, 131: E1148–E1157. http://dx.doi.org/10.1002/ ijc.27603 PMID:22514048
4. Deapen DM, Brody GS (2012). Plast Reconstr Surg, 129:575e–576e. http:// dx.doi.org/10.1097/PRS.0b013e31824199f1 PMID:22374025
5. Lipworth L et al. (2009). Int J Cancer, 124:490–493. http://dx.doi.org/10.1002/ ijc.23932 PMID:19003966
6. Cunningham ML et al. (1991). Toxicol Appl Pharmacol, 107:562–567. http://dx.doi. org /10.1016/0 0 41- 0 0 8X(91) 9 0319 - A PMID:2000642
7. Shanmugam K et al. (2001). Anal Sci, 17:1369–1374. http://dx.doi.org/10.2116/ analsci.17.1369 PMID:11783783
8. Friis S et al. (2006). Int J Cancer, 118:998– 1003. http://dx.doi.org/10.1002/ijc.21433 PMID:16152592
9. Jewell M et al. (2011). Plast Reconstr Surg, 128:651–661. http://dx.doi.org/10.1097/ PRS.0b013e318221db81 PMID:21865998
10. Lipworth L et al. (2009). Plast Reconstr Surg, 123:790–793. http://dx.doi.org/10. 1097/PRS.0b013e318199edeb PMID: 19319041
11. Lavigne E et al. (2012). Cancer Epidemiol Biomarkers Prev, 21:1868–1876. http:// dx.doi.org/10.1158/1055-9965.EPI-12- 0484 PMID:22850806
Chapter 2.10 • Pharmaceutical drugs 165
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
0
the synthesis of estrogen and pro- gesterone in the placenta. In the USA alone, an estimated 5–10 million peo- ple received diethylstilbestrol during pregnancy or were exposed to the drug during fetal life. Subsequently, several adverse effects emerged (Table 2.10.2) [1]. First, slightly in- creased risks of breast cancer and possibly of endometrial cancer were observed among women exposed to diethylstilbestrol during pregnancy. Second, among female children of women who used diethylstilbestrol during pregnancy, a markedly in- creased risk of adenocarcinoma of the vagina or cervix was recorded, typically occurring during adoles- cence. Third, fetal exposure to di- ethylstilbestrol was associated with increased risks of squamous cell car- cinoma of the cervix among female offspring and testicular cancer among male offspring. Currently, evidence is accumulating of an increased risk of cancers of the female reproductive organs among third-generation off- spring of women who used diethyl- stilbestrol [1].
Tamoxifen Tamoxifen is indicated as adjuvant therapy for treatment of postmen- opausal estrogen-receptor-positive or progesterone-receptor-positive breast cancer among both women and men, as well as ductal carci- noma in situ after breast surgery and radiotherapy among women [1]. Furthermore, tamoxifen has been approved as a breast cancer preventive agent among women at high risk of developing breast cancer (see below). Observational epide- miological studies and randomized trials have consistently shown that use of tamoxifen increases the risk of endometrial cancer whether given as adjuvant therapy among women with breast cancer or as preventive therapy among women at high risk of breast cancer (Table 2.10.2) [1]. There is also some indication that tamoxifen may be associated with an increased risk of some types of gas- trointestinal cancer; however, these results are not conclusive.
Phenacetin The analgesic phenacetin and mix- tures containing phenacetin were until recently classified by IARC as probably carcinogenic to humans (Group 2A), since the available evidence had not been able to dis- criminate between potential carci- nogenic effects of phenacetin and those attributable to other analge- sics or components of analgesics used by the same patients. In the recent IARC re-evaluation [1], how- ever, phenacetin and phenacetin- containing mixtures were both clas- sified as carcinogenic to humans (Group 1), causing cancers of the renal pelvis and ureter (Table 2.10.1). Since phenacetin was withdrawn from the market in most countries in about 1980, the pool of patients with excess risk of cancers of the renal pelvis or ureter due to phenacetin therapy is probably limited now, although a long latency period for development of these cancer types cannot be excluded and relevant cancers may yet be diagnosed.
Miscellaneous agents Use of Aristolochia plants contain- ing aristolochic acid has mainly
been limited to traditional Chinese drugs, weight-loss pills in Belgium, and cereals from Balkan fields [1]. Aristolochic acid induces DNA dam- age similar to that seen for antineo- plastic agents and is associated with nephropathy and an increased risk of cancers of the renal pelvis and ureter (Table 2.10.1). Genome-wide sequencing has established the mutational signature of aristolochic acid [5].
Methoxsalen, which is a pso- ralen produced naturally by vari- ous plants, is a photosensitizer: the agent increases skin reactivity to ul- traviolet radiation [1]. Methoxsalen is nowadays typically taken orally in conjunction with ultraviolet radiation as phototherapy for various condi- tions, notably psoriasis. The drug has been suspected to cause vari- ous types of skin cancer, but there is convincing evidence only for causa- tion of squamous cell carcinoma [1]. Some common vegetables naturally contain methoxsalen, and there have been reports of photosensitivity and skin lesions among grocery or farm workers.
Photosensitivity is also induced by frequently used drugs such as certain types of antihypertensives,
Fig. 2.10.3. A patient receives chemotherapy through a port located on his chest.
166
for example hydrochlorothiazide and nifedipine. There is increasing evi- dence that these drugs increase the risk of lip and skin cancers [6].
Many of the drugs evaluated by IARC have been categorized as probably carcinogenic to humans (Group 2A) or possibly carcinogenic to humans (Group 2B) because the epidemiological evidence has not been definitive and is hence cat- egorized as limited or because car- cinogenicity has been demonstrated only in experimental animals. These agents include, for example, gris- eofulvin, metronidazole, phenytoin, chloramphenicol, iron-dextran com- plexes, phenoxybenzamine, pheno- barbital, oxazepam, propylthiouracil, and anti-retroviral agents (e.g. zido- vudine) [7].
Cancer preventive drugs Hormone antagonists Several drugs within the therapeutic groups of selective estrogen-receptor modulators and aromatase inhibi- tors have been approved for endo- crine therapy among women with hormone-receptor-positive breast cancer. However, so far, only the two selective estrogen-receptor modula- tors tamoxifen and raloxifene have been approved for preventive therapy against breast cancer [8]. Tamoxifen reduces the risk of contralateral breast cancer and, when given as
preventive therapy, reduces the risk of estrogen-receptor-positive breast cancers. Tamoxifen has greater effi- cacy than raloxifene and can be used in premenopausal women, but ralox- ifene has fewer side-effects and is not classified as carcinogenic to hu- mans. Newer drugs within the class of selective estrogen-receptor modula- tors, for example lasofoxifene, also show efficacy, and possibly a better overall benefit–risk profile, but need further assessment. Aromatase in- hibitors may be more efficacious, and drugs within this class, for example anastrozole and exemestane, are currently being evaluated as preven- tive agents among women with a his- tory of carcinoma in situ of the breast or other women considered to be at high risk of developing breast cancer.
Aspirin and other non- steroidal anti-inflammatory drugs Several observational epidemio- logical studies, randomized trials of colon adenoma recurrence, and randomized trials in patients with hereditary colorectal cancer syn- dromes have consistently shown that aspirin and other non-steroidal anti-inflammatory drugs (NSAIDs) reduce the risk of colorectal neo- plasia and cancer [9,10]. A cancer preventive effect of aspirin or non- aspirin NSAIDs against other cancer types is less certain [9,10]. Despite the cancer preventive effect of non- aspirin NSAIDs, the use of these agents for prevention of colorectal cancer is hampered by their recently established association with cardio- vascular events [10]. The evaluation of NSAIDs as a potential cancer pre- ventive measure is thus focused on aspirin [10]. However, although the preventive effect of aspirin against colorectal neoplasia has been practi- cally proven, aspirin cannot be rec- ommended for cancer prevention in the general population of healthy individuals due to the risk of serious adverse events, notably upper gas- trointestinal bleeding [10]. Moreover, uncertainty still remains as to the optimal dose and duration of aspi- rin therapy for cancer prevention
[9]. The results of several current randomized trials will hopefully provide more confirmatory data on these issues.
Statins HMG-CoA reductase inhibitors (statins) reduce serum levels of cholesterol and are widely used to manage and prevent cardiovascu- lar and coronary heart disease. In addition, a growing body of experi- mental evidence has suggested that statins have cancer preventive prop- erties [11]. To date, however, sev- eral randomized trials with cancer as secondary end-point have failed to demonstrate a reduced risk of cancer, overall or at specific sites, within statin treatment periods of 5–10 years. Likewise, observational epidemiological studies with longer follow-up have generally produced null results, with the exception of an inverse association between statin use and aggressive prostate cancer [12]. Statins have also been evalu- ated as adjuvant therapy to standard cancer therapy; however, to date this has yielded primarily null results [11]. Several current trials will hope- fully clarify whether statins have a future role in cancer prevention or therapy.
Recent research issues There is substantial evidence that fe- male sex hormones are associated with a decreased risk of colorectal cancer and thus represent further hormonal agents associated with both carcinogenic and cancer pre- ventive properties. Meta-analysis indicates significant duration-de- pendent reductions in ovarian cancer incidence associated with oral con- traceptive use [13]. Observational epidemiological studies also indicate that metformin, a widely used oral an- tidiabetic drug, may reduce the risk of several cancer types; however, these associations require further assess- ment [14]. Several trials of metformin have recently been launched to as- sess whether this old drug may have a cancer preventive potential.
Fig. 2.10.4. Today, the evaluation of non- steroidal anti-inflammatory drugs as a potential cancer preventive measure is focused on aspirin.
Chapter 2.10 • Pharmaceutical drugs 167
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
0
Some epidemiological studies in- dicate that insulin treatment is asso- ciated with increased risk of several cancer types; however, the results are equivocal [15]. Moreover, it has been suggested that insulin ana- logues are more prone to induce can- cer, but this hypothesis also remains to be clarified. Several international research initiatives have recently been launched to clarify these issues. A recent single study indicated that
long-term current use of calcium- channel blockers is associated with increased breast cancer risk among women aged 55–74 years [16].
Two large randomized trials have demonstrated that the two 5α-reductase inhibitors, finasteride and dutasteride, indicated for treat- ment of benign prostatic hyperplasia, reduce the risk of prostate cancer among men by 25%; however, both studies also revealed an increased risk of aggressive prostate cancer [17]. Whether the increased risk of aggressive prostate cancer is due to detection bias or a true biologi- cal association remains unclear [17]. Until this spurious phenomenon is resolved, these agents should not be recommended as cancer preventive agents.
Concerns have been raised about the new biological response modifiers for treatment of rheumatoid arthri- tis, which include tumour necrosis factor inhibitors. Tumour necrosis factor inhibitors have been sug- gested to increase the risk of cer- tain cancers, including lymphoma in children and adolescents, and these agents currently carry a label warning against a possible risk of malignancy. A large meta-analysis
of 61 randomized trials of tumour necrosis factor inhibitors, including a total of 29 423 patients, concluded that use of these drugs was not as- sociated with an increased risk of cancer overall or at specific sites [18]. However, follow-up was short (24–156 weeks) in this analysis, and continued monitoring is warranted.
The carcinogenicity of some drugs was recently evaluated for Volume 108 of the IARC Monographs [19], and two drugs attracted particular attention. Pioglitazone, used for treatment of type 2 diabe- tes, was classified as probably car- cinogenic to humans (Group 2A), with limited evidence that it causes bladder cancer in humans and evi- dent carcinogenicity in experimen- tal animals. Use of digoxin, which is widely prescribed for chronic heart failure and related conditions, is as- sociated with an increased risk of breast cancer; causal interpretation of these data is precluded because of possible confounding, particularly in relation to obesity and alcohol consumption. In the absence of ani- mal bioassay data, digoxin was cat- egorized as possibly carcinogenic to humans (Group 2B).
Fig. 2.10.5. A patient with diabetes uses an “insulin pen” to inject insulin. Drugs that increase circulating insulin, including ex- ogenous insulin, insulin analogues, and insulin secretagogues, may be associated with an increased cancer risk [15].
168
References 1. IARC (2012). Pharmaceuticals. IARC
Monogr Eval Carcinog Risks Hum, 100A: 1–437. PMID:23189749
2. Grosse Y, Baan R, Straif K et al.; WHO IARC Monograph Working Group (2009). A review of human carcinogens – Part A: pharmaceu- ticals. Lancet Oncol, 10:13–14. http://dx.doi. org /10.1016/S1470 -20 45(0 8)7028 6 - 9 PMID:19115512
3. WHO Collaborating Centre for Drug Statistics Methodology (2012). Guidelines for ATC Classification and DDD Assignment. Oslo: WHO. Available at http://www.whocc. no/atc_ddd_index/.
4. Manson JE, Hsia J, Johnson KC et al.; Women’s Health Initiative Investigators (2003). Estrogen plus progestin and the risk of coronary heart disease. N Engl J Med, 349:523–534. http://dx.doi.org/10.1056/ NEJMoa030808 PMID:12904517
5. Hoang ML, Chen C-H, Sidorenko VS et al. (2013). Mutational signature of aristolochic acid exposure as revealed by whole-exome sequencing. Sci Transl Med, 5:ra102. http:// dx.doi.org/10.1126/scitranslmed.3006200 PMID:23926200
6. Friedman GD, Asgari MM, Warton EM et al. (2012). Antihypertensive drugs and lip cancer in non-Hispanic whites. Arch Intern Med, 172:1246–1251. http://dx.doi. org/10.1001/archinternmed.2012.2754 PMID:22869299
7. Friedman GD, Jiang SF, Udaltsova N et al. (2009). Epidemiologic evaluation of pharmaceuticals with limited evidence of carcinogenicity. Int J Cancer, 125:2173– 2178. http://dx.doi.org/10.1002/ijc.24545 PMID:19585498
8. Cuzick J, DeCensi A, Arun B et al. (2011). Preventive therapy for breast cancer: a consensus statement. Lancet Oncol, 12:496–503. http://dx.doi.org/10.1016/S14 70-2045(11)70030-4 PMID:21441069
9. Chan AT, Arber N, Burn J et al. (2012). Aspirin in the chemoprevention of colo- rectal neoplasia: an overview. Cancer Prev Res (Phila), 5:164–178. http://dx.doi. org/10.1158/1940 - 6207.CAPR-11- 0391 PMID:22084361
10. Umar A, Dunn BK, Greenwald P (2012). Future directions in cancer prevention. Nat Rev Cancer, 12:835–848. http://dx.doi. org/10.1038/nrc3397 PMID:23151603
11. Gazzerro P, Proto MC, Gangemi G et al. (2012). Pharmacological actions of statins: a critical appraisal in the management of cancer. Pharmacol Rev, 64:102–146. http://dx.doi.org/10.1124/pr.111.004994 PMID:22106090
12. Boudreau DM, Yu O, Johnson J (2010). Statin use and cancer risk: a com- prehensive review. Expert Opin Drug Saf, 9:603–621. http://dx.doi.org/10.15 17/14740331003662620 PMID:20377474
13. Havrilesky LJ, Moorman PG, Lowery WJ et al. (2013). Oral contraceptive pills as primary prevention for ovarian cancer: a systematic review and meta-analysis. Obstet Gynecol, 122:139–147. http://dx.doi. org/10.1097/AOG.0b013e318291c235 PMID:23743450
14. Kourelis TV, Siegel RD (2012). Metformin and cancer: new applications for an old drug. Med Oncol, 29:1314–1327. http:// dx.doi.org/10.1007/s12032- 011-9846 -7 PMID:21301998
15. Onitilo AA, Engel JM, Glurich I et al. (2012). Diabetes and cancer II: role of diabetes medications and influence of shared risk factors. Cancer Causes Control, 23:991–10 08. http://dx.doi.org/10.1007/s10552-012- 9971-4 PMID:22527174
16. Li CI, Daling JR, Tang MT et al. (2013). Use of antihypertensive medications and breast cancer risk among women aged 55 to 74 years. JAMA Intern Med, http://dx.doi. org/10.1001/jamainternmed.2013.9071 PMID:23921840
17. Azzouni F, Mohler J (2012). Role of 5α-reductase inhibitors in prostate can- cer prevention and treatment. Urology, 79:1197–1205. http://dx.doi.org/10.1016/j. urology.2012.01.024 PMID:22446342
18. Lopez-Olivo MA, Tayar JH, Martinez- Lopez JA et al. (2012). Risk of malignan- cies in patients with rheumatoid arthritis treated with biologic therapy: a meta- analysis. JAMA, 308:898–908. http:// dx.do i.o r g /10.10 01/ 2 012 .jama.10 8 57 PMID:22948700
19. Grosse Y, Loomis D, Lauby-Secretan B et al. (2013). Carcinogenicity of some drugs and herbal products. Lancet Oncol, 14:807–808. http://dx.doi.org/10.1016/S14 70-2045(13)70329-2
Chapter 2.11 • Naturally occurring chemical carcinogens 169
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
1
Summary • Natural products are chemicals
from plants, fungi, lichens, and bacteria, some of which have unique pharmacological effects. Humans are primarily exposed to many of these bioactive naturally occurring chemicals through food and water; natural products are used as pharma- ceuticals and herbal remedies.
• Natural products of bacteria that are possibly or probably carcino- genic to humans include antibi- otics, chemotherapeutic agents, and the water contaminant mi- crocystin-LR. Environmental ex- posure to cyanobacterial toxins is well documented.
• Mycotoxins are secondary me- tabolites of fungi that are known to cause sickness or death in humans or animals. Exposure to carcinogenic mycotoxins can be high in areas where a single commodity is consumed as a dietary staple.
• Several chemicals found in par- ticular plants used as food, food additives, or herbal medicines have been classified as human or animal carcinogens. Some of these plants are weeds and are sometimes harvested with crop
2.11 2 ETIOLOGY
Naturally occurring chemical carcinogens Ronald T. Riley Wentzel C.A. Gelderblom (reviewer)
J. David Miller (reviewer)
plants or are deliberately used as herbal medicines.
Information about possible, prob- able, and known carcinogens that are naturally occurring chemicals of microbial or plant origin is summa- rized in this chapter. Natural prod- ucts are chemicals found in nature, which often have unique pharmaco- logical effects. In the 2011 edition of the Dictionary of Natural Products, more than 230 000 natural products were described [1]. Cancer-causing chemicals produced by bacteria, fungi, and plants are discussed here. Compounds that are semi- synthetic, such as etoposide and teniposide from podophyllotoxin, or are produced by a genetically modified organism, as illustrated by doxorubicin obtained from a geneti- cally altered strain of Streptomyces, are not included. That said, these three agents do provide important insights into the potential of natu- rally occurring chemicals to be carcinogenic. “Natural products” from tobacco, betel quid, or areca nut (see Chapter 2.2), ethanol (see Chapter 2.3), or naturally occurring chemicals produced by anaerobic bacteria such as methylated arse- nic species, which are environmen- tal pollutants (see Chapter 2.9), are discussed elsewhere.
One of the striking attributes of nature is biodiversity. As might be expected, there are a large number of biologically active organic chemi- cals produced in nature by microbes and plants. Biologically produced naturally occurring organic com- pounds are structurally diverse. Not surprisingly, humans are exposed to many of these chemicals primar- ily through food and drinking-water. In addition, exposure to naturally occurring chemicals occurs because they are used as pharmaceuticals and herbal remedies useful in the treatment and/or prevention of diseases.
The discovery of naturally occur- ring chemicals with potentially useful or dangerous biological activity is still in progress and is often proprietary. The current edition of the Dictionary of Natural Products [1] is a compre- hensive survey of the literature on natural product discovery. For exam- ple, from 1995 to 1997 more than 6000 new compounds were added to the dictionary. Making informed decisions about the carcinogenic risk to humans posed by exposure to each of these 230 000 (and grow- ing) biologically active natural prod- ucts is a daunting, if not impossible, task. However, considering bacteria, including actinomycetes, and fungi, there are only 45 000 compounds with known biological activity [2].
170
The proportion of biologically active chemicals derived from plants that are used as therapeutic agents is quite low, despite some important examples [3].
Since 1971, the IARC Monographs Programme has evaluated more than 900 agents. More than 400 have been identified as carcinogen- ic, probably carcinogenic, or pos- sibly carcinogenic to humans. The information gained from assess- ments of representative carcinogenic
chemicals by IARC and the United States National Toxicology Program provides critical data for identifying which known or newly discovered chemicals may be carcinogenic hazards. For example, quantitative structure–activity relationships ob- tained from long-term feeding stud- ies can be used to develop predic- tive models. These can be used for screening and prioritizing the carci- nogenic hazards posed by naturally occurring chemicals when the data from animal studies are inadequate [4]. Accurate predictive modelling is dependent on the quality of the toxi- cology data rather than on the quan- tity of chemicals tested in long-term rodent carcinogenicity studies. Thus, it may be possible to predict which natural products are potential carci- nogenic hazards without conducting long-term feeding studies on newly discovered natural products. This approach can be used effectively only if the toxicology database from long-term feeding studies represents all the relevant structures that have a high probability of contributing to cancer induction in humans. This
task is also daunting, but not im- possible, as would be testing every naturally occurring biologically ac- tive chemical.
Naturally occurring carcinogens produced by bacteria Table 2.11.1 summarizes naturally occurring chemicals produced by bacteria evaluated by IARC as possibly or probably carcinogenic to humans. With the exception of microcystin-LR, human exposure to these agents occurs from treat- ment in a clinical setting. These six agents are produced by bacteria of the order Actinomycetales. The actinomycete genus Streptomyces produces a wide spectrum of natu- ral products [5]. Actinomycetes are commonly found in soil and water and contribute to the decomposition of organic material [6]. The charac- teristic smell of wet soil is due in part to geosmin (meaning “earth smell”), a secondary metabolite produced by Streptomyces. Streptomyces- produced secondary metabolites
Fig. 2.11.1. Streptomyces, or soil bacteria, are commonly found in soil and water and are necessary for the decomposition of organic material.
Table 2.11.1. Naturally occurring organic chemical carcinogens produced by bacteria
Agent (bacterial genera)a
Primary source of exposureb
Proposed mechanism of carcinogenicityc Primary target or tumour sitesd
IARC evaluation of carcinogenicity to humanse
Azacitidine (Streptoverticillium)
Chemotherapy Cytidine analogue/DNA methylation inhibitor Multiple sites Group 2A
Chloramphenicol (Streptomyces)
Antibiotic Inhibits mitochondrial protein synthesis → aplastic anaemia/bone-marrow depression
Blood Group 2A
Azaserine (Glycomyces)
Antibiotic, chemotherapy
Glutamine/purine antagonist → inhibition of purine biosynthesis and glutamine-dependent enzymes
Pancreas, kidney
Group 2B
Daunomycin (Streptomyces)
Antibiotic, chemotherapy
DNA intercalation → inhibition of macromolecu- lar biosynthesis → topoisomerase II inhibition
Mammary, kidney, local sarcomas
Group 2B
Microcystin-LR (Microcystis)
Water contaminant
Inhibition of protein phosphatases 1 and 2A → tumour promotion
Liver, colon Group 2B
Mitomycin C (Streptomyces)
Antibiotic, chemotherapy
DNA/RNA alkylating agent → DNA/RNA damage Dithiol cross-linking agent → protein damage
Multiple sites Group 2B
Streptozotocin (Streptomyces)
Antibiotic, chemotherapy
DNA alkylating agent → DNA damage Oxidative stress → diabetes
Liver, lung, kidney
Group 2B
a For all bacteria the most likely producing genera are given; the listing is not exhaustive. b All of the agents listed can be produced by biosynthesis in situ, but for the antibiotics and chemotherapeutic agents exposure is in a controlled clinical setting. c Where possible, the mechanism described is taken from the most recent IARC Monograph or IARC Scientific Publication. d Includes reference to animal bioassay data. For some studies the route of exposure is oral, but in some studies exposure was intravenous, subcutaneous, or other. e Group 1, carcinogenic to humans; Group 2A, probably carcinogenic to humans; Group 2B, possibly carcinogenic to humans; Group 3, not classifiable as to its carcinogenicity to humans; Group 4, probably not carcinogenic to humans. Only agents in Groups 1, 2A, and 2B are listed here. There are several naturally occurring agents classified by IARC as Group 3, but they are not included in this table.
Chapter 2.11 • Naturally occurring chemical carcinogens 171
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
1
may play important roles in bacte- rial–fungal, bacterial–plant, and bac- terial–animal interactions in nature [6]. For example, Streptomyces are endosymbionts in wasps, and in the process of building cocoons the lar- vae incorporate nine different antibi- otics on the surface of the cocoon for protection from microbial pathogens [7]. There are many other exam- ples of actinomycetales secondary metabolites playing roles in sym- biotic relationships with eukaryotic organisms [6]. Thus, environmental exposure to naturally produced an- tibiotics, including the six listed in Table 2.11.1, is possible, but the lev- els of human exposure would likely be quite low, given the nature of the environment where the producing organisms thrive (soil, compost, manure, fermenting hay).
In contrast, human environmental exposure to the cyanobacterial toxin microcystin-LR is well documented [8]. Exposure occurs most frequently through drinking-water or ingestion of water during recreational activities. Microcystin-LR contamination of di- etary supplements produced using cyanobacteria has been demonstrat- ed. In one documented case, expo- sure resulting from contamination of water used for renal dialysis has resulted in acute toxicity in humans [8]. Many countries have guidelines to minimize exposure from drinking- water and recreational exposure. In rodent studies, microcystin-LR was a promoter of preneoplastic lesions and persistent neoplastic nodules in the liver. Co-exposure to aflatoxin is a concern since in rats microcystin-LR promotes the formation of aflatoxin- induced glutathione S-transferase placental form-positive foci, which are considered preneoplastic le- sions [8]. The inhibition of protein phosphatases 1 and 2A in vivo is consistent with microcystin promotion of liver tumours in animals. Inhibition of phosphatases causes hyperphos- phorylation of intracellular proteins, disruption of intermediate filament assembly, altered expression of on- cogenes and early response genes, and expression of tumour necrosis
factor α, leading to altered cell divi- sion, survival, and apoptosis [8].
Naturally occurring carcinogens produced by fungi As with bacteria, the total number of naturally occurring fungal metabo- lites cannot be quantified; however, an estimate may be made. For exam- ple, all fungi produce secondary me- tabolites and there are approximate- ly two unique secondary metabolites per fungal species [9]. Thus, the total number of unique fungal secondary metabolites can be estimated based on the estimated number of fungal species. In 2001, Hawksworth esti- mated that there were about 100 000 formally described fungal species, and that number was considered to represent 5% of the total worldwide [10]. However, this number is likely to be conservative since there are estimated to be 200 000 fungal spe- cies at the tip of Africa alone [11]. The recent completion of an online world catalogue of fungal names (http://www.indexfungorum.org/) has produced a more accurate figure of about 300 000 for the number of spe- cies names of fungi. Assuming two unique secondary metabolites per fungal species produces an estimate of about 600 000 secondary fungal metabolites. Some of these metabo- lites may be carcinogens. Currently, there are eight fungal compounds (excluding alcoholic beverages) that have been evaluated as hav- ing sufficient evidence of carcino- genicity in humans and/or animals (Table 2.11.2).
Human exposure to ciclosporin (also known as cyclosporin, cy- closporine, or cyclosporin A) and griseofulvin occurs exclusively in clinical settings. Nonetheless, en- vironmental exposure is possible since the fungi that produce these agents are common in soil and other matrices [12]. The fungus most used for producing ciclosporin commer- cially, Tolypocladium inflatum, also produces fumonisins B2 and B4 [13]. Dietary exposure to griseofulvin is also possible since both Penicillium
griseofulvum and griseofulvin have been isolated or detected in honey, service tree fruits, and cured meats. That said, it is unlikely that exposure outside clinical settings is of great significance.
Mycotoxins are secondary metab- olites of microfungi that are known to cause sickness or death in humans or animals [14]. Of the toxins listed in Table 2.11.2, six are almost exclu- sively encountered as contaminants of food, although occupational ex- posure also occurs [14]. The foods most likely to be contaminated with high concentrations of mycotoxins are cereal grains and groundnuts, which are dietary staples in many developing countries [14]. The only fungal toxin for which there is sufficient evidence of carcinogenicity in humans is the group referred to as aflatoxins [14,15]. Aflatoxins include aflatoxins B1, B2, G1, G2, and M1. In naturally contaminated food, aflatoxins B1 and B2 or aflatox- ins G1 and G2 usually occur together; however, aflatoxins B2 and G2 are less biologically active and there is limited or inadequate evidence, respectively, of their carcinogenicity in experimen- tal animals. Aflatoxin G1 is less mu- tagenic than aflatoxin B1, and forms fewer DNA adducts. Aflatoxin M1
Fig. 2.11.2. A woman in sub-Saharan Mali sifts newly harvested grain. Cereal crops subject to storage under tropical condi- tions are susceptible to contamination by mycotoxins.
172
occurs almost exclusively in milk and milk products and is less carcinogenic than aflatoxin B1. The mechanisms of action of aflatoxins B1, G1, and M1 are similar in that they can be metabolized to exo-epoxides that can react with DNA, form chemical adducts, and modify genes. Less is known about the metabolism of sterigmatocystin to the exo-epoxide. For aflatoxins, one affected gene in humans is the tumour suppressor gene TP53, and there is a mechanistic link between levels of genetic biomarkers for aflatoxin expo- sure (codon 249-specific TP53 muta- tions) and hepatocellular carcinoma.
Epidemiological studies of afla- toxin have provided strong evidence of a greater than multiplicative interaction between aflatoxin expo- sure and chronic infection with hepa- titis B virus in relation to increased risk of hepatocellular carcinoma [14,15]. Other factors may potentiate the risk of aflatoxin-induced hepa- tocellular carcinoma. For example, many of the populations at high risk
for hepatocellular carcinoma are also consumers of large amounts of maize that is frequently co-contaminated with both aflatoxins and Fusarium verticillioides toxins, in particular fumonisin B1 [14,16]. For example, fumonisin B1 treatment was shown to synergistically increase the num- ber of preneoplastic liver lesions in aflatoxin B1-initiated rats and signifi- cantly increased liver tumours in a dose-dependent manner in aflatoxin B1-initiated trout in a long-term fu- monisin feeding study [14,16]. The mechanistic basis for fumonisin pro- motion of aflatoxin B1 carcinogenicity in trout and preneoplastic lesions in rats is likely a result of inhibited bio- synthesis of ceramide (a lipid media- tor of cell death) and elevation in lev- els of the mitogenic sphingoid base metabolite sphinganine-1-phosphate, a ligand for the G protein-coupled ex- tracellular receptors S1P1–S1P5 [16].
Ochratoxin A is a liver and kidney carcinogen in animals and is found in maize, many small grain cereals,
and other commodities such as grapes, coffee, and even some meat products [14]. The mechanistic basis for the carcinogenicity of ochratoxin A in animals is unclear. Ochratoxin A is a phenylalanine analogue that accumulates in the kidney via an organic anion transporter. Some of its biological effects are due to phe- nylalanine mimicry [14]. Like fumoni- sin, ochratoxin A causes changes in regulatory processes that affect cell survival and proliferation. There is also evidence for DNA adduct formation by ochratoxin A via both direct and indirect processes [14]. Ochratoxin A has, for many decades, been suspected as the cause of the high incidence of urinary tract tumours in the Balkans referred to as Balkan endemic nephropathy. However, epidemiological support for this is not clear. More recently, natural products and food contami- nants derived from plants have also been implicated in the genesis of this disease [17].
Table 2.11.2. Naturally occurring organic chemical carcinogens produced by fungi
Agent (fungal genera)a
Primary source of exposureb
Proposed mechanism of carcinogenicityc Primary target or tumour sitesd
IARC evaluation of carcinogenicity to humans (NTP class)e
Aflatoxins (Aspergillus sp.)
Food contaminant
Metabolism to DNA/protein-reactive epoxide → TP53 mutations
Liver Group 1 (K)
Ciclosporin (Tolypocladium and some Fusarium species)
Pharmaceutical Binds cyclophilin, blocks calcineurin and inhibits its phosphatase activity → immune suppression
Skin, lymphoma, other sites
Group 1 (K)
Aflatoxin M1 (Aspergillus)
Food contaminant
DNA-reactive epoxide → mutations Liver Group 2B
Fumonisin B1 (Fusarium sp.)
Food contaminant
Inhibition of ceramide synthase → disruption of lipid metabolism and signalling pathways regulating cell growth/tumour promoter
Liver, kidney Group 2B
Fusarium moniliforme toxins (Fusarium sp.)
Food contaminant
Inhibition of ceramide synthase → disruption of lipid metabolism and signalling pathways regulating cell growth/tumour promoter
Liver, kidney Group 2B
Griseofulvin (Penicillium sp.)
Antifungal agent
Disruption of microtubule formation via tubulin binding/mitotic inhibitor
Liver, thyroid Group 2B
Ochratoxin A (Aspergillus and Penicillium species)
Food contaminant
Phenylalanine analogue/disruption of signalling pathways regulating cell growth/DNA adducts?
Kidney, liver Group 2B (R)
Sterigmatocystin (Aspergillus sp.)
Food contaminant
Metabolism to DNA-reactive epoxide → mutations
Liver, lung Group 2B
a For all fungi the most likely producing genera are given; the listing is not exhaustive. b All of the agents listed can be produced by biosynthesis in situ, but for ciclosporin and griseofulvin exposure is in a controlled clinical setting. c Where possible, the mechanism described is taken from the most recent IARC Monograph or IARC Scientific Publication. d For cancer studies, the route of exposure was usually oral. e Group 1, carcinogenic to humans; Group 2A, probably carcinogenic to humans; Group 2B, possibly carcinogenic to humans; Group 3, not classifiable as to its carcinogenicity to humans; Group 4, probably not carcinogenic to humans. Only agents in Groups 1, 2A, and 2B are listed here. There are several naturally occurring agents classified by IARC as Group 3, but they are not included in this table. NTP, United States National Toxicology Program: K, known be a human carcinogen; R, reasonably anticipated to be a human carcinogen.
Chapter 2.11 • Naturally occurring chemical carcinogens 173
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
1
Naturally occurring carcinogens produced by plants A large number of compounds pro- duced by plants have been charac- terized as being dietary pesticides [18]. The number of so-called natural pesticides consumed (5000–10 000) and the amount (1.5 g/person/day) were also estimated, and it was speculated that many such biologi- cally active compounds may exhibit carcinogenic activity in animal bio- assays. Whether or not these esti- mates are correct, a large number of biologically active organic chemicals are biosynthesized by plants to deter a range of threats to their survival, and to promote successful competi- tion with other plants and response to injury. Many may be carcinogenic
in rodent bioassays if administered at high levels.
Table 2.11.3 lists agents classi- fied as human or animal carcinogens by IARC and produced by plants that have been used as food, food addi- tives, or herbal medicines. A study conducted by the United States Centers for Disease Control and Prevention found that 10% of adults in the USA consumed herbal medi- cines in 1999 [19].
Herbal medicines containing extracts or preparations from plants producing aristolochic acid have been used since ancient times. Some herbal medicines and foods continue to be a source of inten- tional or inadvertent exposure to aristolochic acid. Contamination of wheat with seeds from Aristolochia species has been suggested as
contributing to Balkan endemic ne- phropathy [19,20]. This suggestion is supported by mechanistic data involving aristolochic acid-derived DNA adducts detected in tumour tissue from people with Balkan en- demic nephropathy. These same ad- ducts are found in animals treated with aristolochic acid, and the same mutation in the TP53 gene was seen in tumour tissues from patients with Balkan endemic nephropathy and in human TP53 knock-in mouse fibro- blasts treated with aristolochic acid [20]. More recently, whole-exome sequencing, based on urothelial carcinoma from 19 individuals with documented exposure to aristolochic acid, has established a mutational signature involving A:T → T:A trans- versions, indicating the power of this technique to reveal individual expo- sure to particular carcinogens [20].
Lasiocarpine, monocrotaline, and riddelliine (Table 2.11.3) are examples of pyrrolizidine alkaloids produced by plants growing in arid regions. For example, riddelliine- producing plants grow in desert areas of North America and are occasional contaminants of herbal products and food, but are most notable for poisoning of grazing
Fig. 2.11.3. Aspergillus flavus (A) grown as colonies on Czapek yeast extract agar (left) and malt extract agar (right) at 25 °C for 7 days; (B,C) heads, bars = 20 μm; (D) conidia, bar = 5 μm. This organism produces aflatoxins.
Fig. 2.11.4. Aristolochia manshuriensis: (A) fruit twig, (B) flower, and (C) transverse section. Aristolochia species have been linked to endemic diseases of the kidney in the general population and in research.
174
Table 2.11.3. Naturally occurring organic chemical carcinogens produced by plants that are used as herbal medicines, food, or food additives or contaminants
Agent/plantsa Primary source of exposureb
Proposed mechanism of carcinogenicityc Primary target or tumour sitesd
IARC evaluation of carcinogenicity to humanse
Aristolochic acid (Aristolochia/Asarum species)
Herbal medicine, food contaminant
Metabolism to DNA-reactive nitrenium ion → TP53 mutations
Urothelium Group 1
Plants containing aristolochic acid (Aristolochiaceae family)
Herbal medicine, food contaminant
Metabolism to DNA-reactive nitrenium ion → TP53 mutations
Urothelium Group 1
Bracken fern (Pteridium sp.)
Food, food contaminant
DNA alkylating activity (ptaquiloside?) Bladder, gastrointestinal tract
Group 2B
Caffeic acid (Coffee, fruits, vegetables, grains, argan oil)
Herbal medicine, food component
Antioxidant/increased cell proliferation/ lipoxygenase inhibitor
Kidney, gastrointestinal tract
Group 2B
Carrageenan, degraded Food additive Oxidative stress/inflammation Colorectum Group 2B
Dihydrosafrole and safrole (Sassafras, essential oils, spices)
Food additive, food contaminant
Metabolism to DNA-reactive sulfo-oxy metabolites → mutations
Liver, lung, oesophagus
Group 2B
1-Hydroxyanthraquinone (Tabebuia avellanedae, Morinda officinalis, and many other plants)
Herbal medicine
Induces inflammation in the colon/oxidative stress/mutagen?
Large intestine Group 2B
Lasiocarpine (Boraginaceae family)
Food contaminant
Metabolism to DNA-reactive dehydronecine → mutations
Liver, lymphoma, leukaemia
Group 2B
Methyleugenol (Essential oils, basil, tarragon, nutmeg, and others)
Food additive, food contaminant
Metabolism to DNA-reactive sulfo-oxy metabolites → mutations
Liver, stomach, kidney
Group 2B
Monocrotaline (Fabaceae family)
Food contaminant
Metabolism to DNA-reactive dehydronecine → mutations
Liver, multiple sites
Group 2B
Riddelliine (Senecio sp., Leguminosae family)
Herbal medicine
Metabolism to DNA-reactive dehydronecine → mutations
Liver, lung, leukaemia
Group 2B
a For all agents examples of natural sources of exposure are given; the listing is not exhaustive. b All of the agents listed can be produced by biosynthesis in situ. c Where possible, the mechanism described is taken from the most recent IARC Monograph or IARC Scientific Publication. d For all cancer studies, including animal bioassays where relevant, the route of exposure was usually oral. e Group 1, carcinogenic to humans; Group 2A, probably carcinogenic to humans; Group 2B, possibly carcinogenic to humans; Group 3, not classifiable as to its carcinogenicity to humans; Group 4, probably not carcinogenic to humans. Only agents in Groups 1, 2A, and 2B are listed here. There are several naturally occurring agents classified by IARC as Group 3, but they are not included in this table.
animals that consume the producing plants. Consumption of herbal teas prepared from plants that produce pyrrolizidine alkaloids has poisoned humans and caused death; however, there are no available human car- cinogenicity data for riddelliine or any of the other pyrrolizidine alka- loids [21]. The mechanism of toxicity and carcinogenicity of pyrrolizidine alkaloids in animals involves their
metabolism to DNA-reactive me- tabolites that bind DNA and cause mutations in specific genes, most notably TP53 and the K-ras onco- gene [21].
There are numerous other plant natural products classified by the IARC Monographs as being carcinogenic. However, exposure or production at levels sufficient to pose appreciable risk of biological effects is primarily
through exposure in industrial/ occupational or clinical settings or by intake of the agent produced syntheti- cally (Table 2.11.4). Examples are the pharmaceutical methoxsalen [17] and chemicals produced in an industrial setting that are also natural products of plant origin. Examples of agents that are natural products produced synthet- ically for industrial uses are listed in Table 2.11.4 [22].
Chapter 2.11 • Naturally occurring chemical carcinogens 175
2 E
T IO
LO G
Y C
H A
P T
E R
2 .1
1
Table 2.11.4. Naturally occurring organic chemical carcinogens derived from plants that are produced for industrial or medical uses but also can occur naturally in foods
Agent/plantsa Primary source of exposureb Proposed mechanism of carcinogenicity c
Primary target or tumour sitesd
IARC evaluation of carcinogenicity to humanse
Methoxsalen (plus ultraviolet A radiation) (Parsnip, parsley, other vegetables)
Pharmaceutical Photo-activation, DNA reactivity → mutations Skin Group 1
5-Methoxypsoralen (Parsnip, parsley, other vegetables, bergamot and lime oils)
Pharmaceutical, cosmetic additive
Photo-activation, DNA reactivity → mutations Skin Group 2A
Acetaldehyde (Many fruits, vegetables, essential oils)
Industrial chemical
Irritant/DNA damage/protein binding Upper respiratory tract
Group 2Bf
Benzophenone (Grapes, black tea, papaya)
Industrial chemical
Endocrine disruption/oxidative stress Liver, kidney, histiocytic sarcoma in multiple sites
Group 2B
Catechol (Fruits, vegetables, argan oil)
Industrial chemical
Metabolism to benzo-1,2-quinone, which binds to protein?
Stomach Group 2B
Cumene (Zingiber officinale, fruits, vegetables, meat, many foods and beverages)
Industrial chemical
Metabolism to DNA-reactive α-methylstyrene oxide → mutations
Respiratory tract, spleen, liver, kidney
Group 2B
Dantron (danthron) (Xyris semifuscata, Rheum palmatum, some insects)
Industrial chemical
DNA damage? Tumour promoter Liver, gastrointestinal tract
Group 2B
2,4-Hexadienal (Fruits, vegetables, meat, many foods and beverages)
Industrial chemical
Direct-acting DNA alkylating agent → mutations?
Forestomach, tongue, adrenal glands
Group 2B
Methyl isobutyl ketone (Fruits, vegetables, meat, beverages)
Industrial chemical
Cytotoxicity, regenerative cell proliferation Liver, kidney (inhalation)
Group 2B
a For all agents examples of natural sources of exposure are given; the listing is not exhaustive. b All of the agents listed can be produced by biosynthesis in situ. c Where possible, the mechanism described is taken from the most recent IARC Monograph or IARC Scientific Publication. d For cancer studies the route of exposure was most likely oral. e Group 1, carcinogenic to humans; Group 2A, probably carcinogenic to humans; Group 2B, possibly carcinogenic to humans; Group 3, not classifiable as to its carcinogenicity to humans; Group 4, probably not carcinogenic to humans. Only agents in Groups 1, 2A, and 2B are listed here. There are several naturally occurring agents classified by IARC as Group 3, but they are not included in this table. f Acetaldehyde associated with consumption of alcoholic beverages is classified as Group 1.
176
References
1. Buckingham J, ed. (2011). Dictionary of Natural Products. London: Chapman and Hall/CRC.
2. Kurtböke DI (2012). Biodiscovery from rare actinomycetes: an eco-taxonomical perspective. Appl Microbiol Biotechnol, 93:1843–1852. http://dx.doi.org/10.1007/ s00253-012-3898-2 PMID:22297430
3. Newman DJ, Cragg GM (2012). Natural products as sources of new drugs over the 30 years from 1981 to 2010. J Nat Prod, 75:311–335. http://dx.doi.org/10.1021/np200 906s PMID:22316239
4. Valerio LG Jr, Arvidson KB, Chanderbhan RF, Contrera JF (2007). Prediction of ro- dent carcinogenic potential of naturally oc- curring chemicals in the human diet using high-throughput QSAR predictive modeling. Toxicol Appl Pharmacol, 222:1–16. http:// dx.doi.org/10.1016/j.taap.2007.03.012 PMID:17482223
5. Wink J. Compendium of Actinobacteria. Available at http://www.gbif-prokarya.de/ microorganisms/wink.html
6. Seipke RF, Kaltenpoth M, Hutchings MI (2012). Streptomyces as symbionts: an emerging and widespread theme? FEMS Microbiol Rev, 36:862–876. http://dx.doi. org/10.1111/j.1574 - 6976.2011.00313.x PMID:22091965
7. Kroiss J, Kaltenpoth M, Schneider B et al. (2010). Symbiotic streptomycetes pro- vide antibiotic combination prophylaxis for wasp offspring. Nat Chem Biol, 6:261–263. http://dx.doi.org/10.1038/nchembio.331 PMID:20190763
8. Paerl HW, Otten TG (2013). Harmful cyano- bacterial blooms: causes, consequences, and controls. Microb Ecol, 65:995–1010. ht tp://dx.doi.org/10.1007/s00248 - 012- 0159-y PMID:23314096
9. Riley RT (1998). Mechanistic interaction of mycotoxins: theoretical considerations. In: Sinha KK, Bhatnagar D, eds. Mycotoxins in Agriculture and Food Safety. New York: Marcel Dekker, pp. 227–253.
10. Hawksworth DL (2001). The magni- tude of fungal diversity: the 1.5 million species estimate revisited. Mycol Res, 105:1422–1432. http://dx.doi.org/10.1017/ S0953756201004725
11. Crous PW, Rong IH, Wood A et al. (2006). How many species of fungi are there at the tip of Africa? Stud Mycol, 55:13–33. ht t p: //dx.doi.o rg /10. 3114 /sim. 5 5.1.13 PMID:18490969
12. Rodríguez MA, Cabrera G, Godeas A (2006). Cyclosporine A from a nonpatho- genic Fusarium oxysporum suppressing Sclerotinia sclerotiorum. J Appl Microbiol, 100:575–586. http://dx.doi.org/10.1111/ j.1365-2672.2005.02824.x PMID:16478497
13. Mogensen JM, Møller KA, von Freiesleben P et al. (2011). Production of fumonisins B2 and B4 in Tolypocladium species. J Ind Microbiol Biotechnol, 38:1329–1335. http:// dx.doi.org/10.1007/s10295 - 010 - 0916 -1 PMID:21132348
14. Pitt J, Wild CP, Baan RA et al., eds (2012). Improving Public Health through Mycotoxin Control. Lyon: IARC (IARC Scientific Publications Series, No. 158).
15. IARC (2012). Chemical agents and related occupations. IARC Monogr Eval Carcinog Risks Hum, 100F:1–599. PMID:23189753
16. Bulder AS, Arcella D, Bolger M et al. (2012). Fumonisins (addendum). In: Safety Evaluation of Certain Food Additives and Contaminants: Prepared by the Seventy- fourth Meeting of the Joint FAO/ WHO Expert Committee on Food Additives (JECFA). Geneva: WHO (WHO Food Additives Series, No. 65), pp. 325 –794. Available at whqlibdoc.who.int/publi cations/2012/9789241660655_eng.pdf.
17. IARC (2012). Pharmaceuticals. IARC Monogr Eval Carcinog Risks Hum, 100A:1– 437. PMID:23189749
18. Ames BN, Gold LS (2000). Paracelsus to parascience: the environmental cancer distraction. Mutat Res, 447:3–13. http:// dx.doi.org/10.1016/S0027-5107(99)00194-3 PMID:10686303
19. Grollman AP (2013). Aristolochic acid nephropathy: harbinger of a global iat- rogenic disease. Environ Mol Mutagen, 54:1–7. http://dx.doi.org/10.1002/em.21756 PMID:23238808
20. Hoang ML, Chen CH, Sidorenko VS et al. (2013). Mutational signature of aristolochic acid exposure as revealed by whole-ex- ome sequencing. Sci Transl Med, 5:ra102. PMID:23926200
21. Chen T, Mei N, Fu PP (2010). Genotoxicity of pyrrolizidine alkaloids. J Appl Toxicol, 30:183–196. PMID:20112250
22. IARC (2012). Some chemicals present in industrial and consumer products, food and drinking-water. IARC Monogr Eval Carcinog Risks Hum, 101:1–610.
Early-life exposures, birth cohorts, and noncommunicable diseases (with special reference to cancer) 177
Summary There is strong evidence that several early-life factors influence the occur- rence of different types of cancer as well as other noncommunicable dis- eases. Although stand-alone cohorts seldom have the statistical power to study cancer incidence directly, they can provide important information on how early-life exposures affect risk factors for common adult cancers, and thus contribute to their preven- tion. A potential solution is to pool several cohorts to increase sample size; such efforts have been limited by different definitions of exposures, outcomes, and confounders, as well as variable ages at ascertainment. Greater efforts should be made to coordinate data collection in cohorts that are still in early life so that, sev- eral decades from now, the problems that we are currently facing will have been overcome.
Why birth cohorts are important The role of early influences in shap- ing adult life has long been recog- nized. Reviews of this topic often cite John Milton (“The childhood shows the man”; 1671) and William Wordsworth (“The Child is father of the Man”; 1802) as examples of the
long-standing awareness of the im- portance of early factors.
In the 1960s the field was sub- stantially advanced by René Dubos, who proposed the idea of “biologi- cal Freudianism” [1]. He built upon Sigmund Freud’s work on the last- ing effect of early behaviours, and expanded it to study the role of early stimuli in general, with special em- phasis on nutritional and infectious exposures. Among many remark- able insights, Dubos highlighted the importance of the gut microbiota, a topic that is currently high on our re- search agenda. The work of Dubos, however, relied heavily on animal models instead of epidemiological studies.
Current interest in early expo- sures was boosted by the work of David Barker since the late 1970s. This evolved into the field of research that is currently known as develop- mental origins of health and disease [2,3], focusing on how several non- communicable diseases (NCDs) – particularly cardiovascular disease (CVD) and metabolic diseases – may be programmed by early events. So far, research on developmental ori- gins of health and disease has main- ly addressed exposures related to nutrition and growth patterns, both in utero and during the first years of life, but the concept applies to many different types of risk factors.
Early-life exposures, birth cohorts, and noncommunicable diseases (with special reference to cancer)
Cesar G. Victora
Cesar G. Victora is an emeritus
professor of epidemiology at the
Federal University of Pelotas in
Brazil. He has conducted extensive
research on maternal and child health
and nutrition, on equity issues, and
on the evaluation of health services.
Dr Victora obtained his medical
degree from the Federal University
of Rio Grande do Sul and a Ph.D.
in health-care epidemiology from
the London School of Hygiene and
Tropical Medicine. His unit in Pelotas
is a WHO Collaborating Centre in
Maternal Health and Nutrition. He is a
former international associate editor of
the American Journal of Public Health
and was one of the coordinators of
the 2008 Lancet series on maternal
and child undernutrition. Dr Victora
was elected to the Brazilian Academy
of Sciences in 2006. He is currently a
visiting professor in the Department of
International Health at the
Johns Hopkins Bloomberg School
of Public Health.
178
Experimental studies on the long-term consequences of early- life exposures on NCDs are most often unethical, impracticable, or both. There are some remarkable exceptions, such as the nutrition supplementation trials carried out in Guatemala in the 1960s [4]. Other trials, such as the Belarus breast- feeding promotion trial [5] and the Hyderabad nutrition trial [6], are likely to eventually produce results, but their subjects are still too young.
As a result of the dearth of ex- perimental studies, birth cohorts provide by far the largest amount of epidemiological evidence on early- life exposures. For most risk factors, prospective cohorts with primary data collection by trained and stan- dardized field workers offer more re- liable and complete information on exposures and confounders than do retrospective cohorts or case–con- trol studies.
There are many more studies about early-life influences on CVD and related metabolic diseases than is the case for cancer. Two main rea- sons seem to explain this differen- tial. First, incidence rates for CVD and metabolic conditions tend to be higher than those of most, if not all, cancers. Second, there are many well-established, easily measurable risk factors and biomarkers that may be used as surrogates for the risk of CVD and metabolic conditions, even among young adults. For example, studies of the associations between early growth patterns and adult body mass index, blood pressure, serum lipids, C-reactive protein, and so on are plentiful in the literature. The relatively high incidence of these diseases and the substantial preva- lence of reliable risk markers mean that a relatively small sample size is adequate, and such associations may be investigated within cohorts with a few thousand participants, even if the subjects are still young adults.
The situation for cancer studies is more complex. Incidence rates are low compared with those for CVD or metabolic diseases, particularly among young adults. As a conse-
quence, large numbers of subjects must be followed up for several de- cades to study disease incidence. And while early determinants as- sociated with common risk factors for cancer (e.g. smoking and high body mass index) may be studied in individual cohorts, there is a sharp contrast between the large number of easily measureable CVD biomark- ers and the few that are currently available for cancer.
Research on the early determi- nants of cancer is therefore harder to carry out, as well as costlier, than similar studies for other NCDs, and will entail special requirements, which are addressed below.
Birth cohorts for studying NCDs: high-income countries Existing prospective birth cohorts may be classified according to their sample size and duration of follow- up, as well as the level of economic development of the country where they take place.
A recent PubMed search for birth cohort studies produced about 6900 references. Of the 20 countries from which the largest number of articles originated, only Brazil (234 refer- ences) and India (47 references) do not fall into the high-income country category. (As far as I know, there is no comprehensive roster of birth cohorts, but there are ongoing at- tempts to register such studies on web sites – see http://www.birthco horts.net and http://www.chicospro ject.eu – as well as in the Cohort Profiles section in the International Journal of Epidemiology.)
Cohorts from high-income coun- tries may be divided into two groups, according to duration of follow-up to the present time. The first group is made up of prospective cohorts with several decades of follow-up. The best-known examples include British and Scandinavian cohorts [7–12]. Typically, these studies have sam- ple sizes of about 10 000 or fewer participants. While such numbers are sufficient for studying several CVDs, metabolic conditions, or men- tal illnesses, they are seldom large
enough for cancer outcomes. Few such analyses are therefore avail- able in the literature – one example is a report on leukaemia [13] from the Jerusalem birth cohort, which started with more than 90 000 new- borns recruited in the 1960s and 1970s [14].
A limitation of long-running co- horts is that – by definition – expo- sure information was collected many years ago. Few cohort studies col- lected biological samples, question- naires were often very short and not standardized, and information on important exposures and confound- ers was not collected. An example is the Aberdeen Children of the 1950s Study, which failed to collect what is today regarded as essential informa- tion, such as parental smoking sta- tus [15]. This limitation restricts the ability of such studies to answer im- portant questions on consequences of early risk factors.
The second group of cohorts from high-income settings includes young cohorts. Some of these cohorts have very large sample sizes, such as those launched in the past decade or two in Denmark, Norway, and the USA, among other countries [16]. Most in- clude state-of-the-art exposure mea- surement tools, as well as collection and storage of biological samples. In terms of cancer epidemiology, early results from several of the new cohorts are already being pooled in the International Childhood Cancer Cohort Consortium [16], but unfortu- nately it will take many years before these can provide information on how early exposures affect risk factors for common adult cancers. A limitation of very large cohorts is that there is a trade-off between sample sizes and the frequency and intensity of mea- surements on individual participants. The most informative cohorts tend to be those with smaller sample sizes and more detailed measurements.
Birth cohorts for studying NCDs: low- and middle- income countries As discussed above, most publi- cations about birth cohort studies
179 Early-life exposures, birth cohorts, and noncommunicable diseases
(with special reference to cancer)
originate in high-income countries. Yet these countries account for only 11 million births a year, compared with 121 million births in low- and middle-income countries (LMICs) (http://www.unicef.org/sowc2012/ statistics.php). Findings from high- income countries tend to influence global policies, and yet these may not be applicable to LMIC popula- tions. As an example, concerns raised about the long-term con- sequences of higher birth weight or rapid weight gains in infancy, derived from cohorts where infant undernutrition is rare, do not apply to populations where poor nutrition and growth faltering are still highly prevalent [17]. In LMICs, rapid catch- up growth may be essential, not only for short-term survival but also for long-term health and human capital [18,19].
There are other reasons why more cohorts from LMICs are need- ed [20]. Risk factors may operate differently in each setting, as is the case for asthma [21]. The frequency of some exposures, including ma- ternal and fetal undernutrition and infectious diseases, are markedly higher in LMICs. Even if exposures appear to have similar prevalence, their nature may vary; for example, in LMICs physical activity is mostly due to commuting and manual domes- tic labour, whereas in high-income countries leisure-time activities are the prevailing types of physical ac- tivity [22]. Last, confounders may operate in different directions in high-income countries and in LMICs, as was recently shown for the con- founding effect of socioeconomic status on the association between breastfeeding duration and precur- sors for chronic diseases [23].
For these reasons there is a third relevant group of birth cohorts, comprising those from LMICs with long follow-up times [24]. The best- known of these studies are part of the Consortium of Health-Orientated Research in Transitioning Societies collaboration [25]. These five cohorts are providing important information on precursors for NCDs, particu- larly CVD and diabetes, sometimes
with conflicting results from those reported from high-income settings [26–30]. The Consortium of Health- Orientated Research in Transitioning Societies collaboration, however, is unlikely to provide direct evidence on cancer incidence. First, the total number of subjects under follow-up is close to 10 000, resulting in low statistical power for rare outcomes. Second, linkage studies are not possible because of inadequate sources of secondary data such as cancer registration and vital statis- tics. Another limitation is that none of these cohort studies collected biological samples in early life that could be analysed for biomarkers.
Other types of birth cohorts The three categories of birth cohort studies discussed so far are not exhaustive. There are also several new, relatively small cohorts both in high-income countries and in LMICs. In LMICs, as far as I am aware, there is only one new cohort with more than a few thousand subjects: the China Children and Families Cohort Study, which started in 2006–2007 and plans to enrol 300 000 births [16].
There are also several retrospec- tive cohorts, as well as cohorts that are fully based on record linkage, particularly in European countries. Most of these include subjects born in the 1970s or later, so that follow- up to the present time is about four decades. Also, because informa- tion on exposures is derived from routine records, available data are relatively restricted and their quality is variable. This is not to say that these cohorts cannot make impor- tant contributions to answering spe- cific questions – for example, on the importance of birth weight – but their ability to address multiple risk factors and to control for confounding may be limited.
Famine studies, a special type of retrospective cohort, have led to important insights into the timing of nutritional insults and their effect on adult outcomes, including cancer [31–35]. However, these populations faced unusual hardship conditions
during relatively short time periods, and as such may not reflect the na- ture of exposures usually faced by populations, which tend to be less acute but have a longer duration.
Implications for future studies and collaborations Birth cohorts have thus far provided invaluable amounts of evidence on the role of early-life exposures in adult diseases, but there is undoubt- edly much more to be learned from existing and new cohorts.
A critical issue is that birth co- horts take a long time to produce results. Those launched several decades ago, whose members are now at the age of highest incidence for most NCDs, including cancers, tend to have limited information on most early-life exposures. Cohorts launched in the past two decades tend to have much better exposure data as well as larger sample sizes, but will take a long time to produce results. A second trade-off is that large cohorts have greater statistical power but tend to have less frequent and less detailed assessments of early-life exposures than do smaller cohorts.
An obvious solution, therefore, is to pool cohorts to increase sample size and enhance external validity. An excellent example is the Emerging Risk Factors Collaboration, with data from more than 100 prospec- tive studies with more than 2 million participants (http://www.phpc.cam. ac.uk/ceu/research/erfc/). However, these cohorts recruited adults rather than pregnant women or newborns, and were assembled relatively re- cently, so that better and more stand- ardized information on exposures is available, compared with what may be obtained from older birth cohorts. There are also several new initiatives for pooling birth cohorts to address specific health conditions, but these tend to be restricted to relatively young cohorts [16,36,37].
I am unaware of any successful initiatives in pooling several long- term birth cohorts from high-income countries, to study cancer outcomes
180
or other NCDs. Even with large sam- ple sizes, however, long-running co- horts are unlikely to have collected much information that is now consid- ered as essential. For example, few if any cohorts have stored biological samples collected in childhood.
My own experience, and certain- ly that of others involved in such ex- ercises, is that there are major gains from pooling cohorts, but there are also important difficulties. Pooling is easier said than done. We recently summarized the barriers we had to face in the Consortium of Health- Orientated Research in Transitioning Societies network [25]:
Some of the weaknesses include: (i) differences in variable definitions, or in measurement techniques across sites (this affects exposures, out- comes and confounding variables), which means that major effort has gone into producing a common data set; (ii) the different ages of individu- als across the five cohorts and the different time periods they reflect; (iii) the different ages for which data are available throughout infancy and childhood; (iv) heterogeneity in the results for some of the analy- ses, for example, those on body composition. In order to overcome these limitations, we have restricted potential analyses to those includ- ing variables collected consistently across the cohorts; likewise, we have limited the analyses to ages with data available for all (or most) cohorts. In some analyses, different outcome variables have been used (e.g. pre-hypertension for adoles- cents and hypertension for adults) because of the different ages of the individuals across cohorts.
Similar difficulties have been re- ported by investigators involved in pooling younger birth cohorts from high-income countries [16].
Stand-alone, long-term birth cohorts also face important chal- lenges. Attrition is possibly the most
important issue. In the 1982 Pelotas (Brazil) birth cohort study, which I coordinate, we achieved relatively high follow-up rates (68% at the re- cently completed 30-year visit), but even so individuals who were lost to follow-up differ in many ways from those who came to the clinic. Ensuring consistency and standardi- zation of exposure measures over time – when methods are constantly evolving – is another major chal- lenge. Obtaining continued funding is also problematic, because donor fatigue affects even the most suc- cessful studies.
How can these challenges be overcome? Ideally, one would like to design an international birth co- hort, involving multiple countries at different stages of development, with large sample sizes in each country. Measurements would take advan- tage of the variety of new tools that are now available, including genetic, epigenetic, microbiome, and expo- some assessments [38], among others. This design would allow analyses not only at an individual level but also at an ecological level, because some exposures may affect most, if not all, subjects from a given site [39]. Information on exposures would be collected at similar ages at all sites, using similar methods, including questionnaires, biological samples, and environmental mea- surements. Obviously, this would be a massive undertaking with major practical difficulties, as has been shown by the long inception period that has been required for the United States National Children’s Study (http://www.nationalchildrensstudy. gov/Pages/default.aspx). In addition to the major scientific challenges to be faced, finding sufficient finan- cial resources for such a large co- hort would require pooling funding agencies over an extended period of time – possibly an even harder task than pooling research teams.
A special plea must be made for multipurpose birth cohorts, ad- dressing several different outcomes, related not only to health but also to human capital. This will allow us to investigate the full impact of early-life factors, because some exposures may positively affect some adult outcomes and have a detrimental effect on others. For example, adult height is largely determined by the linear growth velocity up to the age of 2 years [40]. Taller adults have greater incidence of several types of cancer [41] but also tend to be more intelligent, have higher schooling, and earn higher wages [19], as well as having lower all-cause mortal- ity, largely due to protection against deaths due to CVD [42]. Only studies that capture multiple dimensions of human health and development can assess the overall effect of early-life exposures and growth patterns over the life-course.
In conclusion, there is strong evidence that several early-life fac- tors influence the occurrence of different types of cancer as well as other NCDs. Whereas stand-alone cohorts seldom have the statistical power to study cancer incidence di- rectly, they may provide important data on how early-life exposures af- fect the adult prevalence of risk fac- tors for cancer, and thus contribute to their prevention. Pooling several cohorts to increase sample size is a potential solution, but so far such efforts have been limited by different definitions of exposures, outcomes, and confounders, as well as variable ages at ascertainment. Greater ef- forts should be made to coordinate data collection efforts in cohorts that are still in early life so that, several decades from now, the problems that we are currently facing will have been overcome.
I would like to thank Ezra Susser and Jørn Olsen for useful inputs to this Perspective.
181 Early-life exposures, birth cohorts, and noncommunicable diseases
(with special reference to cancer)
References
1. Dubos R, Savage D, Schaedler R (2005). Biological Freudianism: lasting effects of early environmental influences. 1966. Int J Epidemiol, 34:5–12. http://dx.doi. org/10.1093/ije/dyh309 PMID:15649968
2. Barker DJ (1992). The fetal and infant ori- gins of adult disease. BMJ, 301:1111. http:// dx.doi.org/10.1136/bmj.301.6761.1111 PMID:2252919
3. Gluckman P, Hanson M, eds (2006). Developmental Origins of Health and Disease. Cambridge, UK: Cambridge University Press.
4. Stein AD, Melgar P, Hoddinott J, Martorell R (2008). Cohort Profile: the Institute of Nutrition of Central America and Panama (INCAP) Nutrition Trial Cohort Study. Int J Epidemiol, 37:716–720. PMID:18285366
5. Patel R, Oken E, Bogdanovich N et al. (2013). Cohort Profile: The Promotion of Breastfeeding Intervention Trial (PROBIT). Int J Epidemiol, 7:7. PMID:23471837
6 . K i n r a S , R a m e s h w a r S a r m a K V, Ghafoorunissa et al. (2008). Effect of inte- gration of supplemental nutrition with public health programmes in pregnancy and early childhood on cardiovascular risk in rural Indian adolescents: long term follow-up of Hyderabad nutrition trial. BMJ, 337:a605. h t t p : / / d x . d o i . o r g / 1 0 .113 6 / b m j . a 6 0 5 PMID:18658189
7. Kajantie E, Barker DJP, Osmond C et al. (2008). Growth before 2 years of age and serum lipids 60 years later: the Helsinki Birth Cohort study. Int J Epidemiol, 37:280– 289. http://dx.doi.org/10.1093/ije/dyn012 PMID:18267964
8. Elliott J, Shepherd P (2006). Cohort Profile: 1970 British Birth Cohort (BCS70). Int J E p i d e m i o l, 3 5:8 3 6 – 8 4 3 . ht t p: //d x .d o i . org/10.1093/ije/dyl174 PMID:16931528
9. Power C, Elliott J (2006). Cohort Profile: 1958 British Birth Cohort (National Child Development Study). Int J Epidemiol, 35:34 – 41. http://dx.doi.org/10.1093/ije/ dyi183 PMID:16155052
10. Stenberg SA, Vågerö D (2006). Cohort Profile: the Stockholm Birth Cohort of 1953. Int J Epidemiol, 35:546–548. http://dx.doi. org/10.1093/ije/dyi310 PMID:16377656
11. Syddall HE, Aihie Sayer A, Dennison EM et al. (2005). Cohort Profile: the Hertfordshire Cohort Study. Int J Epidemiol, 34:1234– 1242. http://dx.doi.org/10.1093/ije/dyi127 PMID:15964908
12. Rantakallio P (1988). The longitudinal study of the Northern Finland Birth Cohort of 1966. Paediatr Perinat Epidemiol, 2:59–88. http://dx.doi.org/10.1111/j.1365-3016.1988. tb00180.x PMID:2976931
13. Paltiel O, Harlap S, Deutsch L et al. (2004). Birth weight and other risk factors for acute leukemia in the Jerusalem Perinatal Study cohort. Cancer Epidemiol Biomarkers Prev, 13:1057–1064. PMID:15184264
14. Harlap S, Davies AM, Deutsch L et al. (2007). The Jerusalem Perinatal Study cohort, 1964–2005: methods and a re- view of the main results. Paediatr Perinat Epidemiol, 21:256–273. http://dx.doi. org/10.1111/j.1365 -3016.2007.00799.x PMID:17439536
15. Leon DA, Lawlor DA, Clark H, Macintyre S (2006). Cohort Profile: the Aberdeen Children of the 1950s Study. Int J Epidemiol, 35:549–552. http://dx.doi. org/10.1093/ije/dyi319 PMID:16452107
16. Brown RC, Dwyer T, Kasten C et al.; International Childhood Cancer Cohort Consortium (I4C) (2007). Cohort Profile: the International Childhood Cancer Cohort Consortium (I4C). Int J Epidemiol, 36:724– 730. http://dx.doi.org/10.1093/ije/dyl299 PMID:17255350
17. Black RE, Allen LH, Bhutta ZA et al.; Maternal and Child Undernutrition Study Group (2008). Maternal and child un- dernutrition: global and regional expo- sures and health consequences. Lancet, 371:243–260. http://dx.doi.org/10.1016/ S0140-6736(07)61690-0 PMID:18207566
18. Victora CG, Barros FC (2001). Commentary: The catch-up dilemma – relevance of Leitch’s ‘low–high’ pig to child growth in developing countries. Int J Epidemiol, 30:217–220. http://dx.doi. org/10.1093/ije/30.2.217 PMID:11369717
19. Victora CG, Adair L, Fall C et al.; Maternal and Child Undernutrition Study Group (2008). Maternal and child undernutrition: consequences for adult health and human capital. Lancet, 371:340–357. http://dx.doi. org/10.1016/S0140 - 673 6(07) 61692- 4 PMID:18206223
20. Victora CG, Barros FC (2012). Cohorts in low- and middle-income countries: from still photographs to full-length movies. J Adolesc Health, 51 Suppl:S3–S4. http://dx.doi. org/10.1016/j.jadohealth. 2012.09.0 03 PMID:23283157
21. Weinmayr G, Genuneit J, Nagel G et al.; ISAAC Phase Two Study Group (2010). International variations in associations of allergic markers and diseases in chil- dren: ISAAC Phase Two. Allergy, 65:766– 775. ht tp: //dx.doi.org /10.1111/ j.13 9 8 - 9995.2009.02283.x PMID:20028376
22. Hallal PC, Andersen LB, Bull FC et al.; Lancet Physical Activity Series Working Group (2012). Global physical activity levels: surveillance progress, pitfalls, and prospects. Lancet, 380:247–257. http:// dx.doi.org/10.1016/S0140-6736(12)60646-1 PMID:22818937
23. Brion MJ, Lawlor DA, Matijasevich A et al. (2011). What are the causal effects of breastfeeding on IQ, obesity and blood pressure? Evidence from comparing high- income with middle-income cohorts. Int J Epidemiol, 40:670–680. http://dx.doi. org/10.1093/ije/dyr020 PMID:21349903
24. Harpham T, Huttly S, Wilson I, De Wet T (2003). Linking public issues with private troubles: panel studies in developing coun- tries. J Int Dev, 15:353–363. http://dx.doi. org/10.1002/jid.988
25. Richter LM, Victora CG, Hallal PC et al.; COHORTS Group (2012). Cohort Profile: the Consortium of Health-Orientated Research in Transitioning Societies. Int J Epidemiol, 41:621–626. http://dx.doi. org/10.1093/ije/dyq251 PMID:21224276
26. Adair LS, Fall CH, Osmond C et al.; COHORTS Group (2013). Associations of linear growth and relative weight gain during early life with adult health and hu- man capital in countries of low and middle income: findings from five birth cohort stud- ies. Lancet, 382:525–534. http://dx.doi. org /10.1016/S0140 - 673 6(13) 6 0103 - 8 PMID:23541370
27. Kuzawa CW, Hallal PC, Adair L et al.; COHORTS Group (2012). Birth weight, postnatal weight gain, and adult body composition in five low and middle in- come countries. Am J Hum Biol, 24:5–13. ht t p: //dx.do i.o r g /10.10 0 2 /ajhb. 212 27 PMID:22121058
28. Adair LS, Martorell R, Stein AD et al. (2009). Size at birth, weight gain in in- fancy and childhood, and adult blood pressure in 5 low- and middle-income- country cohorts: when does weight gain matter? Am J Clin Nutr, 89:1383–1392. http://dx.doi.org/10.3945/ajcn.2008.27139 PMID:19297457
29. Fall CH, Borja JB, Osmond C et al.; COHORTS Group (2011). Infant-feeding patterns and cardiovascular risk factors in young adulthood: data from five cohorts in low- and middle-income countries. Int J Epidemiol, 40:47–62. http://dx.doi. org/10.1093/ije/dyq155 PMID:20852257
30. Norris SA, Osmond C, Gigante D et al.; COHORTS Group (2012). Size at birth, weight gain in infancy and childhood, and adult diabetes risk in five low- or middle- income country birth cohorts. Diabetes Care, 35:72–79. http://dx.doi.org/10.2337/ dc11-0456 PMID:22100968
31. Li QD, Li H, Li FJ et al. (2012). Nutrition deficiency increases the risk of stomach cancer mortality. BMC Cancer, 12:315. http://dx.doi.org/10.1186/1471-2407-12-315 PMID:22838407
182
32. Schouten LJ, van Dijk BAC, Lumey LH et al. (2011). Energy restriction during child- hood and early adulthood and ovarian cancer risk. PLoS One, 6:e27960. http:// dx.doi.org/10.1371/journal.pone.0027960 PMID:22132180
33. Hughes LA, van den Brandt PA, Goldbohm RA et al. (2010). Childhood and adoles- cent energy restriction and subsequent colorectal cancer risk: results from the Netherlands Cohort Study. Int J Epidemiol, 39:1333–1344. http://dx.doi.org/10.1093/ ije/dyq062 PMID:20427463
34. Hughes LA, van den Brandt PA, de Bruïne AP et al. (2009). Early life exposure to famine and colorectal cancer risk: a role for epigenetic mechanisms. PLoS One, 4:e7951. http://dx.doi.org/10.1371/journal. pone.0007951 PMID:19956740
35. Elias SG, Peeters PH, Grobbee DE, van Noord PA (2005). The 1944–1945 Dutch fam- ine and subsequent overall cancer incidence. Cancer Epidemiol Biomarkers Prev, 14:1981– 1985. http://dx.doi.org/10.1158/1055-9965. EPI-04-0839 PMID:16103448
36. Gehring U, Casas M, Brunekreef B et al. (2013). Environmental exposure assess- ment in European birth cohorts: results from the ENRIECO project. Environ Health, 12:12–18. http://dx.doi.org/10.1186/1476- 069X-12-8 PMID:23374669
37. Bousquet J, Anto J, Sunyer J et al. (2012). Pooling birth cohorts in allergy and asthma: European Union-funded initia- tives - a MeDALL, CHICOS, ENRIECO, and GA2LEN Joint Paper. Int Arch Allergy Immunol, 161:1–10. http://dx.doi. org/10.1159/000343018 PMID:23258290
38. Wild CP (2005). Complementing the genome with an “exposome”: the outstanding chal- lenge of environmental exposure measure- ment in molecular epidemiology. Cancer Epidemiol Biomarkers Prev, 14:1847–1850. http://dx.doi.org/10.1158/1055-9965.EPI- 05-0456 PMID:16103423
39. Rose G (1985). Sick individuals and sick populations. Int J Epidemiol, 14:32–38. h t t p: //d x .d o i .o r g /10 .10 9 3 / i j e /14.1. 3 2 PMID:3872850
40. Stein AD, Wang M, Martorell R et al.; COHORTS Group (2010). Growth pat- terns in early childhood and final attained stature: data from five birth cohorts from low- and middle-income countries. Am J Hum Biol, 22:353–359. http://dx.doi. org/10.1002/ajhb.20998 PMID:19856426
41. Green J, Cairns BJ, Casabonne D et al.; Million Women Study collaborators (2011). Height and cancer incidence in the Million Women Study: prospective cohort, and meta-analysis of prospective studies of height and total cancer risk. Lancet Oncol, 12:785–794. http://dx.doi.org/10.1016/S1470- 2045(11)70154-1 PMID:21782509
42. Wormser D, Angelantonio ED, Kaptoge S et al.; Emerging Risk Factors Collaboration (2012). Adult height and the risk of cause- specific death and vascular morbidity in 1 million people: individual participant meta-analysis. Int J Epidemiol, 41:1419– 1433. http://dx.doi.org/10.1093/ije/dys086 PMID:22825588
An understanding of cancer biology is critical to de- velop rationally designed therapy and offer preven- tive options. For decades, relevant knowledge has been sought by describing how cancer cells differ from normal cells – something being elucidated far more rapidly in the post-genomic era. Specification of the human genome has enabled the identification of how various types of tumour cells differ from normal cells in multiple parameters, including relevant so- matic mutations and altered gene expression, which is often determined through epigenetic change such as alterations in DNA methylation patterns. Mutation or epigenetic change may mediate, amongst other effects, altered metabolism or modified intracellular signalling in response to growth-altering stimuli. In parallel, the roles of cancer stem cells and the tumour microenvironment have been recognized. Accordingly, description of how inflammation, growth of new blood vessels, and modification of the immune response mediate tumour growth now also enables options for cancer prevention and treatment to be identified.
Cancer biology
3
184
Summary • All cancers harbour mutations
in their genome, some of which have profound effects on the bi- ology of the cancer cell by driv- ing tumour growth that subse- quently leads to cancer-related death.
• The types and subtypes of can- cer, as well as the spectrum of cancer-causing mutations, dif- fer considerably across world populations because of genetic diversity within communities and between individuals as well as diverse chemical exposures, infections, dietary components, and other factors that give rise to mutations.
• The International Cancer Ge- nome Consortium was estab- lished to generate a catalogue of cancer mutations for more than 25 000 tumours represent- ing 50 types of cancer that are variously of clinical and societal importance across the globe.
• The Consortium is making the data available to the research community as rapidly as pos- sible to accelerate research into the etiology of cancer and development of tailored cancer treatment based on unique
3.1 3 BIOLOGY
Genomics Thomas J. Hudson James D. McKay (reviewer)
Teruhiko Yoshida (reviewer)
mutation profiles that are spe- cific to individual tumours.
From the perspective of genome sci- entists, cancer is a disease of the genome. All cancers harbour muta- tions in their genome. Some of these mutations have profound effects on the biology of the cancer cell by driv- ing tumour growth. Some mutations can inactivate genes that usually protect cells from abnormal prolifera- tion; these are classically referred to as tumour suppressor genes. Other mutations can result in proteins with oncogenic functions that stimulate cell growth or provide other advan- tages to cancer cells, thus affecting biological pathways involved in the transformation of normal cells into cancer cells [1,2].
As the new millennium dawned, scientists working on the Human Genome Project revealed the first comprehensive view of the sequence of a “normal” human genome and its constituents: protein-coding genes, non-coding RNAs, and non-coding sequences conserved across spe- cies, providing clues to regulatory el- ements required for gene expression as well as catalogues of genetic vari- ants (or polymorphisms) representing genomic diversity among individuals within and across populations [3]. When the first draft of the human
genome sequence was released, it was estimated that more than 100 on- cogenes and 30 tumour suppressor genes had been identified [4].
The types and subtypes of cancer, as well as the spectrum of cancer- causing mutations, vary considerably across world populations as a result of environmental differences, such as chemical exposures, infections, diet, and other factors, and popula- tion diversity in the genes encod- ing molecules that interact with the exogenous factors. In a new era of genome sciences and technologies [5,6], it became possible to conceive of systematic genome-wide searches of mutations in tumour genomes from a diversity of cancers originating in various organs of the body and affect- ing individuals in different parts of the world. This concept became a reality relatively quickly as it became clear that cancer-causing mutations could be used as biomarkers to inform clini- cal decisions. Thus, KRAS mutations may indicate non-responsiveness to drugs that inhibit a growth factor (EGFR) often found in colon cancer. Biomarkers may inform application of new targeted drugs such as tras- tuzumab, for breast cancers over- expressing the HER2 protein, and imatinib, a tyrosine kinase inhibitor effective in chronic myeloid leukae- mia harbouring a specific mutation called the Philadelphia translocation
Chapter 3.1 • Genomics 185
3 B
IO LO
G Y
C H
A P
T E
R 3
.1
[7]. Understanding the mutational profiles of tumours in patients in dif- ferent regions of the world is funda- mental to translating genomic knowl- edge into diagnostic and treatment strategies, and for developing new targeted solutions for cancer control.
An international network was established to generate a catalogue of cancer mutations for more than 25 000 tumours representing 50 types of cancer [8]. This monumen- tal effort came together in 2007 as the result of advances in genome sequencing technologies [6] and ex- perience gained in pilot projects in the United Kingdom and the USA. International leaders in cancer and genome research convened in Toronto to establish a framework for an internationally coordinated effort that would generate comprehensive catalogues of cancer mutations [9]. With support from governmental and philanthropic granting agen- cies, teams of clinicians, scientists, ethicists, and computer scientists assumed responsibility for the com- prehensive study of cancer genomes from one or more types of cancer un- der the auspices of the International Cancer Genome Consortium. This decade-long project will ultimately engage more than 1000 clinicians and scientists, who will be involved in generating data and analysing more than 25 000 cancer samples. Following the principles adopted by
the Human Genome Project, the Consortium is making the data avail- able to the research community as rapidly as possible, and with minimal restrictions, to accelerate research into the causes and control of cancer.
Between October 2008 and January 2013, there has been a steady increase in the number of tu- mour genome analyses undertaken by the International Cancer Genome Consortium, and the initial goal of 25 000 has now been surpassed (Fig. 3.1.1). As of January 2013, the halfway stage of the International Cancer Genome Consortium effort, 55 projects were under way in Asia, Australia, Europe, North America, and South America (Fig. 3.1.2). These projects will cover most of the common tumour types in participat- ing countries (Fig. 3.1.3). At the half- way point in a decade-long initiative, the Consortium is providing cancer genome data sets from 7358 tumours (Fig. 3.1.4) to the international com- munity via an Internet portal. To en- able users to access the data and perform queries that will enable fur- ther cancer research, several data portals and search tools have been created by the Consortium and other organizations (Table 3.1.1). Despite this rapid growth of information, the task of completing the project and ac- celerating the translation of cancer genome information into clinically
useful data will face several chal- lenges, which are summarized below.
Biological challenges Biological complexity results from many factors. One is that there are numerous types of cancer-causing mutations. The simplest mutations affect single DNA bases, such as the substitution of one nucleotide for another or the insertion or dele- tion of a nucleotide. Some mutations involve segments of a chromosome that are increased in number, typi- fied by copy number gains and am- plifications; others involve deletions of one or both copies of a gene, or sequences that are aligned in a di- rection opposite to their original ori- entations on normal chromosomes, referred to as inversions. Some mutations involve translocations between chromosomes. For exam- ple, the Philadelphia translocation mentioned previously results from the joining of chromosomes 9 and 22 and is observed in approximately 95% of patients with chronic myeloid leukaemia. In 2–3% of all cancers and more frequently in bone cancers and medulloblastomas, rearrange- ments are widespread and complex, a phenomenon called chromothrip- sis (or chromosome shattering) [10]. Cancer cells may also acquire DNA from viruses such as Epstein–Barr virus, hepatitis B virus, and hu- man papillomavirus. In defining its guidelines, the International Cancer Genome Consortium established recommendations that will lead to all forms of mutations affecting DNA, RNA, and methylation being de- tected and catalogued. In practice, this usually requires the generation of different laboratory protocols for each mutation category, followed by parallel detection of normal and mutated constituents using high- throughput detection technologies [5–7].
A second reason for the biologi- cal complexity is the large number of mutations observed in every tumour. This is not unexpected given the fre- quent observations of chromosome anomalies revealed by cytogenetic
Fig. 3.1.1. Growth in the number of tumour genome analyses undertaken by International Cancer Genome Consortium members.
186
studies over many decades. The mu- tation spectra in cancer cells were further elucidated by whole cancer genome sequencing, which revealed an extensive number of putative so- matic mutations in every cancer ge- nome; for example, 33 345 somatic base substitutions were observed in the first melanoma genome and 22 910 in the first small cell lung cancer genome [11,12]. These large numbers of somatic mutations usu- ally result from repeated exposure to carcinogens (e.g. tobacco smoke or ultraviolet radiation) and defec- tive DNA repair pathways found in cancer cells [1], and thus occur ran- domly across the genome.
Some mutations occur in se- quences that are functional and af- fect cellular mechanisms that pro- vide growth advantages to cells, such as oncogenic mutations or inactivating mutations involving tumour suppressors. These are of- ten referred to as driver mutations, while the remaining majority of so- matic mutations with no biological consequence are called passenger mutations [5]. The large numbers of mutations also indicate the long
history between initiation of cancer and clinical manifestations for most cancers. There are exceptions to this, as some cancer types (usually paediatric) have low mutation rates. For example, rhabdoid tumours are highly aggressive cancers of
early childhood with extremely low numbers of mutations, including a characteristic driver mutation in the SMARCB1 gene that is observed in all cases [13].
Heterogeneity of driver mutations among patients with similar forms
Fig. 3.1.2. National locations of International Cancer Genome Consortium projects as of January 2013.
Fig. 3.1.3. Numbers of tumours, as indicated by numbers of donors, for each tumour type to undergo comprehensive genome analyses by International Cancer Genome Consortium members.
Chapter 3.1 • Genomics 187
3 B
IO LO
G Y
C H
A P
T E
R 3
.1
Fig. B3.1.1. Frequency of TP53 mutations in various cancer types (histogram) and localization of mutation hotspots in the DNA-binding domain of the p53 protein (three-dimensional structure).
TP53 mutations and human cancer Magali Olivier
The tumour suppressor gene TP53 has been recognized as an impor- tant anticancer gene for more than 20 years because of its frequent alteration in most human cancers. TP53 encodes a protein of 393 amino acids, p53, which plays an important role in the maintenance of cell integrity by responding to vari- ous forms of cellular stresses and acting as a multifunctional protein through the control of various cel- lular pathways. p53 has been shown to repress proliferative signalling, enhance the effects of growth sup- pressors, sensitize cells to apoptosis and autophagy, suppress replicative immortality through senescence, promote genetic and genomic sta- bility, control inflammation, exert anti-angiogenic effects, and repress metastasis, thus counteracting a large panel of tumour-promoting cellular activities [1]. Because of this large spectrum of anti-tumour
functions, inactivating p53 function is an essential step during cancer development, and this is most fre- quently achieved by mutating the TP53 gene. TP53 mutations are found in all cancer types with vari- ous frequencies, ranging from 5% to 90% (Fig. B3.1.1). Recent whole- genome and exome sequencing studies have confirmed that TP53 is the most frequently mutated gene in most cancers.
IARC maintains a database that documents all TP53 gene variations reported in the scientific literature (http://p53.iarc.fr). More than 30 000 somatic mutations are compiled, with annotations on tumour phe- notype, patient characteristics, and structural and functional impacts of mutations [2]. Most TP53 mutations are single-amino-acid substitutions that are located in the DNA-binding domain of p53 and disrupt the tran- scriptional activities of p53. Due to
the three-dimensional structure of the p53 DNA-binding domain and its crucial function, most single-amino- acid substitutions would disrupt p53 function. As a result, mutations ob- served in cancers are very diverse and are scattered along the entire coding sequence, although some mutation hotspots are observed. Some hotspots, such as codons 175, 248, 273, 220, and 245, are found in all cancer types, while others are cancer-specific (such as codon 249 in liver cancer, or codons 280 and 285 in bladder cancer) (Fig. B3.1.1).
TP53 mutations may be used as diagnostic and prognostic markers in the clinic, and several drugs that target the p53 pathway are being de- veloped [1]. TP53 mutations are also informative markers for carcinogen exposure. Indeed, specific TP53 mu- tation patterns have been described in relation with exposure to ultravio- let radiation in skin cancer, tobacco smoke in lung cancer, aflatoxins in liver cancer, or aristolochic acid in upper urinary tract cancers [3]. TP53 mutations can thus be useful mark- ers in molecular epidemiology stud- ies to investigate tumour etiology.
References 1. Hainaut P et al. (2013). TP53 somatic mu-
tations: prognostic and predictive value in human cancers. p53 in the Clinics. New York: Springer.
2. Petitjean A et al. (2007). Hum Mutat, 28:622–629. http://dx.doi.org/10.1002/ humu.20495 PMID:17311302
3. Olivier M et al. (2010). Cold Spring Harb Perspect Biol, 2:a001008. http:// dx.doi.org/10.1101/cshperspect.a001008 PMID:20182602
Website IARC TP53 Database: http://p53.iarc.fr
188
of cancer is an important source of complexity. In a survey of 100 breast tumours analysed for somatic muta- tions and copy number alterations, both gains and deletions, in approxi- mately 21 500 protein-coding genes, also called an exome, 40 genes were implicated as driver mutations across the set, of which 7 were seen in more than 10% of cases and 33 were infrequent [14]. While some cells had more than one recognized driver mutation (the maximum seen in this study was six), 28 cases had only one. This study illustrates that most breast cancers differ from one another in the number of mutations and the set of driver mutations.
Another source of complexity is due to tumour masses usually be- ing composed of different cancer cell subpopulations carrying related but distinct mutation profiles [15]. Over time and across metastatic lesions of a single patient, muta- tion spectra evolve. Although not surprising given the high mutation rates in patients, this evolution has many consequences, of which the most serious is the acquisition of drug resistance, either through new mutations in genes that alter critical drug receptors or pharmacogenomic pathways or through the ability of subclones that are insensitive to drugs to outcompete other clones that are not, after therapy. Thus, the dynamic evolution of cancer genomes is emerging as a critical factor in explaining disease relapse and, ultimately, cancer deaths.
Opportunities for research on etiology The first melanoma and small cell lung cancer genomes [11,12] have revealed that genome-wide mutation patterns, or “mutation signatures”, reflect exposure to environmental carcinogens (ultraviolet radiation, tobacco smoke). The mutation sig- natures found genome-wide in a sin- gle tumour were amazingly similar to those described in a single gene, TP53 (see “TP53 mutations and hu- man cancer”). Similar striking muta- tion signatures have been reported
in upper urinary tract cancers linked to exposure to aristolochic acid [16]. More systematic research on mutation signatures across cancer types has revealed recurrent muta- tion signatures in cancer genomes [17]. Through correlation of muta- tion signatures and exposures, the underlying mutagenic mechanisms causing some of these signatures could be deduced from previous knowledge (smoking, ultraviolet ra- diation, age, BRCA1/2 germline mu- tations). However, many signatures remain unexplained and have led to new hypotheses about mutagenic mechanisms. For example, some signatures have been attributed to the APOBEC family of cytidine deaminases. As these enzymes are involved in the innate immune response to viruses and retrotrans- posons, the presence of these sig- natures might reflect an enhanced activity of APOBEC family members caused by a viral infection. These results may thus stimulate research on the potential role of viral infection in the development of new cancer types [18].
Research into associations be- tween mutation signatures and ex- posures, through genomic epidemi- ology studies and model systems of induced mutagenesis, has great potential to uncover new causal links between the mutagenic processes and the influence of environmental
exposures. For these new research opportunities to have worldwide benefits, it is important for such genomic epidemiology studies to consider geographical heterogeneity in etiologies and to focus on expo- sures and cancer types relevant to low- and middle-income countries.
Technological challenges Although genomic technologies are not discussed in detail in this Report, this topic is relevant to data generators, users of data available in International Cancer Genome Consortium and other databases, and future adoptees of genomic technologies for clinical applica- tions. Given the recent development of most of these technologies, as well as the associated computa- tional challenges, it is recognized that each data set contains false- positive results and is incomplete (i.e. contains false negatives). To facilitate the task of future users of Consortium data sets, various quali- ty control metrics, including second- ary validation of potentially deleteri- ous mutations, are conducted and reported. It is thus essential for ex- perts and non-experts to be aware that the technology has limitations and that potentially important muta- tions of clinical relevance require independent validation.
Fig. 3.1.4. Growth in the number of cancer genome data sets available via the International Cancer Genome Consortium Data Portal (http://dcc.icgc.org). Each colour represents a different tumour subtype.
Chapter 3.1 • Genomics 189
3 B
IO LO
G Y
C H
A P
T E
R 3
.1
The rapid expansion of cancer genome data sets being generated is creating formidable challenges [19]. For each tumour, numerous types of primary data are collect- ed, including clinical information, pathological data, DNA sequences (from normal and tumour tissues for the detection of several mutation categories, including simple base substitutions, copy number chang- es, and other structural alterations), methylation, and RNA expression. In addition, interpreted data such as functional consequences of mutations to protein structure and potential clinical correlates also need to be recorded in robust and user-friendly databases. As these individual-specific data sets be- come integrated at the project level and ultimately analysed across all tumour types, the computational challenges are daunting. Similarly, how this information is effectively disseminated to the wider scien- tific community (via descriptive databases or through transfer and analysis of the large amounts of
Table 3.1.1. Internet resources for cancer genomics data
Resource Website Description
International Cancer Genome Consortium (ICGC)
http://dcc.icgc.org/ The ICGC Data Portal provides access to cancer genome data and project data from ICGC members.
The Cancer Genome Atlas (TCGA) https://tcga-data.nci.nih.gov/tcga/ tcgaHome2.jsp
The TCGA Data Portal provides a platform for researchers to search, download, and analyse cancer genome data sets generated by institutions in the USA contributing to TCGA.
Catalogue of Somatic Mutations in Cancer (COSMIC)
http://www.sanger.ac.uk/genetics/CGP/ cosmic
COSMIC stores and displays curated somatic mutation data and other information related to human cancers.
Broad Institute Integrative Genomics Viewer (IGV)
http://www.broadinstitute.org/igv/ IGV is a high-performance visualization tool for interactive exploration of large, integrated genomic data sets.
University of California Santa Cruz (USCS) Cancer Genomics Browser
https://genome-cancer.soe.ucsc.edu The UCSC Cancer Genomics Browser is a suite of web-based tools to visualize, integrate, and analyse cancer genomics and its associated clinical data.
United States National Cancer Institute’s Therapeutically Applicable Research to Generate Effective Treatments (TARGET)
http://target.nci.nih.gov/dataMatrix/ TARGET_DataMatrix.html
TARGET data sets include genomic characterization, full clinical information, and sequencing files for five common types of childhood malignancies: acute lymphoblastic leukaemia, acute myeloid leukaemia, neuroblastoma, osteosarcoma, and Wilms tumour.
Fig. 3.1.5. Figurative depiction of somatic mutations present in a single cancer genome, part of a catalogue of somatic mutations in the small cell lung cancer cell line NCI-H2171. Individual chromosomes are depicted on the outer circle, followed by concentric tracks for point mutation, copy number, and rearrangement data relative to mapping position in the genome. Arrows indicate examples of the various types of somatic mutation present in this cancer genome.
190
raw data) also presents formidable challenges.
International issues The most intense cancer genome re- search was initially concentrated in developed nations with a track record in large-scale genome research. When the International Cancer Genome Consortium was launched, a remarkable effort was made by funding agencies in many countries to participate, not only as a way of ensuring that cancers common in their respective populations were in- vestigated but also to build capacity in next-generation sequencing tech- nologies and allow their communities of scientists to be actively involved in cutting-edge research. For exam- ple, India established the National Institute of Biomedical Genomics in Kalyani, West Bengal, to conduct a Consortium project to study gingi- vo-buccal cancer. This form of oral cancer, which accounts for 30% of all cancers in India [20], is due to the habit of chewing tobacco with betel quid.
At the halfway mark in the Consortium’s decade-long initiative, many types of cancer across the
globe are being studied, but there are still obstacles (economic bar- riers, limited infrastructures for biobanking, paucity of skills, etc.) in analysing some forms of cancer that are common in some low- and middle-income countries (notably in Africa). Several approaches that may be considered to resolve this include collaborations between cancer care workers in these countries and international lead- ers in cancer research (including Consortium members), leveraging the expertise already established in these countries for conducting clinical trials for infectious diseas- es, and establishing a distributed model for tumour banking driven by standard operating procedures.
Legal and regulatory frame- works that differ among internation- al jurisdictions have the potential to hinder the exchange of data among collaborating groups in different countries and also to limit the shar- ing of data via the Internet. Some valid reasons for placing restrictions on data include protection of pri- vacy and the potential that misuse could harm individuals. For these reasons, the International Cancer Genome Consortium implemented
robust governance procedures for handling data from human subjects, including the adoption of common bioethical elements and harmonized consent procedures [8]. Moreover, the Consortium structured its databases to classify data that can- not be used to identify individuals as “open”, or unrestricted and publicly accessible. In parallel to this, the Consortium created mechanisms to classify more sensitive genetic and clinical information that could be misused by third parties to re- identify participants as “controlled”, and available only to scientists and research institutions agreeing to conditions stipulated in an access agreement. Oversight of this pro- cess is conducted by an interna- tional committee [8,21].
Conclusions and next steps Cancer genomics is a research field that is growing rapidly as a result of new genomic technologies. Given the fundamental role that mutations play in the development of cancer, cancer genomes will continue to be analysed and mined for additional clues that will increase our under- standing of cancer cells, for biomark- ers that will be used as indicators of prognosis and treatment response, and for targets that will be evaluated as substrates for new cancer drugs. The impact of this progress is appar- ent not only in the growing number of targeted agents being developed by biotechnology and pharmaceutical companies but also in the reduced time between identification of driver mutations and approval of a specific targeted therapy by regulatory agen- cies [7].
The diversity of mutational pro- cesses underlying development of cancer is evident from an analysis of 4 938 362 mutations from 7042 can- cers. The prevalence of mutations varied over more than 5 orders of magnitude; childhood cancers car- ried the fewest, and cancers related to chronic mutagenic exposures, such as tobacco smoke or ultraviolet ra- diation, exhibited the most (> 400 per megabase of DNA). Mutations were
Fig. 3.1.6. Next-generation gene sequencing equipment in use at the Joint Genome Institute in Walnut Creek, California. Large-scale cancer genome studies are already applying next-generation sequencing technologies to tumours from 50 different cancer types to generate more than 25 000 cancer genomes.
Chapter 3.1 • Genomics 191
3 B
IO LO
G Y
C H
A P
T E
R 3
.1
categorized on the basis of 96 pos- sible scenarios, and 21 distinct mu- tational signatures were defined in relation to 30 cancer classes [22].
It is expected that the growth in cancer genome information will continue exponentially, from thou- sands to millions of tumours in the next decade. Most of this growth is happening as a consequence of new sequencing technologies that are be- coming easier to adapt to clinical lab- oratory environments. The feasibility
of cancer clinical sequencing that is used in selecting therapies has been demonstrated in the clinical trials setting [23]. The challenges in converting cancer genome analyses into improved diagnosis, prognosis, and treatment are formidable, but the impact will be felt increasingly by health-care systems, clinicians, and patients [24].
Recently, the Cancer Genome Atlas Research Network docu- mented point mutations and small
insertions or deletions from 3281 tumours across 12 tumour types, including a clinical association anal- ysis. Overall, mutations in several genes, including BAP1, DNMT3A, KDM5C, FBXW7, and TP53, corre- lated with poor prognosis, whereas mutations in two genes, BRCA2 and IDH1, often correlated with improved prognosis [25].
192
References
1. Hanahan D, Weinberg RA (2000). The hallmarks of cancer. Cell, 100:57–70. http: //dx.doi.org /10.1016/S 0 0 92- 8 674(0 0) 81683-9 PMID:10647931
2. Hanahan D, Weinberg RA (2011). Hallmarks of cancer: the next generation. Cell, 144: 646–674. http://dx.doi.org/10.1016/j.cell. 2011.02.013 PMID:21376230
3. Lander ES, Linton LM, Birren B et al.; International Human Genome Sequencing Consortium (2001). Initial sequencing and analysis of the human genome. Nature, 409:860–921. http://dx.doi.org/10.1038/35 057062 PMID:11237011
4. Futreal PA, Kasprzyk A, Birney E et al. (2001). Cancer and genomics. Nature, 409:850–852. http://dx.doi.org/10.1038/35 057046 PMID:11237008
5. Stratton MR, Campbell PJ, Futreal PA (2009). The cancer genome. Nature, 458:719– 724. http://dx.doi.org/10.1038/nature07943 PMID:19360079
6. Wong KM, Hudson TJ, McPherson JD (2011). Unraveling the genetics of cancer: genome sequencing and beyond. Annu Rev Genomics Hum Genet, 12:407–430. http://dx.doi.org/10.1146/annurev-genom- 082509-141532 PMID:21639794
7. Chin L, Andersen JN, Futreal PA (2011). Cancer genomics: from discovery sci- ence to personalized medicine. Nat Med, 17:297–303. http://dx.doi.org/10.1038/nm. 2323 PMID:21383744
8. Hudson TJ, Anderson W, Artez A et al.; International Cancer Genome Consortium (2010). International network of cancer genome projects. Nature, 464:993–998. http://dx.doi.org/10.1038/nature08987 PMID:20393554
9. Jennings J, Hudson TJ (2013). Reflections on the founding of the International Cancer Genome Consortium. Clin Chem, 59:18–21. http://dx.doi.org/10.1373/clinchem.2012. 184713 PMID:23136248
10. Maher CA, Wilson RK (2012). Chromothripsis and human disease: piecing together the shattering process. Cell, 148:29–32. ht t p: //d x.d o i.o r g /10.1016 / j.c e ll. 2 012 . 01.006 PMID:22265399
11. Pleasance ED, Cheetham RK, Stephens PJ et al. (2010). A comprehensive catalogue of somatic mutations from a human cancer genome. Nature, 463:191–196. http://dx.doi. org/10.1038/nature08658 PMID:20016485
12. Pleasance ED, Stephens PJ, O’Meara S et al. (2010). A small-cell lung cancer genome with complex signatures of tobacco expo- sure. Nature, 463:184–190. http://dx.doi. org/10.1038/nature08629 PMID:20016488
13. Lee RS, Stewart C, Carter SL et al. (2012). A remarkably simple genome underlies highly malignant pediatric rhabdoid cancers. J Clin Invest, 122:2983–2988. http://dx.doi. org/10.1172/JCI64400 PMID:22797305
14. Stephens PJ, Tarpey PS, Davies H et al.; Oslo Breast Cancer Consortium (OSBREAC) (2012). The landscape of cancer genes and mutational processes in breast cancer. Nature, 486:400– 404. ht tp://dx.doi.org/10.1038/nature11017 PMID:22722201
15. Samuel N, Hudson TJ (2013). Translating genomics to the clinic: implications of cancer heterogeneity. Clin Chem, 59:127–137. http:// dx.doi.org/10.1373/clinchem.2012.184580 PMID:23151419
16. Hoang ML, Chen CH, Sidorenko VS et al. (2013). Mutational signature of aris- tolochic acid exposure as revealed by whole-exome sequencing. Sci Transl Med, 5:197ra102. http://dx.doi.org/10.1126/ scitranslmed.3006200 PMID:23926200
17. Alexandrov LB, Nik-Zainal S, Wedge DC et al. (2013). Signatures of mutational pro- cesses in human cancer. Nature, 500:415– 421. http://dx.doi.org/10.1038/nature12477 PMID:23945592
18. Roberts SA, Lawrence MS, Klimczak LJ et al. (2013). An APOBEC cytidine deaminase mutagenesis pattern is widespread in hu- man cancers. Nat Genet, 45:970–976. http:// dx.doi.org/10.1038/ng.2702 PMID:23852170
19. Chin L, Hahn WC, Getz G, Meyerson M (2011). Making sense of cancer genomic data. Genes Dev, 25:534–555. http://dx.doi. org/10.1101/gad.2017311 PMID:21406553
20. Khan Z (2012). An overview of oral cancer in Indian subcontinent and recommenda- tions to decrease its incidence. Cancer, 3:WMC003626.
21. Joly Y, Dove ES, Knoppers BM et al. (2012). Data sharing in the post-genomic world: the experience of the International Cancer Genome Consortium (ICGC) Data Access Compliance Office (DACO). PLoS Comput Biol, 8:e1002549. http://dx.doi.org/10.1371/ journal.pcbi.1002549 PMID:22807659
22. Alexandrov LB, Nik-Zainal S, Wedge DC et al.; Australian Pancreatic Cancer Genome Initiative; ICGC Breast Cancer Consortium; ICGC MMML-Seq Consortium; ICGC PedBrain (2013). Signatures of mutation- al processes in human cancer. Nature, 500:415–412. http://dx.doi.org/10.1038/ nature12477 PMID:23945592
23. Tran B, Brown AM, Bedard PL et al. (2013). Feasibility of real time next generation se- quencing of cancer genes linked to drug response: results from a clinical trial. Int J Cancer, 132:1547–1555. http://dx.doi.org/10. 1002/ijc.27817 PMID:22948899
24. Dancey JE, Bedard PL, Onetto N, Hudson TJ (2012). The genetic basis for cancer treatment decisions. Cell, 148:409–420. http://dx.doi.org/10.1016/j.cell.2012.01.014 PMID:22304912
25. Kandoth C, McLellan MD, Vandin F et al. (2013). Mutational landscape and sig- nificance across 12 major cancer types. Nature, 502:333–339. http://dx.doi.org/10. 1038/nature12634
Website International Cancer Genome Consortium: www.icgc.org
Chapter 3.2 • Genome-wide association studies 193
3 B
IO LO
G Y
C H
A P
T E
R 3
.2Summary • Cancer genome-wide associa-
tion studies have successfully identified many new susceptibil- ity alleles in the human genome, most of which are unique to each type of cancer.
• Cancer genome-wide asso- ciation studies will continue to discover an increasing fraction of the common susceptibility al- leles, and have uncovered dis- tinct genetic architectures for specific cancer susceptibilities.
• Findings from cancer genome- wide association studies have provided new mechanistic in- sights into cancer etiology, in- cluding changes in the regula- tion of key genes and pathways. Investigations of the relationship between germline susceptibility alleles and somatic alterations should uncover new pathways and targets for therapeutic and preventive measures.
• Clinical cancer genetics will profit from the determination of a more comprehensive catalogue of susceptibility alleles, across a spectrum of frequencies and ef- fect sizes, which could be imple- mented in precision medicine.
3.2 3 BIOLOGY
Genome-wide association studies Stephen J. Chanock Dongxin Lin (reviewer)
Paul Pharoah (reviewer)
For nearly half a century, the herita- ble contribution to cancer has been investigated, beginning with studies of families in which multiple mem- bers developed the same type of cancer [1]. As the draft sequence of the human genome was completed [2,3], its annotation revealed a wide spectrum of genetic variation, from single base changes to large struc- tural alterations and copy number variations [4]. Traditionally, investiga- tors have focused on the single base substitution, known as the single- nucleotide polymorphism (SNP); the frequency of the alternative nu- cleotide is captured by the minor al- lele frequency (MAF). Differences in human populations are etched in the patterns of genetic variation; this includes both the correlation between nearby variants, known as linkage disequilibrium, and the ac- tual frequencies of common variants, measured by the MAF. In turn, these differences have become attractive for investigating differences in inci- dence for distinct cancers, by either population or exposure.
Cancer susceptibility alleles can be discovered by different approaches, such as linkage and association analyses; the tools have improved substantively, shifting from genotyping of sin- gle tandem repeats and SNPs to whole-genome sequence analysis
using next-generation sequencing technologies. Not all alleles have comparable estimated effects (Fig. 3.2.1) [5]. Linkage analyses in family studies are used to discover highly penetrant mutations, as in BRCA1 or TP53, which are rare but have a strong predictive value. Due to the limitations in available tools, only a small fraction of pedigrees have been explained by mutations in “cancer susceptibility genes”. These have provided a foundation for clinical cancer genetics.
Common susceptibility alleles are discovered by association stud- ies, which compare allele frequen- cies between affected and unaf- fected individuals (Fig. 3.2.2). The estimated effect sizes are smaller for common variants and are neither necessary nor sufficient for cancer susceptibility (Fig. 3.2.1). Moreover, common variants contribute to a polygenic model for susceptibil- ity, akin to complex diseases like diabetes and neurodegenerative disorders.
In 1996, Risch and Merikangas argued that linkage analysis for com- plex diseases would be less efficient than association analyses in popula- tions for mapping common variants with smaller effect sizes [6]. The field turned towards association studies, initially using the candidate gene approach, in which investigators
194
chose favourite genes or variants for testing, but with little success. Generally, reported findings failed to be replicated in subsequent stud- ies, for a variety of reasons that included issues in study design, small sample sizes, and unrealistic expectations for effectively choos- ing functional variants. The candi- date gene approach yielded less than a dozen conclusive variants that confer susceptibility to different cancers. Notable examples include the association of bladder cancer with NAT2 variants or a common deletion of GSTM1 [7] as well as of aerodigestive cancers with coding variants in alcohol dehydrogenase genes (ADH1B and ADH7) [8]. Still, what emerged was an appreciation of the value of robust replication to- gether with the scaling of studies to a size that would provide adequate power to conclusively detect variants on statistical grounds.
Principles of cancer GWAS The advent of genome-wide asso- ciation studies (GWAS) has substan- tially accelerated discovery of com- mon genetic susceptibility variants for a wide range of human diseases and traits. Advances in microar- ray technologies have enabled re- searchers to interrogate hundreds of thousands of SNPs in parallel using new analytical tools and standards. The underlying approach has been to conduct a statistical – or “ag- nostic” – analysis of commercial SNP microarrays designed to tag common variants across the entire genome. Consequently, GWAS find markers, or surrogates of suscepti- bility alleles, and rarely determine the “functional” variant, which can explain the association.
Many statistical tests are con- ducted, because the primary ap- proach has been to identify common SNPs with a main effect using the trend test, thus raising the spectre of
false positives. The community has embraced a threshold of genome- wide significance for reporting GWAS results, defined as a trend as- sociation test with P ≤ 5 × 10 –8 after adjustment as per the GWAS study design. Because of the small effect sizes of most common SNPs conclu- sively associated with cancer risk, it is unlikely that the alleles incor- porate a dominant effect. Follow-up studies or large meta-analyses are required to establish a conclusive finding [9]. Independent replication guards against the pursuit of false positives, which is important since the downstream cost of mapping and laboratory investigation is high with respect to time and resources [10]. The actual functional marker does not have to be tested; instead, a surrogate can be in linkage dis- equilibrium and thus replicated in follow-up [11]. Occasionally, a com- mon genetic marker may point to- wards a less common variant with a
Fig. 3.2.1. Mapping genetic susceptibility. Distribution of susceptibility alleles by frequency and strength of genetic effect, illustrat- ing the distribution of susceptibility alleles as well as the feasibility of identifying variants through GWAS and sequence analysis.
Chapter 3.2 • Genome-wide association studies 195
3 B
IO LO
G Y
C H
A P
T E
R 3
.2
stronger effect, known as a synthetic association. However, most GWAS signals map to common variants with direct functional effects on cancer susceptibility.
GWAS are scalable for discov- ery. For example, the Collabora- tive Oncological Gene-environment Study has pooled existing scans and conducted large-scale replica- tion to discover new loci in three hormone-related cancers: breast, ovarian, and prostate cancers [12– 14]. These studies have identified a
larger fraction of the common vari- ants that contribute to the compre- hensive panel of independent loci. Since GWAS genotyping has been performed with different commercial and custom SNP microarrays, tech- niques for imputation of data have been developed to combine data sets. Several imputation programmes successfully infer untested and highly correlated SNPs based on reference data sets, such as the International HapMap Project, the 1000 Genome Project, or the Division of Cancer
Epidemiology and Genetics (United States National Cancer Institute) im- putation set [4,15,16].
Cancer GWAS discoveries In the past 6 years, the pace of GWAS findings has accelerated (Box 3.2.1). Since the first cancer GWAS were published in 2007, nearly 400 distinct genetic loci have been con- clusively identified for more than two dozen different cancers – including common cancers, such as breast, colon, and prostate cancers, as well as rarer paediatric cancers, like Ewing sarcoma and neuroblastoma [17–20]. Moreover, GWAS have iden- tified susceptibility loci for specific subtypes of cancers, such as estro- gen-negative breast cancer [21].
Reported cancer GWAS find- ings are associated with cancer susceptibility. With rare exceptions, the etiological markers are not asso- ciated with clinical outcomes, includ- ing metastatic disease or survival. Several neuroblastoma loci discrimi- nate between aggressive and milder disease. Of the more than 75 inde- pendent loci identified for prostate cancer, not one accurately discrimi- nates between aggressive and non- aggressive disease. Recently, vari- ants in SLC39A6 were conclusively associated with survival in individu- als with oesophageal squamous cell carcinoma [22]. Still, the lack of correlation between etiological loci and clinical outcomes suggests that distinct regions of the genome may contribute to the development of cancer but not necessarily the progression of cancer.
Fig. 3.2.2. Genetic analysis of GWAS. Multiple steps are conducted, including the choice of SNPs across the genome (usually included on a commercial SNP microarray) based on linkage disequilibrium in a region, enabling the selection of a surrogate to test for the region. Association analysis is conducted in a case–control setting, examining all SNPs in a Manhattan plot, followed by replication analyses that pinpoint markers on chromosomes, which are fine-mapped and investigated in the laboratory. MAF, minor allele frequency.
1. Discovery of new regions in the genome associated with diseases/traits • New candidate genes and
regions
2. Clues for mechanistic insights into the contribution of com- mon genetic variation to can- cer biology • Etiology
• Gene–environment/lifestyle interactions
• Outcomes and pharmacogenomics
3. Challenge of genetic markers for risk prediction for individ- ual or public health decisions • Common variants represent a
fraction of the genetic contri- bution to risk
• Polygenic risk models
Box 3.2.1. Current status of genome-wide association studies.
196
Gene–environment interactions and breast cancer Federico Canzian
GWAS have recently identified sev- eral genetic loci associated with risk of breast cancer. They are the latest addition to a list of established breast cancer risk factors, which are mainly environmental or lifestyle-related.
Exploring the possible interplay between genetic variants and estab- lished breast cancer risk factors can (i) provide hints on the function of the novel genetic loci, and in gen- eral on biological mechanisms of breast cancer etiology, and (ii) help in formulating better risk prediction models, with a potential impact on prevention.
Given the small risk modification conferred by each environmental and genetic risk factor, large-scale association studies, involving many thousands of cases and controls, are needed to explore gene–envi- ronment interactions. Several such studies have recently been pub- lished, on up to 26 000 breast can- cer cases and 32 000 controls [e.g. 1–3]. They involved between 7 and 17 SNPs and established risk factors including age at menarche, parity, age at first live birth, breastfeeding, menopausal status, age at meno- pause, use of hormone replacement therapy, body mass index, height,
tobacco use, alcohol consumption, number of previous breast biopsies, and family history of breast cancer. Both prospective cohorts [1,2] and retrospective series of breast cancer cases and controls [3] were used. Most of the study subjects were Caucasians, with some studies done in Asian populations.
Results obtained so far [1–3] suggest that two-way interactions between known risk SNPs and es- tablished breast cancer risk factors are unlikely to be strong. None of the large-scale studies reported sta- tistically significant gene–environ- ment interactions, after correcting for multiple testing. These studies were well powered, having greater than 90% power to detect interaction odds ratios as low as 1.06. Higher- order interactions were not tested as they would require much larger sample sizes.
Previous reports suggested the existence of a few statistically sig- nificant gene–environment interac- tions, for example between SNPs in FGFR2 and use of hormone re- placement therapy, family history of breast cancer, age at menarche, and parity. None of these associa- tions were observed in the larger
studies. Considering the relatively small sample size and the modest statistical significance of the interac- tions reported in the original studies, the most likely explanation is that these could be chance findings.
New GWAS on breast cancer risk, and meta-analyses of exist- ing ones, are still being performed; therefore, more breast cancer susceptibility loci will be known in the near future. Consequently, new gene–environment interaction stud- ies will be needed. Furthermore, all studies published so far assumed a multiplicative model of interaction. It remains to be seen whether signifi- cant gene–environment interactions can be detected by exploring differ- ent models, such as supra-additive interactions.
References 1. Travis RC et al. (2010). Lancet, 375:2143–
2151. http://dx.doi.org/10.1016/S0140-67 36(10)60636-8 PMID:20605201
2. Campa D et al. (2011). J Natl Cancer Inst, 103:1252–1263. http://dx.doi.org/10.1093/ jnci/djr265 PMID:21791674
3. Milne RL et al. (2010). Breast Cancer Res, 12:R110. http://dx.doi.org/10.1186/bcr2797 PMID:21194473
Nearly all cancer GWAS markers have a MAF greater than 10%, with a handful in the 5–10% range [23]. This is a consequence of the design of the first-generation SNP micro- array chips and the smaller sample sizes used. With larger meta-analy- ses and follow-up studies planned, further discovery in the MAF space between 2% and 10% will occur. An emerging concept is that there could be major differences in the underly- ing genetic architecture of different types of cancer [24]. For example, the range of effect sizes and allele frequencies varies between prostate cancer and breast cancer, which in- cludes BRCA1, BRCA2, TP53, and a
set of moderately penetrant genetic mutations.
In the first wave of cancer GWAS, the per-allele estimated effect sizes were small, with odds ratios between 1.1 and 1.4. In paediatric cancer GWAS, estimates of 1.6–1.8 are not unusual, perhaps suggesting stronger drivers of cancer devel- opment. One notable exception for young adults is in testicular cancer GWAS, for which the per-allele ef- fect estimate is greater than 2.5 for KITLG on chromosome 12 [25]. This is not surprising since testicu- lar cancer has a high heritability in family studies. It is notable that as sample sizes increase, it is possible to find smaller estimated effect sizes.
The Collaborative Oncological Gene- environment Study consortium has discovered new breast and pros- tate cancer signals but with effect sizes of 1.05–1.15 (see “Gene– environment interactions and breast cancer”).
Few of the cancer GWAS sig- nals map to a coding change in a plausible candidate gene. Most signals map to non-coding regions. Approximately one quarter map to intergenic regions, in which there are no adjacent correlated markers that map to characterized genes. Together, these findings suggest that the majority of the common genetic variants contribute to cancer through
Chapter 3.2 • Genome-wide association studies 197
3 B
IO LO
G Y
C H
A P
T E
R 3
.2
perturbations of the regulation of known or novel pathways.
Most cancer GWAS signals are unique to a cancer type, but there are a few highly informative regions that harbour susceptibility alleles for multiple cancers, thus linking distinct cancers and suggesting shared pathways. A region flanking the MYC oncogene on 8q24 harbours at least five independent loci associated with prostate cancer as well as loci as- sociated with four other cancers (breast, colon, and bladder cancers and chronic lymphocytic leukaemia) [23]. GWAS have detected strong signals for the human leukocyte an- tigen (HLA) regions for cancers of the immune system or those driven by viral infections (e.g. cervical, liver, and nasopharyngeal cancers) (see “DNA repair polymorphisms and hu- man cancer”).
A region on 5p15.33 contain- ing the telomerase gene, TERT, harbours multiple susceptibility al- leles for many cancers, including rare and common SNP alleles [26]. Rare mutations in TERT track with dyskeratosis congenita (an inher- ited bone-marrow failure syndrome) and idiopathic pulmonary fibrosis. Cancer GWAS have discovered susceptibility alleles for at least 10 different types over at least five independent signals. Moreover, a protective allele for one cancer can confer susceptibility to another cancer; notably, the same allele has inverse effects for two skin cancers, basal cell carcinoma and melanoma. The pleiotropy in this region hints at complex gene–gene or gene– environment interactions.
Whereas most of the initial GWAS were conducted in subjects of European ancestry, recently more have been conducted in subjects of Asian and African ancestry. Prostate cancer GWAS in men of African an- cestry identified new independent loci on 8q24 as well as a new re- gion on 17q21 [27]. Studies in Asia have concentrated on cancers nota- ble for high incidence, such as gas- tric adenocarcinoma, oesophageal squamous cell carcinoma, and lung cancer in nonsmokers [28,29].
The underlying population genetic histories of distinct groups have im- portant implications for the discovery of susceptibility alleles because the underlying linkage disequilibrium can vary greatly. In many regions of the genome, differences in MAFs between continental populations (e.g. European, Asian, and African descent) can account for variance in the detection of susceptibility alleles. For example, the MAF of a SNP is sufficiently high in men of African ancestry for the SNP to be detected in a GWAS of prostate cancer in African-American men, but because of its low MAF in Europeans, it is not easily detected in GWAS of European men [27].
Until recently, cancer GWAS lumped disparate study designs to- gether and rarely addressed envi- ronmental exposures. Still, a hand- ful of alleles have been shown to interact with specific exposures. In a study of oesophageal squamous cell carcinoma in China, a gene– environment interaction was ob- served between the variants in the ADH/ALDH genes and alcohol con- sumption [30]. In bladder cancer, a novel marker for NAT2 interacts with smoking, correlating with low or intermediate acetylation status. In fact, common genetic variants have been shown to modify the ef- fect of smoking on bladder cancer [31]. A GWAS of lung cancer in non- smoking Asian women discovered regions not observed in lung can- cer GWAS dominated by smokers [29]. Also, there was no signal in nonsmoking women across 15q24, a region strongly correlated with smoking behaviour and lung cancer in smokers, suggesting that the re- ported signals are primarily related to tobacco use.
Analysis of associated risk factors collected in cancer GWAS has been extremely informative for tobacco use, alcohol consumption, and an- thropometric measures (e.g. height, weight, and waist size). Cross- disease consortia (e.g. diabetes and cardiovascular disease) have com- bined numbers to accelerate discov- ery of common risk factors. Already,
more than 200 independent loci have been identified for height [32].
Investigation of GWAS signals Each region should be fine-mapped to determine optimal variants for functional analyses, mainly because of the large number of highly corre- lated variants. The pattern of linkage disequilibrium, often with apparent differences between ancestral popu- lations, is examined to winnow down possible variants. Choosing the opti- mal ones also requires bioinformat- ics assessment using new public da- tabases, such as the Encyclopedia of DNA Elements (ENCODE) project [33]. ENCODE has begun to system- atically shed light on the biology of the regulation of the genome, spe- cifically cataloguing signposts and markers of biological activity. This resource can be used for individual variants but also to look at patterns or pathways, as recently suggested for breast cancer susceptibility loci identified by GWAS [34]. A greater- than-expected fraction of cancer susceptibility alleles map to regula- tory regions, suggesting that com- mon variants confer susceptibility primarily through perturbations in regulatory events [23].
Laboratory investigation is re- quired to explain the biological basis of the SNPs that directly associate with the cancer of interest. Each indi- vidual variant has to be studied sepa- rately, and a combination of different approaches and tools is required, which explains the markedly slower pace of characterization (Fig. 3.2.3). For example, investigation of a pros- tate cancer GWAS marker, which maps to the β-microseminoprotein gene (MSMB) on 10q11, revealed that the risk allele in the promoter region decreases transcriptional ac- tivity, lowering expression in pros- tate cancer tissue [35]. Expression decreases as prostate cancer pro- gresses from early to late stages; loss of MSMB expression can be associated with disease recurrence after radical prostatectomy. Still, the finding may not be fully explained by
198
DNA repair polymorphisms and human cancer Paolo Vineis
A wide range of DNA damage can be inflicted, both from extracellular agents – including environmental exposures – and via endogenous mechanisms. Genotoxic chemicals bind to DNA, forming adducts that in turn can be repaired by the DNA repair machinery (Fig. B3.2.1) or lead to permanent DNA damage. DNA damage can lead to cancer and other diseases. Therefore, it is plausible that inherited sequence variants in DNA repair genes are involved in cancer development.
Five main mechanisms are in- volved in DNA repair: base excision repair, which corrects non-bulky damage; nucleotide excision repair, which corrects lesions that disrupt the double helical structure of DNA; mismatch repair, which corrects replication errors; double-strand break repair, which corrects dou- ble-strand breaks through two dif- ferent pathways, homologous and
non-homologous recombination; and direct repair, which corrects methylated or alkylated bases [1]. Rare, highly penetrant mutations of DNA repair genes (such as the XP family, involved in the heritable disease xeroderma pigmentosum) are associated with a high risk of cancer, whereas the evidence is weaker for common SNPs with low penetrance.
A systematic review and meta- analysis, based on systematic cri- teria, showed that out of 241 as- sociations investigated, only three associations were graded as having strong epidemiological credibility of cumulative evidence [2]. These associations were between two SNPs – rs1799793 and rs13181 – in the ERCC2 gene and lung cancer (recessive model), and between rs1805794 in the NBN gene and bladder cancer (dominant model). An update of this meta-analysis
has since been performed [3], and the authors found partially incon- sistent results. In addition, none of the cancer GWAS published until 2011 showed highly statistically significant associations for any of the common DNA repair gene vari- ants that would place DNA repair genes among the top 10–20 hits identified in GWAS. This suggests that it is unlikely that DNA repair gene polymorphisms per se play a major role. This is not surprising since in a multistep, multigenic pro- cess such as carcinogenesis, single polymorphisms in single genes are unlikely to alter the expression or function of specific proteins to the extent of producing a pathological phenotype. However, the combined effect of several SNPs in a gene or in multiple genes could have more impact.
Pathway approaches using novel genotyping technologies will en- able more comprehensive studies of multiple SNPs in multiple genes, and it will be possible to investigate gene–environment interactions more rigorously than before, using novel statistical methodology. The com- bined effect of multiple SNPs in sev- eral genes in one or more relevant DNA repair pathways could have a greater impact on pathological phe- notypes than SNPs in single genes, but this has been investigated only occasionally.
References 1. Friedberg E et al. (2006). DNA Repair
and Mutagenesis. Washington, DC: ASM Press.
2. Vineis P et al. (2009). J Natl Cancer Inst, 101:24–36. http://dx.doi.org/10.1093/jnci/ djn437 PMID:19116388
3. Ricceri F et al. (2012). Mutat Res, 736:117– 121. http://dx.doi.org/10.1016/j.mrfmmm. 2011.07.013 PMID:21864546
Fig. B3.2.1. DNA damage, repair mechanisms, and consequences. (A) Common DNA damaging agents (top), examples of DNA lesions induced (middle), and the most relevant DNA repair mechanisms responsible for the removal of those lesions (bottom). (B) Acute effects of DNA damage on cell-cycle progression, leading to transient arrest in the G1, S, G2, and M phases (top), and on DNA metabolism (middle). Long-term consequences of DNA injury (bottom) include point mutations or chromosome aberrations and their biological effects. cis-Pt, cisplatin; CPD, cyclobutane pyrimidine dimer; EJ, end joining; HR, homologous recombination; MMC, mitomycin C; (6-4)PP, 6-4 photoproduct; UV, ultraviolet.
Chapter 3.2 • Genome-wide association studies 199
3 B
IO LO
G Y
C H
A P
T E
R 3
.2
MSMB alone. The gene on the other side of MSMB, NCOA4, is also up- regulated by the GWAS promoter SNP, and NCOA4 forms chimeric transcripts with MSMB [36,37]. Also promising is the relationship between the MSMB allele and uri- nary levels of β-microseminoprotein, which was validated as a serum bio- marker of prostate cancer in previ- ous investigations.
The value of fine-mapping is critical for exploring GWAS markers. Prokunina-Olsson et al. uncovered a novel gene, IFNL4, created by a complex dinucleotide variant (e.g. a single bp missense contiguous to a deletion) in strong linkage disequi- librium with the published marker on chromosome 19 associated with clearance of hepatitis C virus infec- tion, a major risk factor for liver can- cer [38]. Earlier investigators focused on variants upstream of the nearest gene, IFNL3 (previously IL28B). The
deletion allele generates a frameshift, which leads to a novel protein that induces an interferon-type response, apparent on close inspection of RNA sequencing data in primary human hepatocytes. The complex IFNL4 ge- netic variant has a strong effect on hepatitis C virus clearance, especial- ly in individuals of African ancestry. The strength of the estimated effect on both spontaneous and treatment- induced clearance of hepatitis C virus is significantly larger than most can- cer GWAS signals, prompting pursuit in clinical studies.
Understanding the biological basis of a GWAS signal can even- tually lead to clinical translation. A bladder cancer GWAS locus maps to the prostate stem cell antigen gene (PSCA) [39]. Based on RNA sequencing followed by functional analysis, a promoter SNP, charac- terized in fine-mapping, influences messenger RNA (mRNA) PSCA
expression [40]. The creation of an alternative translation start site leads to increased expression of PSCA on the cell surface [41]. The actual genotype could predict PSCA protein expression and identify blad- der cancer patients harbouring the PSCA variant, who may benefit from immunotherapy with anti-PSCA humanized antibody, an emerging therapy for different cancers.
SNP markers have yet to be incorporated into clinical practice, despite the opinions of some com- mercial groups. Adequate testing and validation have not been con- ducted in the clinical setting. Once a catalogue approaching a compre- hensive set of common susceptibility alleles has been defined for a given cancer, it will be possible to explore applications in large validation stud- ies. Already, it is possible to begin to develop the models to design and conduct the validation studies [42].
Fig. 3.2.3. Laboratory investigation of GWAS SNPs. Cartoon depiction of the steps after fine-mapping, beginning with the assess- ment of whether a marker resides in a coding region and continuing through the bioinformatics analysis and assessment of functional elements before the experimental studies are conducted. mRNA, messenger RNA; miRNA, microRNA, RNASeq, RNA sequencing.
200
In a recent analysis of ovarian can- cer, it appears that women harbour- ing germline mutations in BRCA1 or BRCA2 may actually have improved survival, partly due to increased sen- sitivity to standard chemotherapeu- tic agents [43]. So far, analyses of small numbers of cancer susceptibil- ity SNPs have failed to confirm their utility using receiver operator curves [44]. Continued discovery is required to generate the comprehensive set of common and uncommon variants per disease, which could be used to reclassify risk status for either pre- vention or early intervention.
Future directions Cancer GWAS will continue to dis- cover susceptibility alleles but are now poised to investigate pharma- cogenomics and outcome analy- ses, particularly with large studies on the horizon that are well phe- notyped. The major discoveries of cancer GWAS represent an impor- tant transition in the development of integrative scientific collaborations because they have relied on a net- work of epidemiologists, geneticists, and analysts, who have uncovered
genetic markers for risk. The next frontier lies in the investigation of the underlying biology that can ex- plain the contribution of susceptibil- ity alleles to disease pathogenesis or progression, which in turn could lead to more effective strategies for prevention or treatment.
The development of next-gen- eration sequence analysis tools represents a major opportunity to map common and uncommon sus- ceptibility alleles as well as highly penetrant mutations in families. There are daunting challenges in quality control of genome sequence data, including a substantively larg- er number of variants for testing, which compounds the challenge of distinguishing true signal from background noise [45]. To define the comprehensive set of uncommon variants (with a MAF between 0.5% and 5%), hybrid approaches will be required. Recently, two groups con- verged on susceptibility mutations in MITF for melanoma based on not only family- and population-based association studies but also labo- ratory insights into the underlying biology of disruption of sumoylation of MITF protein [46,47].
The next generation of stud- ies will focus on two major exten- sions of the GWAS approach: (i) investigation of the interrelationship between germline susceptibility alleles and somatic alterations [43]; and (ii) testing the clinical and public health applicability of the full spec- trum of cancer susceptibility alleles identified through GWAS, which in- cludes rare, highly penetrant muta- tions, less common alleles with mod- erate penetrance, and the common alleles with small effect sizes. The susceptibility alleles discovered by cancer GWAS are not yet ready for clinical implementation. To develop accurate and useful recommenda- tions for genetic counselling in the new age of complex disease, more discovery is needed that extends into the fraction of the uncommon variants and rare mutations. New paradigms for implementing suscep- tibility alleles in precision medicine represent long-term goals and are expected to be applied to classifica- tion of risk groups that could be of- fered early detection or intervention.
Chapter 3.2 • Genome-wide association studies 201
3 B
IO LO
G Y
C H
A P
T E
R 3
.2
References
1. Knudson AG (2000). Chasing the cancer demon. Annu Rev Genet, 34:1–19. http:// dx.doi.org/10.1146/annurev.genet.34.1.1 PMID:11092820
2. Lander ES, Linton LM, Birren B et al.; International Human Genome Sequencing Consortium (2001). Initial sequenc- ing and analysis of the human genome. Nature, 409:860–921. http://dx.doi.org/ 10.1038/35057062 PMID:11237011
3. Venter JC, Adams MD, Myers EW et al. (2001). The sequence of the human genome. Science, 291:1304–1351. http:// d x .d o i.o r g /10.112 6 /s c i e n c e.10 5 8 0 4 0 PMID:11181995
4. Frazer KA, Ballinger DG, Cox DR et al.; International HapMap Consortium (2007). A second generation human haplotype map of over 3.1 million SNPs. Nature, 449:851–861. http://dx.doi.org/10.1038/ nature06258 PMID:17943122
5. Manolio TA, Collins FS, Cox NJ et al. (2009). Finding the missing heritability of complex diseases. Nature, 461:747–753. http://dx.doi.org/10.1038/nature08494 PMID:19812666
6. Risch N, Merikangas K (1996). The future of genetic studies of complex human dis- eases. Science, 273:1516–1517. http://dx. doi.org/10.1126/science.273.5281.1516 PMID:8801636
7. Moore LE, Baris DR, Figueroa JD et al. (2011). GSTM1 null and NAT2 slow acety- lation genotypes, smoking intensity and bladder cancer risk: results from the New England bladder cancer study and NAT2 meta-analysis. Carcinogenesis, 32:182– 189. http://dx.doi.org/10.1093/carcin/bgq 223 PMID:21037224
8. Hashibe M, McKay JD, Curado MP et al. (2008). Multiple ADH genes are associ- ated with upper aerodigestive cancers. Nat Genet, 40:707–709. http://dx.doi. org/10.1038/ng.151 PMID:18500343
9. Chanock SJ, Manolio T, Boehnke M et al.; NCI-NHGRI Working Group on Replication in Association Studies (2007). Replicating genotype-phenotype associa- tions. Nature, 447:655–660. http://dx.doi. org/10.1038/447655a PMID:17554299
10. Burton PR, Clayton DG, Cardon LR et al.; Wellcome Trust Case Control Consortium (2007). Genome-wide association study of 14,000 cases of seven common dis- eases and 3,000 shared controls. Nature, 447:661–678. http://dx.doi.org/10.1038/ nature05911 PMID:17554300
11. Orr N, Chanock S (2008). Common genetic variation and human disease. Adv Genet, 62:1–32. http://dx.doi.org/10.1016/S0065- 2660(08)00601-9 PMID:19010252
12. Eeles RA, Olama AA, Benlloch S et al.; COGS–Cancer Research UK GWAS– ELLIPSE (part of GAME-ON) Initiative; Australian Prostate Cancer Bioresource; UK Genetic Prostate Cancer Study Collaborators/British Association of Urological Surgeons’ Section of Oncology; UK ProtecT (Prostate testing for cancer and Treatment) Study Collaborators; PRACTICAL (Prostate Cancer Association Group to Investigate Cancer-Associated Alterations in the Genome) Consortium (2013). Identification of 23 new prostate cancer susceptibility loci using the iCOGS custom genotyping array. Nat Genet, 45:385–391, e1–e2. http://dx.doi.org/10.1038/ng.2560 PMID:23535732
13. Michailidou K, Hall P, Gonzalez-Neira A et al.; Breast and Ovarian Cancer Susceptibility Collaboration; Hereditary Breast and Ovarian Cancer Research Group Netherlands (HEBON); kConFab Investigators; Australian Ovarian Cancer Study Group; GENICA (Gene Environment Interaction and Breast Cancer in Germany) Network (2013). Large-scale genotyp- ing identifies 41 new loci associated with breast cancer risk. Nat Genet, 45:353–361, e1–e2. http://dx.doi.org/10.1038/ng.2563 PMID:23535729
14. Pharoah PD, Tsai YY, Ramus SJ et al.; Australian Cancer Study; Australian Ovarian Cancer Study Group (2013). GWAS meta-analysis and replication identifies three new susceptibility loci for ovarian cancer. Nat Genet, 45:362–370, e1–e2. http://dx.doi.org/10.1038/ng.2564 PMID:23535730
15. Abecasis GR, Altshuler D, Auton A et al.; 1000 Genomes Project Consortium (2010). A map of human genome variation from population-scale sequencing. Nature, 467:1061–1073. http://dx.doi.org/10.1038/ nature09534 PMID:20981092
16. Wang Z, Jacobs KB, Yeager M et al. (2012). Improved imputation of common and un- common SNPs with a new reference set. Nat Genet, 44:6–7. http://dx.doi.org/ 10.1038/ng.1044 PMID:22200770
17. Chung CC, Magalhaes WC, Gonzalez- Bosquet J, Chanock SJ (2010). Genome- wide association studies in cancer – current and future directions. Carcinogenesis, 31:111–120. http://dx.doi.org/10.1093/carcin/ bgp273 PMID:19906782
18. Hindorff LA, Gillanders EM, Manolio TA (2011). Genetic architecture of cancer and other complex diseases: lessons learned and future directions. Carcinogenesis, 32:945–954. http://dx.doi.org/10.1093/ carcin/bgr056 PMID:21459759
19. Maris JM, Mosse YP, Bradfield JP et al. (2008). Chromosome 6p22 locus as- sociated with clinically aggressive neu- roblastoma. N Engl J Med, 358:2585– 2593. http://dx.doi.org/10.1056/NEJMoa 0708698 PMID:18463370
20. Postel-Vinay S, Véron AS, Tirode F et al. (2012). Common variants near TARDBP and EGR2 are associated with suscepti- bility to Ewing sarcoma. Nat Genet, 44: 323–327. http://dx.doi.org/10.1038/ng.1085 PMID:22327514
21. Garcia-Closas M, Couch FJ, Lindstrom S et al.; Gene ENvironmental Interaction and breast CAncer (GENICA) Network; kCon- Fab Investigators; Familial Breast Cancer Study (FBCS); Australian Breast Cancer Tissue Bank (ABCTB) Investigators (2013). Genome-wide association studies identify four ER negative-specific breast cancer risk loci. Nat Genet, 45:392–398, e1–e2. http://dx.doi.org/10.1038/ng.2561 PMID:23535733
22. Wu C, Li D, Jia W et al. (2013). Genome-wide association study identifies common vari- ants in SLC39A6 associated with length of survival in esophageal squamous-cell carci- noma. Nat Genet, 45:632–638. http://dx.doi. org/10.1038/ng.2638 PMID:23644492
23. Chung CC, Chanock SJ (2011). Current status of genome-wide association studies in cancer. Hum Genet, 130:59–78. http:// dx.doi.org/10.1007/s00439 - 011-1030 -9 PMID:21678065
24. Park JH, Gail MH, Weinberg CR et al. (2011). Distribution of allele frequencies and effect sizes and their interrelationships for com- mon genetic susceptibility variants. Proc Natl Acad Sci U S A, 108:18026–18031. ht t p: //d x.d o i.o r g /10.1073 /p nas.11147 59108 PMID:22003128
25. Kanetsky PA, Mitra N, Vardhanabhuti S et al. (2009). Common variation in KITLG and at 5q31.3 predisposes to testicular germ cell cancer. Nat Genet, 41:811–815. http:// dx.doi.org/10.1038/ng.393 PMID:19483682
26. Rafnar T, Sulem P, Stacey SN et al. (2009). Sequence variants at the TERT-CLPTM1L locus associate with many cancer types. Nat Genet, 41:221–227. http://dx.doi.org/ 10.1038/ng.296 PMID:19151717
27. Haiman CA, Chen GK, Blot WJ et al. (2011). Genome-wide association study of prostate cancer in men of African ancestry identifies a susceptibility locus at 17q21. Nat Genet, 43:570–573. http://dx.doi.org/ 10.1038/ng.839 PMID:21602798
28. Abnet CC, Freedman ND, Hu N et al. (2010). A shared susceptibility locus in PLCE1 at 10q23 for gastric adenocarcinoma and esophageal squamous cell carcinoma. Nat Genet, 42:764–767. http://dx.doi.org/ 10.1038/ng.649 PMID:20729852
202
29. Lan Q, Hsiung CA, Matsuo K et al. (2012). Genome-wide association analysis identi- fies new lung cancer susceptibility loci in never-smoking women in Asia. Nat Genet, 44:1330–1335. http://dx.doi.org/10.1038/ ng.2456 PMID:23143601
30. Wu C, Kraft P, Zhai K et al. (2012). Genome-wide association analyses of esophageal squamous cell carcinoma in Chinese identify multiple susceptibility loci and gene-environment interactions. Nat Genet, 44:1090–1097. http://dx.doi. org/10.1038/ng.2411 PMID:22960999
31. Garcia-Closas M, Rothman N, Figueroa JD et al. (2013). Common genetic poly- morphisms modify the effect of smok- ing on absolute risk of bladder cancer. Cancer Res, 73:2211–2220. http://dx.doi. org/10.1158/0 0 08 - 5 472.CAN -12-2388 PMID:23536561
32. Lango Allen H, Estrada K, Lettre G et al. (2010). Hundreds of variants clustered in genomic loci and biological pathways affect human height. Nature, 467:832–838. http://dx.doi.org/10.1038/nature09 410 PMID:20881960
33. Bernstein BE, Birney E, Dunham I et al.; ENCODE Project Consortium (2012). An integrated encyclopedia of DNA elements in the human genome. Nature, 489:57– 74. http://dx.doi.org/10.1038/nature11247 PMID:22955616
34. Li Q, Seo JH, Stranger B et al. (2013). Integrative eQTL-based analyses reveal the biology of breast cancer risk loci. Cell, 152:633–641. http://dx.doi.org/10.1016/j. cell.2012.12.034 PMID:23374354
35. Lou H, Yeager M, Li H et al. (2009). Fine mapping and functional analysis of a com- mon variant in MSMB on chromosome 10q11.2 associated with prostate cancer susceptibility. Proc Natl Acad Sci U S A, 106:7933–7938. http://dx.doi.org/10.1073/ pnas.0902104106 PMID:19383797
36. Lou H, Li H, Yeager M et al. (2012). Promoter variants in the MSMB gene associated with prostate cancer regu- late MSMB/NCOA4 fusion transcripts. Hum Genet, 131:1453–1466. http://dx. d o i .o r g /10 .10 07/s 0 0 4 3 9 - 012-118 2- 2 PMID:22661295
37. Pomerantz MM, Shrestha Y, Flavin RJ et al. (2010). Analysis of the 10q11 cancer risk locus implicates MSMB and NCOA4 in hu- man prostate tumorigenesis. PLoS Genet, 6:e1001204. http://dx.doi.org/10.1371/journal. pgen.1001204 PMID:21085629
38. Prokunina-Olsson L, Muchmore B, Tang W et al. (2013). A variant upstream of IFNL3 (IL28B) creating a new interferon gene IFNL4 is associated with impaired clearance of hepatitis C virus. Nat Genet, 45:164–171. http://dx.doi.org/10.1038/ng. 2521 PMID:23291588
39. Wu X, Ye Y, Kiemeney LA et al. (2009). Genetic variation in the prostate stem cell antigen gene PSCA confers susceptibil- ity to urinary bladder cancer. Nat Genet, 41:991–995. http://dx.doi.org/10.1038/ng. 421 PMID:19648920
40. Fu YP, Kohaar I, Rothman N et al. (2012). Common genetic variants in the PSCA gene influence gene expression and blad- der cancer risk. Proc Natl Acad Sci U S A, 109:4974– 4979. http://dx.doi.org/10.1073/ pnas.1202189109 PMID:22416122
41. Kohaar I, Porter-Gill P, Lenz P et al. (2013). Genetic variant as a selection marker for anti-prostate stem cell antigen immuno- therapy of bladder cancer. J Natl Cancer Inst, 105:69–73. http://dx.doi.org/10.1093/ jnci/djs458 PMID:23266392
42. Burton H, Chowdhury S, Dent T et al. (2013). Public health implications from COGS and potential for risk stratification and screen- ing. Nat Genet, 45:349–351. http://dx.doi. org/10.1038/ng.2582 PMID:23535723
43. Bolton KL, Chenevix-Trench G, Goh C et al.; EMBRACE; kConFab Investigators; Cancer Genome Atlas Research Network (2012). Association between BRCA1 and BRCA2 mutations and survival in women with in- vasive epithelial ovarian cancer. JAMA, 307:382–390. http://dx.doi.org/10.1001/ jama.2012.20 PMID:22274685
44. Wacholder S, Hartge P, Prentice R et al. (2010). Performance of common ge- netic variants in breast-cancer risk mod- els. N Engl J Med, 362:986 –993. http:// dx.doi.org /10.10 5 6/ N EJ M oa 0 9 07 727 PMID:20237344
45. Mardis ER (2011). A decade’s perspective on DNA sequencing technology. Nature, 470:198–203. http://dx.doi.org/10.1038/ nature09796 PMID:21307932
46. Bertolotto C, Lesueur F, Giuliano S et al.; French Familial Melanoma Study Group (2011). A SUMOylation-defective MITF germline mutation predisposes to melano- ma and renal carcinoma. Nature, 480:94– 98. http://dx.doi.org/10.1038/nature10539 PMID:22012259
47. Yokoyama S, Woods SL, Boyle GM et al. (2011). A novel recurrent mutation in MITF predisposes to familial and sporadic mela- noma. Nature, 480:99–103. http://dx.doi. org/10.1038/nature10630 PMID:22080950
Website A Catalog of Published Genome-Wide Association Studies: http://www.genome.gov/gwastudies
Chapter 3.3 • Gene expression 203
3 B
IO LO
G Y
C H
A P
T E
R 3
.3Summary • Expression of coded genetic
information contained in DNA determines an organism’s phe- notype through synthesis of functional gene products; the synthesis of such proteins and RNA is regulated at the tran- scriptional and translational levels.
• In cancer, malignant transforma- tion of normal cells results from structural changes in DNA and is mediated by aberrant expres- sion of multiple genes.
• By reference to gene expres- sion profiles, tumours have been subclassified into catego- ries with distinct biological and clinical properties; this process sometimes identifies genetic changes and gives rise to thera- peutic targets.
• The human genome is esti- mated to encode for more than 1000 microRNAs that regulate the translation of more than 60% of genes and hence determine gene expression.
• Currently, large-scale inter- national collaborative cancer studies, based on next-genera- tion sequencing technology, are generating a comprehensive
3.3 3 BIOLOGY
Gene expression Magdalena B. Wozniak Paul Brennan
Emmanuel Barillot (reviewer) Dongxin Lin (reviewer)
catalogue of oncogenic alter- ations evident in particular tu- mour types.
Gene expression is the process of translating the information speci- fied through the sequence of bases in DNA into functional products, including proteins and non-coding RNA, such as ribosomal RNA, trans- fer RNA, microRNA (miRNA), and small nuclear RNA, among others (Fig. 3.3.1). In eukaryotes, transcrip- tion is performed by three types of RNA polymerases, each of which requires a promoter DNA sequence and a set of DNA-binding transcrip- tion factors to initiate the process. The promoter sequence is almost always located just upstream of the transcription start site, which cor- responds to the 5′ end. Binding of regulatory proteins to an enhancer sequence results in a shift in chro- matin structure, which either pro- motes or inhibits RNA polymerase and transcription factor binding. Euchromatin, which has a more open structure, is associated with transcription, whereas heterochro- matin is associated with transcrip- tional inactivity (Fig. 3.3.2A).
Only a fraction of the genes in a particular cell are expressed at a given time. The variety of gene ex- pression profiles is determined by
distinct sets of transcription regula- tors. Transcription creates precursor messenger RNA (pre-mRNA), which further undergoes a series of modi- fications, leading to mature mRNA. These modifications include the en- zymatic reactions of 5′ capping and 3′ cleavage and polyadenylation. Protein-coding sequences in DNA account for only 1–2% of the human genome [1]. The increased diversity at the level of primary mRNA and proteins can be accounted for by alternative RNA splicing, which cre- ates a series of different transcripts originating from a single gene by re- moving or retaining some introns or exons in particular mature mRNAs. The resulting transcripts can po- tentially be translated into different proteins. Each mRNA is composed of three parts: a 5′ untranslated re- gion, an open reading frame, and a 3′ untranslated region.
Regulation of gene expression Regulation of gene expression can occur at any step of expression and includes various systems that control and determine which genes are switched on or off and to what extent the genes are expressed. Transcriptional regulation of gene expression levels occurs in three main ways: genetic (involving direct interaction with DNA), modulation
204
(involving interaction with transcrip- tion machinery), and epigenetic (involving non-sequence changes in DNA). Genes contain in their se- quence several protein-binding sites around the coding regions, such as enhancers, insulators, and silenc- ers. These sites facilitate binding of transcription factors, resulting in blocking or activation of transcription. Furthermore, the activity of transcrip- tion factors can also be modulated by post-translational protein modifica- tions such as acetylation, phosphory- lation, or glycosylation (Fig. 3.3.2B).
Epigenetic changes, and spe- cifically DNA methylation, modulate transcription by altering the accessi- bility of DNA to proteins. Particularly important for regulation of gene ex- pression levels are post-transcrip- tional modifications, including 5′ cap- ping and 3′ polyadenylation, which protect mRNA from degradation during nuclear export. Once generat- ed, mature RNA isoforms are subject to many levels of regulation, includ- ing the regulation of translation by
miRNAs and regulatory factors, the use of alternative translation start sites, RNA localization, and mRNA stability and turnover. Direct inhibi- tion of protein translation by toxins and antibiotics is less frequent. In addition, a substantial number of isoforms can be an effect of single- nucleotide polymorphisms (SNPs), small genetic differences between alleles of the same gene. Finally, pro- tein expression can also be reduced by degradation, in which unneeded, damaged, and improperly folded pro- teins are labelled for degradation by ubiquitination.
Dysregulation of gene expression in cancer Malignant transformation is accom- panied and characterized by disrup- tion of genetic information and aber- rant expression of multiple genes. These changes enhance the survival of tumour cells and alter their prolif- erative capacity, leading to tumour formation. The fact that the majority
of cells within a tumour mass share a common gene expression profile has encouraged systematic analyses of differential gene expression in hu- man tumour samples to provide an estimate of the degree of genetic and epigenetic dysregulation.
Measurement of gene expression To monitor gene expression, ideally the amount of each final gene prod- uct, which is in most cases a protein, should be measured. However, due to the limitations of protein-based assays, pre-mRNA expression is frequently studied. Several methods have been developed for quantifica- tion of mRNA levels. These include northern blotting, quantitative re- verse transcription polymerase chain reaction (qRT-PCR), RNase pro- tection assays, microarrays, serial analysis of gene expression (SAGE), and next-generation RNA sequenc- ing (RNA-seq). While northern blot- ting and qRT-PCR are often used to
Fig. 3.3.1. Schematic overview of the protein-coding gene expression pathway in eukaryotic organisms. E, enhancer sequence; T, terminator (transcription termination site); m7G, 7-methylguanylate cap.
Chapter 3.3 • Gene expression 205
3 B
IO LO
G Y
C H
A P
T E
R 3
.3
Therapeutic implications of dysregulated DNA damage signalling Peter Bouwman and Jos Jonkers
The intimate link between cancer and DNA damage signalling is clear from the fact that many tumour sup- pressor genes are involved in DNA repair [1,2]. Well-known examples are BRCA1 and BRCA2, which are important for DNA repair by homol- ogous recombination. In addition, many known carcinogens such as tobacco smoke, alcohol, ultraviolet
radiation, and ionizing radiation are known to cause DNA damage.
Paradoxically, tumours can also be treated by DNA-damaging ir- radiation or chemical compounds. Although DNA damage also affects normal cells, tumour cells are often more vulnerable because of defects in DNA repair pathways or critical cell-cycle checkpoints. Especially defects in homology-directed repair
(HDR) of DNA double-strand breaks can be exploited therapeutically. It has recently become clear that this is possible not only by directly dam- aging the DNA but also via inhibi- tion of PARP enzymes. Inhibition of PARP1 and PARP2 can block alter- native repair pathways and result in trapping of PARP to DNA, which is detrimental for HDR-deficient cells [3].
Fig. B3.3.1. DNA damage response (DDR)-related mechanisms of therapy resistance. Loss of function of DNA repair genes can cripple the affected pathway to the extent that tumour cells become highly sensitive to certain types of DNA damage. In addition, the remaining repair pathways may become essential for survival. Resistance to chemotherapy can be a result of partial activity of mutated alleles. Such hypomorphic mutations can be sufficiently affected to lead to tumour formation, yet too active to result in hypersensitivity to targeted therapy. In case mutations lead to a completely inactivated (null) allele, chemotherapy may select for mutations that revert the original defect or select for mutations in other DDR genes that result in pathway rewiring. Crosses symbolize inactivation. Tumour cells with wild-type levels of DDR activity are indicated in green, cells with a partially active DDR pathway are shown in beige, and cells with impaired DDR are shown in purple.
206
Despite their exquisite sensitivity to DNA damage, tumours with HDR defects can eventually become re- sistant to therapy. Apart from mech- anisms that result in decreased intracellular drug concentrations, resistance may also be caused by altered DNA damage signalling [1,2]. Of note, tumours with mutations in a specific DNA repair gene do not always show the same response to therapy; missense mutations in a specialized region will have differ- ent effects compared with mutations that completely abolish expression. Furthermore, tumours often con- sist of heterogeneous populations of cells that may share a common ancestor but have acquired different additional mutations or epigenetic alterations. Data from preclinical
models suggest that this may result in increased capacity to repair DNA damage, and hence resistance to therapy. It has also been shown that therapy-resistant BRCA1- and BRCA2-associated tumours may re- store their HDR defect by genetic re- version. Secondary mutations may overcome the original defect, result- ing in the expression of (partially) functional BRCA1 or BRCA2 protein. Alternatively, DNA repair pathways may be perturbed to withstand DNA- damaging therapy. Recent data sug- gest that this is possible in respect of the two major pathways for repair of DNA double-strand breaks, error- free homologous recombination and intrinsically error-prone non-homol- ogous end joining (NHEJ). The HDR defect of BRCA1-deficient cells
can, for instance, be suppressed by loss of the NHEJ protein 53BP1. Stratification of patients with specific DNA repair defects and combina- tions of targeted therapeutics will be required to maximize the likelihood of successful treatment. To avoid resistance, the development of bio- markers for the activity of factors involved in DNA damage response will be of critical importance.
References 1. Bouwman P, Jonkers J (2012). Nat Rev
Cancer, 12:587–598. http://dx.doi.org/10. 1038/nrc3342 PMID:22918414
2. Lord CJ, Ashworth A (2012). Nature, 481:287–294. http://dx.doi.org/10.1038/ nature10760 PMID:22258607
3. Helleday T (2011). Mol Oncol, 5:387–393. http://dx.doi.org/10.1016/j.molonc.2011. 07.001 PMID:21821475
verify the results for particular genes or for some low-expression genes, microarrays, SAGE, and RNA-seq are used to obtain broad coverage of the genome. Now that identifica- tion of genome-wide sets of protein- coding genes is feasible, combined with the discovery of miRNAs, many researchers have taken advantage of array-based approaches to search for gene expression patterns within most cancer types.
In the context of cancer, gene expression profiling with DNA mi- croarrays has been used to iden- tify individual transcripts and regu- latory gene networks mediating disease pathogenesis, to discover molecular targets for drug develop- ment, to more accurately classify tumours, and to identify molecular markers useful for disease diagno- sis and prognosis or as predictors of clinical outcomes. In addition to detecting changes in transcrip- tional levels, the recently developed RNA-seq technique can detect other abnormalities in the cancer transcriptome, such as SNPs, al- ternative splice variants, novel cod- ing and non-coding transcripts, and gene fusions. Due to decreasing costs, RNA-seq is becoming more
common as a method for cancer gene expression profiling. RNA-seq is also superior to microarray tech- niques, given its unbiased approach in probe selection. However, the
requirement for increased amounts of sample RNA and the challeng- ing bioinformatics analyses of gen- erated data remain limitations of this technique.
Fig. 3.3.2. (A) Representation of “closed” and “open” chromatin structures. (B) Overview of major post-translational modifications that may play essential roles in the regulation of gene expression and be altered in the course of disease processes. Modifications are specified in relation to part of the N-terminal tail of a histone protein; the single letters indicate specific amino acid residues that may be subject to particular modifications.
Chapter 3.3 • Gene expression 207
3 B
IO LO
G Y
C H
A P
T E
R 3
.3
Gene expression microarray studies in cancer Microarray profiling technology, which has been most widely used to study gene expression in cancer, represents an important develop- ment in genomic marker research and has been evaluated as a tool for decision-making in clinical practice. Microarray studies aim to identify genes that are differentially expressed between known groups of samples (class comparison), to discover unknown patterns of gene expression (class discovery), and to develop classifiers associated with disease or treatment outcome (class prediction). For that purpose, both supervised and unsupervised methods of analysis have been used to search for molecular signatures associated with known classes and to discover new classes of a dis- ease. Even though diagnosis and tumour subclassification are still dominated by histological analysis for most cancers, these methods have proven to be insufficient to distinguish between subclasses in several types of cancer.
Specifically, consistent definition of diffuse large B-cell lymphoma has proven difficult due to discrepancies between inter- and intra-observer reproducibility [2]. Gene expres- sion profiling has enabled the sub- classification of diffuse large B-cell lymphoma tumours into categories with distinct biological properties and prognoses. This has led to the identification of two subtypes: ger- minal centre B-like and activated B-like diffuse large B-cell lymphoma [3]. For patients with germinal centre B-like disease, the 10-year survival rate was about 80%, whereas for those with activated B-like disease, the 8-year survival rate was about 40%.
In some tumour types, such as breast cancer [4] and leukaemia [5], molecular markers have been ad- juncts to histological classification for decades. Perhaps the best known example is the contribution of expression profiling to the clas- sification of breast tumours, which has resulted in the development of
the PAM50 Breast Cancer Intrinsic Classifier, which sorts breast cancer into the major molecular subtypes of basal-like, human epidermal growth factor receptor 2 (HER2)- positive, luminal A, and luminal B (Fig. 3.3.3A). Gene expression sig- nature studies performed in melano- ma have also led to the development of a protein-based test measuring expression of five genes that can distinguish malignant melanoma from benign tissue [6].
The use of gene expression sig- natures has also enabled the iden- tification of prognostic subclasses. The more successful examples of gene expression profiles as prog- nostic indicators have been their use to define the risk of relapse in patients with early-stage breast cancer. Three profiles have proven prognostic ability: the MammaPrint 70-gene profile, the Oncotype DX 21-gene recurrence score, and the Rotterdam 76-gene prognos- tic signature [7–9]. These profiles were obtained using tumour-derived mRNA either applied on DNA micro- arrays or by RT-PCR assay. All three tests are used in clinical practice to predict prognosis and assist thera- peutic intervention. Using a similar approach, a 12-gene expression sig- nature has been developed for colon cancer as an independent predictor of recurrence in patients with stage II disease, which aids the more ac- curate identification of patients who will not benefit from chemotherapy [10]. In addition, other studies have identified a 133-gene prognostic sig- nature in adult acute myeloid leukae- mia with normal karyotype as well as new prognostic subtypes in diffuse large B-cell lymphoma [11,12]. Gene expression signatures to assess the risk of recurrence in patients with non-small cell lung cancer [13] and in several other cancer types have also been proposed.
Involvement of non-coding RNAs in cancer Recently, the functional relevance of the non-protein-coding portion of the genome became evident in the
context of both normal development and disease pathogenesis. Currently, a total of 9173 small RNAs are anno- tated by the GENCODE project (ver- sion 16) [14], the majority of which correspond to four major classes: small nuclear RNAs, small nucleolar RNAs, miRNAs, and transfer RNAs. Other classes of short and midsize non-coding RNAs include PIWI- interacting RNAs (piRNAs), transcrip- tion initiation RNAs, promoter-associ- ated small RNAs, transcription start site-associated RNAs, and promoter upstream transcripts. GENCODE also contains 13 220 long non-coding RNA genes, a heterogeneous group of non-coding transcripts longer than 200 nucleotides encompassing main- ly large intergenic non-coding RNAs and transcribed ultraconserved re- gions (T-UCRs).
The most widely studied class of non-coding RNAs with functional importance are miRNAs. miRNAs are small RNA polymers, of approxi- mately 21 nucleotides, that mediate post-transcriptional gene silencing through translational repression [15]. The human genome is estimated to encode for more than 1000 miRNAs that regulate the translation of more than 60% of protein-coding genes. The biogenesis of miRNAs involves a multistep process (Fig. 3.3.4). miRNAs are implicated in essentially all biological processes, given that a single miRNA can target hundreds of mRNAs. Aberrant miRNA expres- sion has been observed in many diseases, including cancer, where miRNAs can operate as either onco- genes or tumour suppressor genes. In addition to single miRNAs, such as miR-21 and miR-10b, clusters of miRNAs, such as the miR-17-92 cluster, have been reported as on- cogenes that are upregulated in cancers and that regulate tumour cell proliferation and metastasis. In contrast, members of the let-7 and miR-34 families are tumour suppressors, which are downregu- lated in many cancers. Under normal circumstances, these miRNAs inhibit cancer growth by targeting various oncogene-encoding mRNAs for deg- radation. Notably, some miRNAs can
208
Fig. 3.3.3. Successful gene expression profiling applications in breast cancer. (A) Hierarchical clustering for the PAM50 classifier genes normalized to the 5 control genes using 171 formalin-fixed, paraffin-embedded procured breast samples. The cluster den- drogram shows the five significant groups previously identified and designated as luminal A, luminal B, HER2-positive, basal-like, and normal. (B) Expression data matrix of 70 prognostic marker genes from sporadic breast tumours from 78 patients diagnosed at age less than 55 years with tumour size less than 5 cm and no lymph node involvement (I). Each row represents a tumour and each column a gene. Genes are ordered according to their correlation coefficient with the two prognostic groups. Tumours are ordered by the correlation to the average profile of the good prognosis group (II). Solid line, prognostic classifier with optimal accuracy; dashed line, with optimized sensitivity. Above the dashed line, patients have a good prognosis signature; below the dashed line, the prog- nosis signature is poor. The metastasis status for each patient is also shown (III): white indicates patients who developed distant metastases within 5 years after the primary diagnosis; black indicates patients who continued to be disease-free for at least 5 years.
function as both an oncogene and a tumour suppressor, depending on the cell and cancer type; hence, miRNA expression profiles should be identi- fied in each cancer type.
Genome-wide miRNA expres- sion profiling microarray studies have shown that effectively all can- cers present an altered miRNA pro- file. In one of the first studies, two
miRNAs were identified as tumour suppressors controlling the patho- genesis of B-cell chronic lympho- cytic leukaemia. Both miR-15a and miR-16-1 were found in a region
Chapter 3.3 • Gene expression 209
3 B
IO LO
G Y
C H
A P
T E
R 3
.3
of chromosome 13 (13q14) that is frequently deleted in chronic lympho- cytic leukaemia cells [16]. In solid tumours, specific miRNA expres- sion profiles were also investigated, including colon adenocarcinoma, breast carcinoma, glioblastoma, hepatocellular carcinoma, thyroid carcinoma, and lung cancer. A large miRNA profiling analysis of 540 samples covering 6 solid tumours
(breast, lung, stomach, prostate, co- lon, and pancreas) identified 43 dys- regulated miRNAs compared with normal tissues [17]. Attempts have been made to assess the correlation of outcome and relapse with miRNA expression levels in lung [18], colon [19], and liver [20] cancers. These findings, together with the high sta- bility of miRNAs in serum samples, make miRNAs attractive candidates
as both diagnostic and prognostic indicators in the clinical setting.
Several studies have demon- strated an association of circulat- ing miRNAs with overall survival and early detection. A group of four miRNAs – miR-486, miR-30d, miR-1, and miR-499 – are related to de- creased overall survival of patients with non-small cell lung cancer [21]. In a study of ovarian cancer, miR-21,
Fig. 3.3.4. Model for the biogenesis and post-transcriptional suppression of microRNAs (miRNAs). The nascent primary miRNA (pri-miRNA) transcripts are first processed into precursor miRNAs (pre-miRNAs) of approximately 70 nucleotides by the RNase III enzyme Drosha inside the nucleus. Pre-miRNAs are transported to the cytoplasm by exportin 5 and are processed into miRNA:miRNA* duplexes by another RNase III enzyme, Dicer. Only one strand of the miRNA:miRNA* duplex is preferentially assembled into the RNA-induced silencing complex (RISC), which subsequently acts on its target by translational repression or messenger RNA (mRNA) cleavage, depending, at least in part, on the level of complementarity between the small RNA and its target. ORF, open reading frame; UTR, untranslated region.
210
miR-92, miR-93, miR-126, and miR-29a were found to be overex- pressed in cancer patients before therapy compared with healthy controls. The same study suggest- ed miR-21, miR-92, and miR-93 as potential early-detection biomarkers for ovarian cancer [22]. Other class- es of non-coding RNAs, such as small nucleolar RNAs, piRNAs, and piRNA-like transcripts, have been found to be implicated in malignant processes in a range of tumour types, although the mechanisms underlying the oncogenic effects of these non-coding RNAs are largely unknown. The differential expression of small nucleolar RNAs has recently been demonstrated in non-small cell lung cancer [23]. T-UCRs and large intergenic non-coding RNAs have also been shown to be relevant in tumorigenesis. Many large intergen- ic non-coding RNAs are differentially expressed in cancers with different clinical outcomes. Aberrant expres- sion signatures of T-UCRs have been described for chronic lympho- cytic leukaemia, colorectal cancer, and hepatocellular carcinoma [24]. In addition, many T-UCRs show im- portant complementarity to specific miRNAs, suggesting that T-UCRs could act as miRNA targets [25].
One of the best understood exam- ples of the involvement of large in- tergenic non-coding RNAs in cancer is the overexpression of homeobox transcript antisense RNA (HOTAIR) in human neoplasia. This large inter- genic non-coding RNA may have an active role in modulating the cancer epigenome and in mediating cell transformation [25,26].
Gene fusions Fusion genes are hybrid genes formed through translocations, in- terstitial deletions, or chromosomal inversions of two previously sepa- rate genes (Fig. 3.3.5). Gene fusions are a known mechanism for onco- gene activation in several malignan- cies, such as haematological can- cers (e.g. leukaemia and lymphoma), sarcoma, and prostate cancer [27]. Oncogenic gene fusions may lead to a gene product with a different function from those of the two fu- sion partners. Examples of note in- clude the oncogenic gene fusions BCR-ABL in chronic myeloid leukae- mia [28], TEL-AML1 in acute lympho- blastic leukaemia [29], recurrent translocations involving ETS family members such as TMPRSS2-ERG in prostate cancer [30], EWS-FLI1 in
Ewing sarcoma [31], and EML4-ALK in lung cancer [32]. Even though gene fusions have been widely de- scribed in rare haematological ma- lignancies and sarcomas [27], the discovery of recurrent gene fusions in prostate [30], lung [32], and breast cancer [33] suggests a putative role in other solid tumours.
Clinical relevance Despite great advances in molecu- lar profiling of cancers, only a few findings have been successfully translated into clinical applications. Many studies have shown the abil- ity of gene expression profiling to identify diagnostic, prognostic, and predictive molecular markers or signatures. However, a substantial proportion of these biomarkers, identified in retrospective studies, have failed in subsequent validation studies. Also, many of the signatures were developed without a clear fo- cus on the intended clinical use, and proper independent validation stud- ies establishing their medical utility have rarely been performed. Indeed, the contribution that expression sig- natures may make to clinical practice has been questioned [34].
Fig. 3.3.5. Schematic overview of the process of formation and structure of fusion genes. The two vertical red lines at the top of the panel indicate break points on each of two chromosomes, to generate, as one product, a fused chromosome giving rise to a mes- senger RNA encoding the corresponding fusion protein.
Chapter 3.3 • Gene expression 211
3 B
IO LO
G Y
C H
A P
T E
R 3
.3
Of all the gene expression signa- tures identified for cancer, only a few have become commercially avail- able. These include the MammaPrint 70-gene signature for breast cancer prognosis (Fig. 3.3.3B) [35], which is the only assay for breast cancer management that has been cleared by the United States Food and Drug Administration (FDA), in February 2007, and the Oncotype DX 21-gene signature for breast cancer [36].
Since then, some academic in- stitutions in the Netherlands and the USA have initiated translation of these gene expression profiles into clinical practice. There is still concern about incorporation of gene expression signatures into routine clinical practice before confirmation of their efficacy from suitably pow- ered randomized controlled trials. This has resulted in the initiation of large prospective clinical trials called MINDACT and TAILORx, to test the efficacy of MammaPrint and Oncotype DX, respectively, in pre- dicting the benefit of adjuvant che- motherapy in breast cancer patients. Interestingly, a recent multicentre study showed that the medical on- cologist’s treatment recommenda- tion was changed for almost one third of patients on the basis of the Oncotype DX 21-gene assay [37].
Development of targeted therapy Gene expression profiling can con- tribute to more effective therapy through its diagnostic, prognos- tic, and predictive value as well as through target discovery. Accurate diagnosis will minimize the side-ef- fects from non-beneficial treatments. Perhaps most revealing is the fact that 70–80% of breast cancer pa- tients who receive chemotherapy based on traditional predictors would have survived without it. Since the early 1990s, targeted therapies have been developed, such as inhibitors targeted to proteins encoded by mutated or overexpressed cancer genes. One of the most remark- able examples is imatinib, a potent inhibitor of ABL kinase in chronic
myeloid leukaemia, which was ap- proved as the first-line treatment for this disease by the FDA in 2001 [38]. Imatinib was discovered in 1992 and is regarded as the first generation of targeted drugs. Furthermore, trastu- zumab, a monoclonal antibody that targets HER2, was a groundbreak- ing drug in the treatment of HER2- positive breast cancer.
Recently, non-coding RNAs and the proteins involved in their biogen- esis and activity have become tar- gets of new therapeutic approaches. Therapies inhibiting miRNA function based on base-pair complementarity have been developed and include three main classes: locked nucleic acids, anti-miRNA oligonucleotides, and antagomirs. Several strategies, called miRNA replacement therapy, have also been proposed to restore the function of downregulated tumour suppressor miRNAs. However, these therapies are currently in the early stages of development and have been studied mostly in mouse mod- els and non-human primates. Finally, an approach aimed at restoring the “global miRNAome” entails the use of histone deacetylase inhibitors and
DNA demethylating agents. These agents, even without target specific- ity, have been shown to release the epigenetic silencing of tumour sup- pressor non-coding RNAs, such as miRNAs and T-UCRs [25,39].
Large-scale cancer genome studies, such as the International Cancer Genome Consortium and the Cancer Genome Atlas, are already applying next-generation sequenc- ing technologies to tumours from 50 different cancer types to generate more than 25 000 cancer genomes at the genomic, epigenomic, and transcriptomic levels and should generate a complete catalogue of oncogenic alterations. In addition, initiatives such as the Encyclopedia of DNA Elements integrate multiple technologies to encode all known functional elements of the genome, including the study of novel RNA transcripts, protein-coding regions, transcription factor binding sites, and chromatin structure such as DNA methylation patterns. These will provide a comprehensive cata- logue of oncogenic alterations, some of which may prove to be new thera- peutic targets.
Fig. 3.3.6. A GeneChip loaded with RNA from a single tumour, to identify mutations in expressed genes that may account for the tumour’s growth and response to therapy.
212
Strengths and limitations The advent of high-throughput gene expression technologies such as mi- croarrays and RNA-seq has made it possible to assess the expression of tens of thousands of genes in a single experiment. Initial microarray studies raised concerns about reproducibil- ity. To prevent common procedural failures and to establish high-quality tools for the microarray community, the FDA initiated the MicroArray Quality Control project in 2005 [40]. In addition, to standardize all microarray (and now RNA-seq) data, every new publication in the field is required to follow MIAME (Minimum Information About a Microarray Experiment) guidelines and to deposit expres- sion profile data in an open data- base. The largest publicly accessible repositories of microarray and other
high-throughput functional geno- mics data are the Gene Expression Omnibus at the United States National Center for Biotechnology Information [41] and ArrayExpress [42].
Microarray technology allows the identification of both single genes and multigene alterations at the level of molecular pathways and networks, providing a broader view of a disease. However, gene expression profiling studies, both microarray and RNA-seq, of- ten generate immense amounts of data that are challenging to interpret and need to be subjected to more advanced bioinformatics data analysis. Furthermore, even though changes at the transcriptional level are important, it is essential to keep in mind that they are just a part of
the bigger picture. The majority of cellular functions are carried out by proteins, and biological changes can be modulated not only through alter- ations at the protein level but also by post-translational modifications such as glycosylation, methylation, acety- lation, and phosphorylation. These modifications could change protein conformation and lead to changes in activity. Several consortia are systematically studying the genome, the transcriptome, the epigenome, and the proteome. The key to making the best use of this information lies in the integration of these data. These technological advances coupled with systems biology promise exciting prospects of understanding the basis of cancer and developing improved diagnostic tools and therapies.
References
1. Claverie JM (2005). Fewer genes, more noncoding RNA. Science, 309:1529–1530. http://dx.doi.org/10.1126/science.1116800 PMID:16141064
2. The Non-Hodgkin’s Lymphoma Classification Project (1997). A clinical evaluation of the International Lymphoma Study Group clas- sification of non-Hodgkin’s lymphoma. Blood, 89:3909–3918. PMID:9166827
3. Alizadeh AA, Eisen MB, Davis RE et al. (2000). Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature, 403:503–511. http://dx.doi. org/10.1038/35000501 PMID:10676951
4. Sørlie T, Perou CM, Tibshirani R et al. (2001). Gene expression patterns of breast carcinomas distinguish tumor sub- classes with clinical implications. Proc Natl Acad Sci U S A, 98:10869–10874. http://dx.doi.org/10.1073/pnas.191367098 PMID:11553815
5. Haferlach T, Kohlmann A, Wieczorek L et al. (2010). Clinical utility of microar- ray-based gene expression profiling in the diagnosis and subclassification of leukemia: report from the International Microarray Innovations in Leukemia Study Group. J Clin Oncol, 28:2529–2537. http:// dx.doi.org/10.1200/JCO.2009.23.4732 PMID:20406941
6. Kashani-Sabet M, Rangel J, Torabian S et al. (2009). A multi-marker assay to distinguish malignant melanomas from benign nevi. Proc Natl Acad Sci U S A, 106:6268–6272. http://dx.doi.org/10.1073/pnas.0901185106 PMID:19332774
7. Paik S, Shak S, Tang G et al. (2004). A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer. N Engl J Med, 351:2817–2826. http://dx.doi.org/10.1056/NEJMoa041588 PMID:15591335
8. van ’t Veer LJ, Dai H, van de Vijver MJ et al. (2002). Gene expression profiling pre- dicts clinical outcome of breast cancer. Nature, 415:530–536. http://dx.doi.org/10. 1038/415530a PMID:11823860
9. Wang Y, Klijn JG, Zhang Y et al. (2005). Gene-expression profiles to predict distant metastasis of lymph-node-negative pri- mary breast cancer. Lancet, 365:671–679. PMID:15721472
10. Gray RG, Quirke P, Handley K et al. (2011). Validation study of a quantitative multigene reverse transcriptase-polymerase chain reaction assay for assessment of recur- rence risk in patients with stage II colon cancer. J Clin Oncol, 29:4611– 4619. http:// dx.doi.org/10.1200/JCO.2010.32.8732 PMID:22067390
11. Dave SS, Fu K, Wright GW et al.; Lymphoma/Leukemia Molecular Profiling Project (2006). Molecular diagnosis of Burkitt’s lymphoma. N Engl J Med, 354:2431–2442. http://dx.doi.org/10.1056/ NEJMoa055759 PMID:16760443
12. Rosenwald A, Wright G, Chan WC et al.; Lymphoma/Leukemia Molecular Profiling Project (2002). The use of molecular profiling to predict survival after chemotherapy for dif- fuse large-B-cell lymphoma. N Engl J Med, 346:1937–1947. http://dx.doi.org/10.1056/ NEJMoa012914 PMID:12075054
13. Potti A, Mukherjee S, Petersen R et al. (2006). A genomic strategy to refine prognosis in early-stage non-small-cell lung cancer. N Engl J Med, 355:570–580. http://dx.doi.org/10.1056/NEJMoa060467 PMID:16899777
14. Djebali S, Davis CA, Merkel A et al. (2012). Landscape of transcription in human cells. Nature, 489:101–108. http://dx.doi. org/10.1038/nature11233 PMID:22955620
15. He L, Hannon GJ (2004). MicroRNAs: small RNAs with a big role in gene regulation. Nat Rev Genet, 5:522–531. http://dx.doi. org/10.1038/nrg1379 PMID:15211354
16. Croce CM (2009). Causes and conse- quences of microRNA dysregulation in cancer. Nat Rev Genet, 10:704–714. h t t p : / / d x . d o i . o r g / 1 0 .1 0 3 8 / n r g 2 6 3 4 PMID:19763153
Chapter 3.3 • Gene expression 213
3 B
IO LO
G Y
C H
A P
T E
R 3
.3
17. Volinia S, Calin GA, Liu CG et al. (2006). A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci U S A, 103:2257–2261. http://dx.doi.org/10.1073/pnas.0510565103 PMID:16461460
18. Yu SL, Chen HY, Chang GC et al. (2008). MicroRNA signature predicts survival and relapse in lung cancer. Cancer Cell, 13:48–57. http://dx.doi.org/10.1016/j.ccr. 2007.12.008 PMID:18167339
19. Schetter AJ, Leung SY, Sohn JJ et al. (2008). MicroRNA expression profiles as- sociated with prognosis and therapeutic outcome in colon adenocarcinoma. JAMA, 299:425– 436. http://dx.doi.org/10.1001/ jama.299.4.425 PMID:18230780
20. Chung GE, Yoon JH, Myung SJ et al. (2010). High expression of microRNA-15b predicts a low risk of tumor recurrence following curative resection of hepatocel- lular carcinoma. Oncol Rep, 23:113–119. PMID:19956871
21. Cortez MA, Bueso-Ramos C, Ferdin J et al. (2011). MicroRNAs in body fluids – the mix of hormones and biomarkers. Nat Rev Clin Oncol, 8:467–477. http://dx.doi.org/10.1038/ nrclinonc.2011.76 PMID:21647195
22. Resnick KE, Alder H, Hagan JP et al. (2009). The detection of differentially expressed mi- croRNAs from the serum of ovarian cancer patients using a novel real-time PCR plat- form. Gynecol Oncol, 112:55–59. http:// dx.doi.org/10.1016/j.ygyno.2008.08.036 PMID:18954897
23. Liao J, Yu L, Mei Y et al. (2010). Small nu- cleolar RNA signatures as biomarkers for non-small-cell lung cancer. Mol Cancer, 9:198. http://dx.doi.org/10.1186/1476-4598- 9-198 PMID:20663213
24. Calin GA, Liu CG, Ferracin M et al. (2007). Ultraconserved regions encoding ncRNAs are altered in human leukemias and carcinomas. Cancer Cell, 12:215–229. http:/dx.doi.org/10.1016/j.ccr.2007.07.027 PMID:17785203
25. Esteller M (2011). Non-coding RNAs in human disease. Nat Rev Genet, 12:861–874. http://dx.doi.org/10.1038/nrg 3074 PMID:22094949
26. Gupta RA, Shah N, Wang KC et al. (2010). Long non-coding RNA HOTAIR repro- grams chromatin state to promote can- cer metastasis. Nature, 464:1071–1076. http://dx.doi.org/10.1038/nature08975 PMID:20393566
27. Mitelman F, Johansson B, Mertens F (2007). The impact of translocations and gene fusions on cancer causation. Nat Rev Cancer, 7:233–245. http://dx.doi.org/10.1038/ nrc2091 PMID:17361217
28. de Klein A, van Kessel AG, Grosveld G et al. (1982). A cellular oncogene is translocated to the Philadelphia chromosome in chronic myelocytic leukaemia. Nature, 300:765– 767. http://dx.doi.org/10.1038/300765a0 PMID:6960256
29. Golub TR, Barker GF, Bohlander SK et al. (1995). Fusion of the TEL gene on 12p13 to the AML1 gene on 21q22 in acute lymphoblastic leukemia. Proc Natl Acad Sci U S A, 92:4917– 4921. http://dx.doi.org/10.1073/pnas.92.11. 4917 PMID:7761424
30. Tomlins SA, Rhodes DR, Perner S et al. (2005). Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science, 310:644–648. http://dx.doi.org/10.1126/science.1117679 PMID:16254181
31. Delattre O, Zucman J, Plougastel B et al. (1992). Gene fusion with an ETS DNA- binding domain caused by chromo- some translocation in human tumours. Nature, 359:162–165. http://dx.doi.org/10. 1038/359162a0 PMID:1522903
32. Choi YL, Takeuchi K, Soda M et al. (2008). Identification of novel isoforms of the EML4-ALK transforming gene in non-small cell lung cancer. Cancer Res, 68:4971– 4976. http://dx.doi.org/10.1158/0008-5472. CAN-07-6158 PMID:18593892
33. Edgren H, Murumagi A, Kangaspeska S et al. (2011). Identification of fusion genes in breast cancer by paired-end RNA-sequencing. Genome Biol, 12:R6. http://dx.doi.org/10.1186/gb-2011-12-1-r6 PMID:21247443
34. Chibon F (2013). Cancer gene expression signatures – the rise and fall? Eur J Cancer, 49:2000–2009. http://dx.doi.org/10.1016/j. ejca.2013.02.021 PMID:23498875
35. van de Vijver MJ, He YD, van ’t Veer LJ et al. (2002). A gene-expression signa- ture as a predictor of survival in breast cancer. N Engl J Med, 347:1999–2009. http://dx.doi.org/10.1056/NEJMoa021967 PMID:12490681
36. Perou CM, Sørlie T, Eisen MB et al. (2000). Molecular portraits of human breast tu- mours. Nature, 406:747–752. http://dx.doi. org/10.1038/35021093 PMID:10963602
37. Lo SS, Mumby PB, Norton J et al. (2010). Prospective multicenter study of the impact of the 21-gene recurrence score assay on med- ical oncologist and patient adjuvant breast cancer treatment selection. J Clin Oncol, 28:1671–1676. http://dx.doi.org/10.1200/ JCO.2008.20.2119 PMID:20065191
38. O’Brien SG, Guilhot F, Larson RA et al.; IRIS Investigators (2003). Imatinib com- pared with interferon and low-dose cytar- abine for newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med, 348:994–1004. http://dx.doi.org/10.1056/ NEJMoa022457 PMID:12637609
39. Lujambio A, Portela A, Liz J et al. (2010). CpG island hypermethylation-associated silencing of non-coding RNAs transcribed from ultraconserved regions in human can- cer. Oncogene, 29:6390–6401. http://dx.doi. org/10.1038/onc.2010.361 PMID:20802525
40. Shi L, Reid LH, Jones WD et al.; MAQC Consortium (2006). The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene ex- pression measurements. Nat Biotechnol, 24:1151–1161. http://dx.doi.org/10.1038/ nbt1239 PMID:16964229
41. Barrett T, Troup DB, Wilhite SE et al. (2009). NCBI GEO: archive for high-throughput functional genomic data. Nucleic Acids Res, 37:D885–D890. http://dx.doi.org/10.1093/ nar/gkn764 PMID:18940857
42. Parkinson H, Kapushesky M, Shojatalab M et al. (2007). ArrayExpress – a public data- base of microarray experiments and gene expression profiles. Nucleic Acids Res, 35:D747–D750. http://dx.doi.org/10.1093/ nar/gkl995 PMID:17132828
Websites ArrayExpress: http://www.ebi.ac.uk/arrayexpress/
Gene Expression Omnibus: http://www.ncbi.nlm.nih.gov/geo/
International Cancer Genome Consortium: www.icgc.org/
The Cancer Genome Atlas: http://cancergenome.nih.gov/
The Encyclopedia of DNA Elements: http://encodeproject.org/ENCODE/
214
Summary • Epigenetics refers to all heri-
table changes in gene expres- sion and associated phenotypic traits that are not coded in the DNA sequence itself, and these changes are mediated by DNA methylation, histone modifica- tions, and non-coding RNAs.
• Epidemiological and laboratory- based studies either confirm or implicate that epigenetic chang- es are induced by environmen- tal and lifestyle factors and that they are involved in a variety of human cancers, and possibly in other chronic disorders.
• Epigenetic changes in cancer provide unique biomarkers, and those in non-cancer tissues can reflect past exposure to carcino- genic factors.
• The suppression of epigenetic changes can, in principle, be used for cancer prevention, and reversibility of these changes is now exploited in cancer treatment.
Recent advances in the field of epi- genetics have had a tremendous impact on our understanding of biological processes and complex
3.4 3 BIOLOGY
Epigenetics Toshikazu Ushijima Zdenko Herceg
Saverio Minucci (reviewer)
human diseases. As a result, aca- demic, medical, and public attention has turned to the potential applica- tion of these new advances in medi- cine and various fields of biomedical research. In addition to conceptual advances, the emergence of pow- erful technologies that enable the detection of epigenetic changes in high-throughput and genome-wide settings has dramatically acceler- ated cancer research and opened up new perspectives, resulting in a broader appreciation of the impor- tance of epigenetics in the etiology of human cancer.
In the past, the term “epigenet- ics” was used to describe all biologi- cal phenomena that do not follow normal genetic principles; nowa- days, the term refers to the study of all changes in gene expression transmitted across cell generations that do not involve changes in DNA sequence, i.e. mutations. One of the most remarkable recent discoveries is that different epigenetic phenom- ena share common underlying mo- lecular mechanisms. There are three main epigenetic mechanisms: DNA methylation, histone modifications, and non-coding RNAs (Fig. 3.4.1), all of which are known to be critical for high-fidelity propagation of gene activity states in a cell type-specific manner.
Consistent with the importance of epigenetic mechanisms in criti- cal cellular processes, dysregula- tion of epigenetic mechanisms has been linked to disease, most nota- bly cancer. Virtually all critical pro- cesses found in cancer cells, such as silencing of tumour suppressor genes, activation of oncogenes, ab- errant cell-cycle regulation, and de- fects in DNA repair, can be caused not only by genetic changes, includ- ing mutation, but also by epigenetic alterations. Accordingly, exploiting epigenetics may have potential in the prevention and treatment of cancer.
Epigenetic mechanisms The three main epigenetic mecha- nisms (DNA methylation, histone modifications, and RNA-mediated gene silencing) have been studied primarily in the context of gene ex- pression. However, they are increas- ingly recognized as important for other chromatin-based processes, such as DNA repair and replication, as well as in the integration of envi- ronmental signals [1]. The patterns of DNA methylation and histone modifications are cell type-specific and are propagated autonomously and with high fidelity over many cell generations.
The best-studied epigenetic mechanism is DNA methylation. The methylation of DNA refers to the
Chapter 3.4 • Epigenetics 215
3 B
IO LO
G Y
C H
A P
T E
R 3
.4
covalent addition of a methyl group to the 5-carbon position of cytosine in a CpG dinucleotide. DNA methylation has long been considered a highly stable epigenetic modification; how- ever, it is now accepted that DNA methylation marks may be more
dynamic than previously thought. Indeed, recent studies showed that the Tet family of proteins may be in- volved in active DNA demethylation. The capacity of Tet proteins to hy- drolyse methyl cytosines (producing 5-hydroxymethylcytosine) and to act
on fully methylated or hemi-methyl- ated DNA has been reported [2].
The second epigenetic mecha- nism encompasses various histone modifications. These are post- translational alterations of histone proteins that interact with DNA to
Fig. 3.4.1. Different types of epigenetic information. DNA methylation, histone modifications, and RNA-mediated gene silencing constitute three distinct mechanisms of epigenetic regulation. DNA methylation (Me) is a covalent modification of the cytosine (C) that is located 5′ to a guanosine (G) in a CpG dinucleotide. Histone (chromatin) modifications refer to covalent post-translational modifications of N-terminal tails of four core histones (H3, H4, H2A, and H2B). The most recently discovered mechanism of epigenetic inheritance involves RNAs, which in the form of microRNA (miRNA) acting with messenger RNA (mRNA) can alter gene expression states in a heritable manner.
216
form a complex known as chroma- tin. Histone modifications include acetylation, methylation, phos- phorylation, and ubiquitination of specific residues in the N-terminal tails of histones. Histone modifica- tions regulate several cellular pro- cesses, including gene transcription, DNA repair, and DNA replication [3]. There is growing evidence that aber- rant histone modifications and the dysregulation of gene products in- volved in these modifications are implicated in human neoplasms, and thus the importance of histone modifications in cancer and other diseases is becoming increasingly recognized [4].
The third class of epigenetic mechanisms is mediated by non- coding RNAs, in the form of either long non-coding RNAs or small RNAs (microRNAs), which also par- ticipate in the stable maintenance of gene activity states over cell divi- sions [5] (see “The non-coding RNA
revolution in medical research”). Experimental evidence suggests that there is intimate and mutually reinforcing cross-talk between the three epigenetic mechanisms in set- ting up and maintaining the genome- wide expression programme in a tissue-specific and lineage-specific manner.
Epigenome changes in cancer Consistent with the critical role of epigenetic mechanisms in the con- trol of cellular processes, a plethora of studies have revealed that the epi- genome is markedly dysregulated in virtually all malignancies [6]. Two forms of aberrant DNA methylation are found in human cancer: the over- all loss of 5-methylcytosine (global hypomethylation) and gene promot- er-associated (CpG island-specific) hypermethylation [7]. While the pre- cise consequences of genome-wide
hypomethylation, including activa- tion of cellular proto-oncogenes and induction of chromosome instability, are still debated, hypermethylation of gene promoters is associated with gene inactivation. When hyper- methylated, gene promoters become unable to bind the factors that are responsible for gene expression [8], and therefore the gene is not tran- scribed. A large number of studies have indicated that the silencing of tumour suppressor genes and other cancer-related genes may occur through hypermethylation of their promoters.
Histone modifications are also dysregulated in cancers. The iden- tification and functional charac- terization of chromatin modifying and remodelling complexes have revealed an important role for his- tone modifications in normal cellular processes and cancer development. Histone modification patterns have been found to modulate accessibility
The non-coding RNA revolution in medical research Pier Paolo Pandolfi
The existence of a previously unrec- ognized RNA language by which all RNA transcripts communicate in extensive interconnected networks has been proposed. The lone code currently available to decipher gene function, the triplet amino acid code described almost 50 years ago, can be applied to only 2% of the human genome. Recent work by several groups of investigators has chal- lenged the traditional view of RNAs as solely templates for protein synthesis, and suggested instead a model in which these complex molecules have important protein- independent functions, including a crucial role in post-transcriptional gene regulation. It may now be pos- sible to systematically functionalize the entire transcriptome, which en- compasses both protein-coding and non-coding transcripts. Notably, the transcriptome represents the ma- jority of the human genome and is heavily dysregulated in cancer.
A proposed RNA language is based on the hypothesis that tran- scripts regulate and communicate with each other by competing for common microRNAs. Transcripts acting in this fashion are termed com- peting endogenous RNAs (ceRNAs). CeRNA-mediated regulation may involve cross-talk between multi- ple unrelated RNA molecules in complex networks. This ceRNA language imparts a novel trans- regulatory function to transcripts independent of their protein-coding function, and may shed light on regulatory networks that have been overlooked by conventional protein- coding studies.
This hypothesis was ultimately confirmed by studying the interac- tion between the messenger RNA (mRNA) encoding for the PTEN tu- mour suppressor gene and its close- ly related pseudogene, PTENP1, which acts as a tumour suppres- sor through this new mechanism.
Extending the analysis to mRNAs has identified several new regula- tors of PTEN, such ZEB2 and VAPA, which are novel tumour suppressor molecules whose function was not previously associated with the proto- oncogenic PI3K/AKT pathway. More recently, work has focused on the BRAF pseudogene BRAFps and its role in promoting MAPK activation through regulation of its parental gene, BRAF. BRAFps is overex- pressed in human cancers and is oncogenic in mice engineered to express high levels of it. This trans- formative new dimension of gene regulation has significant implica- tions for the systematic functionali- zation of the entire transcriptome, and represents a paradigm shift for biomedical research at large.
Reference Karreth FA et al. (2011). Cell, 147:382–395. http://dx.doi.org/10.1016/j.cell.2011.09.032 PMID:22000016
Chapter 3.4 • Epigenetics 217
3 B
IO LO
G Y
C H
A P
T E
R 3
.4 to DNA during gene transcription and DNA repair. Dysregulation of histone modifications may promote induction of mutations and genomic instability in cancer. Interestingly, a recent study revealed that chro- matin organization has a major im- pact on regional mutation rates in human cancer. Furthermore, recent sequencing efforts have revealed that many genes involved in his- tone modifications and chromatin
remodelling are recurrently mutated in a wide range of human cancers. Many recent studies have also pro- vided evidence that the dysregula- tion of non-coding RNAs is involved in the development of human neo- plasia [6].
Although epigenetic changes have been implicated in differ- ent stages of tumour development and progression, the challenge is to identify functionally important
epigenetic changes, which may be referred to as “drivers” in the same way this term is used for mutations, and hence differentiated from “pas- senger” events, which are evident but not functionally important. Current epigenetic studies, includ- ing major international sequencing projects, are expected to generate information establishing the com- prehensive epigenomic profiles of human cancers. These findings seem likely to facilitate mechanistic studies, leading to the development of epigenetic therapies and new biomarkers.
Environmental factors Unlike the genome, which is replicat- ed identically in every single cell of an organism, the epigenome shows wide-ranging variability across dif- ferent cell types, and may also vary within populations of the same cell type under the influence of environ- mental stressors. Epigenetic mecha- nisms have been suggested to play critical roles in physiological re- sponses to environmental exposures (Fig. 3.4.2). Different changes in the epigenome are involved in mediat- ing gene expression programmes re- quired for an organism’s adaptation to environmental exposures. Studies in different model systems have pro- vided important information on the role of epigenetic mechanisms in response to environmental stresses [1,9–12].
Many physical and chemical carcinogens and relevant infectious agents, to which individuals are en- vironmentally exposed, are consid- ered to promote cancer development through disruption of epigenetically maintained gene expression pro- grammes. In addition, studies on monozygotic twins have highlighted epigenomic changes in response to environmental factors and associ- ated disease susceptibility. While the importance of environment in the de- velopment of a wide variety of can- cers is well supported by both epi- demiological and laboratory-based studies, the mechanisms by which en- vironmental exposures dysregulate
Fig. 3.4.2. Epigenetics as an interface between the genetic code and the environment. Epigenetic mechanisms regulate many cellular processes directly or indirectly and play critical roles in cellular responses to environmental and endogenous stimuli. There is intimate and self-reinforcing cross-talk between different types of epigenetic information. This is proposed to constitute the “epigenetic code” that modulates the genetic code in response to endogenous and environmental cues. The epigenetic code is important to maintain gene expression profiles and chromatin structure in a heritable manner over many cell generations and may dictate cellular outcomes by regulating cellular processes such as gene transcription, proliferation, and DNA repair. Dysregulation of epigenetic mechanisms may promote the development of abnormal phenotypes and diseases, including cancer.
218
the epigenome remain poorly char- acterized [1,9–13].
An interesting connection be- tween the epigenome and metabo- lism in cancer has recently emerged. For example, cells with mutations in the isocitrate dehydrogenase genes IDH1 and IDH2 (events that frequent- ly occur in human gliomas) exhibit accumulation of the oncometabo- lite d-2-hydroxyglutarate, resulting in inhibition of α-ketoglutarate de- pendent dioxygenases involved in DNA and histone demethylation [14]. Therefore, dysregulation of IDH1/
IDH2 represents a novel oncogenet- ic mechanism linking cellular metab- olism and epigenome dysregulation in cancer.
The development of powerful epigenomic technologies and the availability of large population-based cohorts offer excellent opportuni- ties to test the role of repeated and chronic exposure to environmental epimutagens and aberrant changes in the epigenome in the etiology of specific cancers. Prospects for epigenetic testing in the context of
carcinogen identification and evalu- ation may be discerned [15].
Epigenetic changes as novel biomarkers Epigenetic changes, especially DNA methylation, are expected to be use- ful as novel biomarkers. Before a tumour develops, DNA methylation changes can accumulate at high lev- els in normal-appearing tissues, un- like mutations [16], and methylation patterns often reflect past exposure to carcinogenic stimuli [17]. Also, the accumulation level of aberrant DNA methylation can be correlated with risk of cancer development [18]. Such a correlation has been well documented for gastric cancers, and has been implicated for many other cancers, including liver, colorectal, oesophageal, breast, and renal can- cers. Therefore, DNA methylation profiles and patterns of methylation in non-cancer tissues can be used as a unique biomarker for past ex- posure to carcinogens and the risk of future cancers (Fig. 3.4.3).
Sites of DNA methylation in par- ticular cancers can be used as a biomarker to detect cancer cells or cancer cell-derived DNA and also to elucidate the pathophysiology of tumours. DNA methylation can be sensitively detected by technolo- gies based on polymerase chain reaction; DNA itself is more stable than RNA or proteins. In addition, most cancer cells have an ample number of aberrantly methylated genes [16]. Therefore, DNA meth- ylation has multiple advantages as a biomarker to detect cancer cells or cancer cell-derived DNA. Once specificity of DNA methylation of a given gene in a population of can- cer cells is established, such aber- rant DNA methylation can be de- tected sensitively and stably. Such cancer detection systems are already commercially available, and multi- ple targets are under development [19,20].
The pathophysiology of tumours, as determining or relevant to patient prognosis, responsiveness to thera- py, and the presence of lymph node
Fig. 3.4.3. DNA methylation as a biomarker for past exposure and future risk of cancers. Exposure to specific carcinogens, such as Helicobacter pylori infection and tobacco smoke, is known to induce aberrant DNA methylation of specific genes in normal- appearing tissue, which can be used as biomarkers for past exposure. The degree of epigenome damage, measured as the level of aberrant DNA methylation (the fraction of cells with methylation), correlates with risk of cancer development for some types of cancers, and can be used as a cancer risk marker.
Chapter 3.4 • Epigenetics 219
3 B
IO LO
G Y
C H
A P
T E
R 3
.4
metastasis, can also be characterized using DNA methylation. In sharp distinction from patterns of gene ex- pression, DNA methylation may indi- cate that a particular gene cannot be expressed even if its expression is induced in the future (Fig. 3.4.4). For example, if the promoter region of O6-methylguanine methyltransferase is determined to be methylated at biopsy of a brain tumour, this gene will never be expressed even after future chemotherapy involving an alkylating agent, and in the absence of such repair, such chemotherapy will be effective. In contrast, if the same gene is not expressed simply because of the lack of its induction at biopsy, its expression can be in- duced upon chemotherapy, and the chemotherapy may be ineffective [21].
DNA methylation of multiple genes (the CpG island methylator phenotype) is associated with
patient prognosis in several types of cancers, including colorectal and gastric malignancies, as well as neuroblastomas. Specifically, the CpG island methylator phenotype in neuroblastoma provides prognostic information that is more precise than amplification of the MYCN onco- gene, one of the clearest prognostic indicators in clinical oncology [22].
Epigenetic therapy and prevention One of the most important as- pects of epigenetic change, which distinguishes such change from mutation, is reversibility induced by drugs [23,24]. DNA methyla- tion can be reversed by the action of DNA demethylating agents, and two such drugs have been approved by the United States Food and Drug Administration and are used for treatment of haemato-
logical disorders [25]. Notably, using these two drugs, clinical responses were achieved in patients with my- elodysplastic syndrome, for whom blood transfusions used to be the sole effective treatment. Histone deacetylation can be reversed by histone deacetylase inhibitors, and in this context two drugs have been approved for treatment of cutaneous lymphoma [26]. The mode of action of epigenetic drugs is different from that of cytotoxic drugs, and low- dose and long-term administration is seen to be important, especially for DNA demethylating agents [23]. Recent cancer genome studies have revealed previously unknown muta- tions of epigenetic regulators, and these mutated regulators are ex- pected to provide novel therapeu- tic targets in the future [24]. Other potential targets for therapy may be microRNAs (see “Causes and con- sequences of microRNA dysregula- tion in cancer”).
The reversibility of epigenetic changes is also useful for cancer prevention. As a proof of concept, in experimental animals tumours such as those of the colon and prostate have been suppressed by DNA demethylating agents [27,28]. However, it must be recognized that DNA methylation is physiologically essential to repress some genes, and nonspecific demethylation is ex- pected to lead to long-term adverse effects [29]. Therefore, to render epigenetic cancer prevention prac- ticable for the human population, the specificity of preventive agents for genes with aberrant epigenetic mod- ifications must be improved. Also, it is now possible to identify individuals at extremely high risk of some can- cers by assessing accumulation levels of aberrant DNA methylation. These individuals represent a popu- lation likely to benefit from effective chemoprevention. Since epigenetic cancer prevention has great poten- tial, multiple relevant studies are re- quired in a timely manner.
Fig. 3.4.4. Detection of inability to be expressed. The statuses of being unable to be expressed and of not being expressed are different. For example, expression of the DNA repair protein O6-methylguanine methyltransferase can be repressed by DNA methylation or by the lack of its transcriptional activation. If O6-methylguanine meth- yltransferase expression was repressed by DNA methylation at the time of biopsy, it cannot be induced even by implementation of chemotherapy with an alkylating agent. In contrast, if expression of the gene was repressed by lack of transcriptional activa- tion, it can be induced by chemotherapy. Since O6-methylguanine methyltransferase repairs DNA damage introduced by the alkylating agent, the effect of chemotherapy is diminished by its induction. DNA methylation can predict whether such induction of gene expression will happen. Pol II, RNA polymerase II.
220
Causes and consequences of microRNA dysregulation in cancer Carlo M. Croce
Since the discovery of miR-15a and miR-16-1 deletions in chronic lym- phocytic leukaemia, many labora- tories around the world have shown microRNA (miRNA) dysregulation in all tumours studied, including the most common, such as lung, breast, prostate, and gastrointestinal can- cers. Such dysregulation, like the dysregulation of oncogenes and tumour suppressor genes, can be caused by multiple mechanisms, such as deletion, amplification, mu- tation, transcriptional dysregulation, and epigenetic changes.
Since miRNAs have multiple tar- gets, their function in tumorigenesis could be due to their regulation of a few specific targets – possibly of even one – or of many. A future challenge will be identifying all of the
targets of the miRNAs involved in cancer and establishing their con- tributions to malignant transforma- tion. An additional challenge will be identifying all of the miRNAs that are dysregulated by pathways that are consistently dysregulated in vari- ous types of human cancers. This point is of particular importance be- cause instead of focusing on specific alterations in protein-coding onco- genes or tumour suppressor genes – which may be difficult to treat – the focus may be on their downstream miRNA targets. If these miRNA tar- gets are crucial for the expression of the malignant phenotype and the cancer cells depend on their dysreg- ulation for proliferation and survival, the use of miRNAs or anti-miRNAs
may be expected to result in tumour regression.
Genomic analyses for alterations in miRNA genes or for copy num- ber alterations in various human tumours by deep sequencing are in progress but have not been com- pleted. These studies could provide additional information about the involvement of miRNAs in cancer and in many other diseases. Over the past few years, a shift from con- ventional chemotherapy to targeted therapies has occurred, and miRNAs and anti-miRNAs will contribute extensively to targeted therapies.
Reference Iorio MV et al. (2012). Cancer J, 18:215–222. ht t p: //d x .d o i .o r g /10 .10 97/ P P O. 0 b 013 e 318250c001 PMID:22647357
Chapter 3.4 • Epigenetics 221
3 B
IO LO
G Y
C H
A P
T E
R 3
.4
References 1. Feil R, Fraga MF (2011). Epigenetics and the
environment: emerging patterns and impli- cations. Nat Rev Genet, 13:97–109. http:// dx.doi.org/10.1038/nrg3142 PMID:22215131
2. Tahiliani M, Koh KP, Shen Y et al. (2009). Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mamma- lian DNA by MLL partner TET1. Science, 324:930–935. http://dx.doi.org/10.1126/ science.1170116 PMID:19372391
3. Bannister AJ, Kouzarides T (2011). Regulation of chromatin by histone modifi- cations. Cell Res, 21:381–395. http://dx.doi. org/10.1038/cr.2011.22 PMID:21321607
4. Sawan C, Herceg Z (2010). Histone modi- fications and cancer. Adv Genet, 70:57– 85. http://dx.doi.org/10.1016/B978-0-12- 380866-0.60003-4 PMID:20920745
5. Pauli A, Rinn JL, Schier AF (2011). Non- coding RNAs as regulators of embryogen- esis. Nat Rev Genet, 12:136–149. http:// dx.doi.org/10.1038/nrg2904 PMID:21245830
6. Rodríguez-Paredes M, Esteller M (2011). Cancer epigenetics reaches mainstream on- cology. Nat Med, 17:330–339. http://dx.doi. org/10.1038/nm.2305 PMID:21386836
7. Sinčić N, Herceg Z (2011). DNA meth- ylation and cancer: ghosts and angels above the genes. Curr Opin Oncol, 23:69–76. http://dx.doi.org/10.1097/CCO. 0b013e3283412eb4 PMID:21119515
8. Vaissière T, Sawan C, Herceg Z (2008). Epigenetic interplay between histone mod- ifications and DNA methylation in gene silencing. Mutat Res, 659:40–48. http:// dx.doi.org/10.1016/j.mrrev.2008.02.004 PMID:18407786
9. Herceg Z (2007). Epigenetics and cancer: towards an evaluation of the impact of envi- ronmental and dietary factors. Mutagenesis, 22:91–103. http://dx.doi.org/10.1093/muta ge/gel068 PMID:17284773
10. Herceg Z, Vaissière T (2011). Epigenetic mechanisms and cancer: an interface be- tween the environment and the genome. Epigenetics, 6:804–819. http://dx.doi.org/ 10.4161/epi.6.7.16262 PMID:21758002
11. Herceg Z, Paliwal A (2011). Epigenetic mechanisms in hepatocellular carcinoma: how environmental factors influence the epi- genome. Mutat Res, 727:55–61. http://dx. d o i.o r g /10.1016 / j.m r r ev. 2 011.0 4.0 01 PMID:21514401
12. Hou L, Zhang X, Wang D, Baccarelli A (2012). Environmental chemical exposures and human epigenetics. Int J Epidemiol, 41:79–105. http://dx.doi.org/10.1093/ije/ dyr154 PMID:22253299
13. Ushijima T, Hattori N (2012). Molecular pathways: involvement of Helicobacter pylori-triggered inflammation in the forma- tion of an epigenetic field defect, and its usefulness as cancer risk and exposure markers. Clin Cancer Res, 18:923–929. http://dx.doi.org/10.1158/1078-0432.CCR- 11-2011 PMID:22205689
14. Xu W, Yang H, Liu Y et al. (2011). Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate- dependent dioxygenases. Cancer Cell, 19:17–30. http://dx.doi.org/10.1016/j.ccr. 2010.12.014 PMID:21251613
15. Herceg Z, Lambert MP, van Veldhoven K et al. (2013). Towards incorporating epige- netic mechanisms into carcinogen identi- fication and evaluation. Carcinogenesis, 34:1955–1967. http://dx.doi.org/10.1093/ carcin/bgt212 PMID:23749751
16. Ushijima T, Asada K (2010). Aberrant DNA methylation in contrast with mutations. Cancer Sci, 101:300–305. http://dx.doi. org/10.1111/j.1349 -7006.2009.01434.x PMID:19958364
17. Takeshima H, Ushijima T (2010). Methylation destiny: Moira takes account of histones and RNA polymerase II. Epigenetics, 5:89– 95. http://dx.doi.org/10.4161/epi.5.2.10774 PMID:20160507
18. Ushijima T (2007). Epigenetic field for cancerization. J Biochem Mol Biol, 40:142–150. http://dx.doi.org/10.5483/BMB Rep.2007.40.2.142 PMID:17394762
19. Liloglou T, Bediaga NG, Brown BR et al. (2012). Epigenetic biomarkers in lung can- cer. Cancer Lett, http://dx.doi.org/10.1016/j. canlet.2012.04.018 PMID:22546286
20. Nogueira da Costa A, Herceg Z (2012). Detection of cancer-specific epigenomic changes in biofluids: powerful tools in biomarker discovery and application. Mol Oncol, 6:704–715. http://dx.doi.org/10.1016/j. molonc.2012.07.005 PMID:22925902
21. Jacinto FV, Esteller M (2007). Mutator path- ways unleashed by epigenetic silencing in human cancer. Mutagenesis, 22:247–253. http://dx.doi.org/10.1093/mutage/gem009 PMID:17412712
22. Abe M, Ohira M, Kaneda A et al. (2005). CpG island methylator phenotype is a strong determinant of poor prognosis in neuroblastomas. Cancer Res, 65:828–834. PMID:15705880
23. Boumber Y, Issa JP (2011). Epigenetics in cancer: what’s the future? Oncology (Williston Park), 25:220 –226, 228. PMID:21548464
24. Dawson MA, Kouzarides T (2012). Cancer epigenetics: from mechanism to therapy. Cell, 150:12–27. http://dx.doi.org/10.1016/j. cell.2012.06.013 PMID:22770212
25. Fahy J, Jeltsch A, Arimondo PB (2012). DNA methyltransferase inhibitors in can- cer: a chemical and therapeutic patent overview and selected clinical studies. Expert Opin Ther Pat, 22:1427–1442. http:// dx.doi.org/10.1517/13543776.2012.729579 PMID:23033952
26. Batty N, Malouf GG, Issa JP (2009). Histone deacetylase inhibitors as anti-neoplastic agents. Cancer Lett, 280:192–200. http:// dx.doi.org/10.1016/j.canlet.2009.03.013 PMID:19345475
27. McCabe MT, Low JA, Daignault S et al. (2006). Inhibition of DNA methyltrans- ferase activity prevents tumorigenesis in a mouse model of prostate cancer. Cancer Res, 66:385–392. http://dx.doi. org/10.1158/0008 - 5472.CAN - 05 -2020 PMID:16397253
28. Yoo CB, Chuang JC, Byun HM et al. (2008). Long-term epigenetic therapy with oral zebularine has minimal side effects and prevents intestinal tumors in mice. Cancer Prev Res (Phila), 1:233–240. http://dx.doi. org/10.1158/1940- 6207.CAPR- 07- 0008 PMID:19138966
29. Bojang P Jr, Ramos KS (2013). The prom- ise and failures of epigenetic therapies for cancer treatment. Cancer Treat Rev, http:// dx.doi.org/10.1016/j.ctr v. 2013.05.0 09 PMID:23831234
222
Summary
• Cancers are increasingly seen as metabolic diseases, as ex- emplified by several oncogenes coding for metabolic enzymes and the impact of metabolic im- balance associated with obesity and other lifestyle factors known to increase cancer risk.
• The metabolome is the sum of low-molecular-weight metabo- lites in a particular sample and at a given time. Metabolomics is the systematic study of varia- tions in the metabolome in par- ticular situations, consequent upon various environmental or health conditions.
• Metabolomics provides a glob- al view of metabolism and fa- cilitates discovery of the role of metabolic pathways essential for cancer cell proliferation and the identification of new biomarkers for cancer diagnosis, prognosis, and recurrence.
• Metabolomics applied to cohorts in the context of metabolome- wide association studies should contribute to the identification of novel risk factors for cancer.
3.5 3 BIOLOGY
Metabolic change and metabolomics Augustin Scalbert Isabelle Romieu
James R. Krycer (reviewer)
Genetic alterations have been con- sidered as critical to the develop- ment of cancers. However, close connections also exist between metabolic change and cancer devel- opment. From one perspective, the metabolism of tumour cells differs from that of normal cells, and sev- eral oncogenes code for metabolic enzymes or alter the expression of these enzymes. From another per- spective, associations between life- style, metabolism, and cancer risk are increasingly being reported. Obesity, which results in major met- abolic imbalance, increases the risk of several malignancies, including colon, postmenopausal breast, en- dometrial, kidney, liver, pancreatic, thyroid, and oesophageal cancers. Cancers are thus increasingly seen as metabolic diseases. However, the contribution of metabolic dysfunc- tion to the onset or development of cancer remains poorly understood.
The development of metabo- lomics and major progress in related analytical techniques have contrib- uted to an improved understanding of the links between metabolism and cancer. Hundreds or thousands of metabolites can now be measured simultaneously in biospecimens to give a highly detailed picture of metabolic profiles in tumour tissues or biofluids. Metabolomics combines these highly sensitive analytical
techniques with powerful methods for data analysis to characterize subtle metabolic changes associ- ated with cancer. This approach has already led to the discovery of new biomarkers for cancer and to a better understanding of the mechanisms underlying the etiology of cancers.
Metabolism and cancer Salient features of cancer cells com- pared with normal cells are their high rates of glycolysis and glutamine metabolism, which allow malignant cells to meet their high energy and anabolic requirements. Production of lactic acid from glucose, even under non-hypoxic conditions, as commonly exhibited by tumour tis- sue is a feature that has been recog- nized for more than 70 years and is known as the Warburg effect. Such metabolism offers several advan- tages to cancer cells, including a high rate of adenosine triphosphate (ATP) production and the production of intermediate metabolites like cit- rate in a truncated tricarboxylic acid cycle to feed biosynthetic pathways and support the production of mac- romolecules for cell proliferation, of- ten resulting in a fine-tuning of these metabolic pathways. Otto Warburg sought to explain this effect as the result of damage to respiratory pro- cesses. However, it is now known
Chapter 3.5 • Metabolic change and metabolomics 223
3 B
IO LO
G Y
C H
A P
T E
R 3
.5
that dysregulation of glycolysis is a better explanation [1].
Connections between some ma- jor oncogenes or tumour suppressor genes and metabolism have been identified, and this finding has fo- cused attention on metabolic dys- regulation as a possible key mecha- nism leading to cancer. Mutations in the tumour suppressor gene TP53 enhance glucose transport and me- tabolism, whereas mutations in the oncogenic transcription factor MYC enhance the expression of most gly- colytic enzymes, lactase dehydroge- nase A, and enzymes required for nucleotide biosynthesis and one- carbon metabolism [2].
Oncogenes also play a key role in enabling cancer cells to adapt to environmental changes and adjust their metabolism accordingly. At an early stage of tumorigenesis, the on- cogene ERBB2 was demonstrated to commit the metabolism of cancer cells detached from their normal ma- trix towards proliferation rather than apoptosis [3]. Detached cells usually undergo apoptosis and autophagy,
but overexpression of this oncogene rescues the cancer cells, restores glucose transport and the pentose phosphate pathway, and improves the redox status of the cells.
Metabolic genes associated with aggressive cancer and stem cell character were systematically investigated within a list of about 2700 genes encoding all known hu- man metabolic enzymes and trans- porters [4]. Some of these genes had previously been shown to be highly expressed in tumours com- pared with normal tissues and also highly expressed in stem cells as assessed in a human breast cancer xenograft model. Sixteen of these genes were demonstrated to be required for tumorigenesis, includ- ing genes encoding transporters for mitochondrial ATP and for lactate as well as enzymes contributing to the glycolysis, pentose phosphate, and nucleotide biosynthesis pathways. This study also established the key role played by phosphoglycerate de- hydrogenase, the first enzyme in the serine biosynthetic pathway, which drives the conversion of glutamine to tricarboxylic acid cycle intermedi- ates for anabolic processes.
The role of metabolic alterations and adaptations in tumorigenicity is thus increasingly recognized. These metabolic alterations are now con- sidered a new hallmark of cancer [5]. This has led to the development of new therapeutic approaches tar- geted at the modulation of corre- sponding signalling pathways or at the direct inhibition of some of the enzymes in the metabolic pathways involved in tumorigenicity. These approaches include inhibition of glycolysis, amino acid biosynthe- sis, and the mTOR and phosphati- dylinositol 3-kinase (PI3K) signalling pathways [6].
Understanding the interactions between altered metabolism and cancer biology is important not only for therapeutic intervention for cancer but also for prevention of malignant development. Obesity and certain metabolic characteristics (the metabolic syndrome) are well- established risk factors for several
cancers, and various mechanisms have been proposed to explain such interactions [7]. These include the di- rect effects of adiposity on the secre- tion of pro-inflammatory cytokines and adipokines, and indirect effects such as insulin resistance and high insulinaemia commonly associated with obesity or an increase in the lev- els of steroid hormones. Alterations of metabolic profiles associated with obesity (e.g. excess circulating free fatty acids), other lifestyle factors, or some genetic factors may also increase cancer risk.
Metabolomics, through an in- depth characterization of metabolic profiles in cells, tissues, and human biofluids, may help elucidate both mechanisms of carcinogenesis and those underlying associations between lifestyle and cancer risk. Metabolomics can also be used to identify novel early biomarkers for cancer and to characterize new risk factors for cancer.
Metabolomics to better understand mechanisms of carcinogenesis The metabolome – the sum of low-molecular-weight metabolites at a given time – is the most fun- damental biochemical indicator of biological systems and is an in- dicator of both genetic and envi- ronmental conditions (Fig. 3.5.2). Metabolomics has been defined as the systematic study of variations of the metabolome under particular conditions [7] and involves assess- ing the metabolome in blood, urine, or tissues using modern and pow- erful analytical techniques based on mass spectrometry or nuclear magnetic resonance (NMR) spec- troscopy. The sensitivity of these analytical techniques and the large number of metabolites detected al- low the characterization of biologi- cal phenotypes with unprecedented precision, the identification of novel biomarkers for diseases and expo- sure to disease risk factors, and a better understanding of the etiology of diseases (Fig. 3.5.3).
Fig. 3.5.1. Campaign poster from the Collectif National des Associations d’Obèses (France): “Obesity kills. Do you still think it’s funny? Obesity is a serious condition that kills 55 000 people per year in France. It is neither a fault nor an inevi- tability, still less a joke.” Obesity has been shown to increase susceptibility to diabe- tes and cancer.
224
Two different metabolomic ap- proaches can be distinguished. In the targeted approach, sets of tens or hundreds of metabolites identified a priori are quantified simultaneously in a single analytical operation, most often by mass spectrometry. In the untargeted approach, hundreds (with NMR spectroscopy) or thousands (with mass spectrometry) of me- tabolites are detected in individual profiles from large sets of samples (Fig. 3.5.4). Metabolites overex- pressed or underexpressed in differ- ent groups of samples (e.g. plasma samples from case and control sub- jects) are identified using multivari- ate statistics and their structures es- tablished a posteriori by comparing their spectra to those stored in large metabolite databases.
When applied to tumour cells or animal models of cancers, metabo- lomics, through the wealth of data generated in a typical experiment, allows the formulation of novel hy- potheses, which can then be tested in hypothesis-driven targeted ex- periments. Two recent publications show how metabolomics contributed to establishing the oncogenicity of glycine and glycine decarboxylase. In a metabolic survey of 60 primary human cancer cell lines from 9 com- mon tumour types, 111 metabolites were measured in the cell culture me- dium [8]. About two thirds of the me- tabolites were found to be secreted
and one third consumed during cell growth. Hierarchical clustering analy- sis of the metabolites permitted the recognition of several clusters related to different metabolic pathways (gly- colysis, tricarboxylic acid cycle, nu- cleotide and polyamine metabolism) with no major distinction according to tissue of origin. Glycine consump- tion was found to be highly correlated to proliferation of transformed cells. This unique observation led to the recognition of the key role played by the mitochondrial glycine synthesis pathway in cancer cell proliferation.
Further confirmation of the role of glycine in cell proliferation was derived from another study, on stem cells from human lung tumours [9]. In this investigation, metabolomics was applied to characterize the role of a specific enzyme, glycine decarboxylase, in overall metabo- lism. Glycine decarboxylase was demonstrated to be one of the most upregulated genes in stem cells compared with other lung tumour cells. This enzyme catalyses the breakdown of glycine to the methyl donor 5,10-methylenetetrahydro- folate, which is required for the syn- thesis of nucleotides. Comparison of metabolic profiles of several cell lines overexpressing glycine de- carboxylase or with this enzyme knocked down revealed alterations in the glycine pathway, glycolysis, and pyrimidine synthesis. The on- cogenic character of glycine decar- boxylase was further confirmed by higher mortality among non-small cell lung cancer patients exhibiting high expression of this enzyme in their tumours.
Metabolomics to identify novel biomarkers for cancer Metabolomics has been used in various case–control studies to identify novel biomarkers for cancer
Fig. 3.5.2. The metabolome is the most downstream biochemical expression of the phenotype and is influenced by both genetic and environmental factors.
Fig. 3.5.3. The human metabolomes and the applications of metabolomics to cancer research.
Chapter 3.5 • Metabolic change and metabolomics 225
3 B
IO LO
G Y
C H
A P
T E
R 3
.5
diagnosis, prognosis, or recurrence. New biomarkers were discovered by comparing metabolic profiles in tu- mours and benign adjacent tissues or blood or urine samples from patients and matched controls. Good exam- ples of these novel biomarkers include hydroxylated ultra long-chain fatty acids for colorectal cancer [10] and 27-nor-5β-cholestane-3,7,12,24,25 pentol glucuronide for ovarian cancer [11]. In these two particular studies, serum samples were analysed by high-resolution mass spectrometry to detect features in metabolic profiles that provided for clearest discrimi- nation between cases and controls. Relevant chemical structures were established by mass spectrometry and NMR spectroscopy.
The potential presented by me- tabolomics may be evident with reference to particular tumours. Because of its asymptomatic nature, hepatocellular carcinoma is usually diagnosed at late and advanced stages. For this tumour type, emerg- ing high-throughput metabolomics technologies have been applied to the discovery of candidate biomark- ers for cancer staging, prediction of recurrence and prognosis, and treat- ment selection [12].
Certain novel biomarkers are rel- atively rare metabolites. More com- mon metabolites, when assembled in marking sets (metabolic signa- tures), may also constitute powerful biomarkers for cancer. Such marking sets characteristic of particular can- cers have also been identified using metabolomics. Amino acid profiles in plasma were shown to permit the discrimination of patients with five different cancers from matched con- trols regardless of disease stage [13]. These characteristic profiles were also evident at an early stage, showing the value of these amino acids as biomarkers for detection of disease. Some of these amino acids were systematically increased or de- creased independent of the cancer site, suggesting the existence of a generic metabolic signature for cancer. Other metabolomic studies also showed consistent alteration in the level of other metabolites.
A systematic analysis of metabolic signatures from 117 preclinical and clinical metabolomic studies revealed that, besides alterations in amino acids, metabolism of glu- tathione, bile acids, galactose, triglycerides, phospholipids, and nucleotides as well as ammonia re- cycling and the glycolytic and glu- coneogenic pathways were consis- tently affected in cancers [14].
In all such metabolomic studies, a key step is biomarker validation. The large number of metabolic vari- ables and the low number of sub- jects (commonly 50–200, depend- ing on the study) increase the risk of false discovery. It is therefore es- sential to validate the new biomark- ers in an independent population. Unfortunately, validation has been largely overlooked in most metabo- lomic studies published thus far. The identification of sarcosine as a biomarker for prostate cancer [15] raised expectations of a possible way to improve diagnosis, but this result could not be validated in a later study [16]. Examples of good
study design for biomarker valida- tion can be recognized in a few me- tabolomic studies [10,11]. To limit the risk of false discovery and better ap- praise the value of biomarkers, it will be important to gain further insight into the potential confounding effects of factors such as body mass, age, gender, postmenopausal status, mi- crobiome, drug treatment, or smok- ing – all known to affect the meta- bolome – and then conduct studies taking into account the influence of these factors.
Metabolomics to identify risk factors for cancer in cohort studies Metabolomics is a powerful means to compare individual phenotypes in populations. A good illustration of the power of the approach is given by its application to the identification of the invariant part of the metabolome in a given individual. The analyses of the metabolome by NMR spectros- copy in a small set of urine samples from 22 subjects allowed the rec- ognition, with 100% confidence, of
Fig. 3.5.4. Metabolomic workflow. MS, mass spectrometry; NMR, nuclear magnetic resonance spectroscopy.
226
the donor of the samples [17]. The rapid development of more sensitive mass spectrometry techniques now permits the generation of metabolite profiles with much greater detail and the comparison of individual pheno- types in large groups of subjects, as is done in most epidemiological studies.
Several thousands of different metabolites constituting the human metabolome have been described in diverse biospecimens. So far, this considerable amount of metabolic information has been exploited to a very limited extent. The human me- tabolome includes both a stable and an unstable fraction. These were characterized by NMR spectros- copy in repeated plasma and urine samples collected 4 months apart, and the stable fraction was found to account for 60% and 47% of the plasma and urine metabolomes, re- spectively [18]. The unstable fraction reflects exposures to environmental and lifestyle factors, which occur only episodically. This fraction is not easily exploited in epidemiological studies unless repeated samples are available.
In contrast, the stable fraction captures both genetic and envi- ronmental variations of individual phenotypes, and its evolution over a lifetime can be described. Its mea- surement opens up wide perspec- tives to study the prevalence and risk of diseases in cohort studies. The
first metabolome-wide association studies (MWAS) (to be compared to the well-established genome-wide association studies) have recently been published. These studies are aimed at identifying the metabo- lites and metabolic pathways most strongly associated with a disease or with intermediate end-points. In a case–control study nested in the Framingham Offspring cohort, 61 metabolites were measured in plas- ma samples collected at baseline (average follow-up, 12 years; 189 cases and 189 controls), and 5 ami- no acids were identified that were strongly associated with the risk of type 2 diabetes [19]. These amino acids include three branched-chain amino acids previously known for their ability to promote insulin resis- tance. No such MWAS studies have yet been published on cancer.
The fraction of the human meta- bolome that varies with environmen- tal exposure rather than genetic fac- tors has been called the exposome [20]. The exposome notably reflects dietary exposure. Characteristic metabolic features could be identi- fied in vegetarians and in low and high meat consumers, or in sub- jects consuming specific foods [21]. Variations in the urinary metabolome across countries as measured in the International Population Study on Macro/Micronutrients and Blood Pressure (INTERMAP) cohort were found to be partly explained
by metabolites directly derived from foods consumed in various amounts in these countries [22,23]. Measurement of the exposome in exposome-wide association stud- ies (EWAS) should enable the iden- tification of the environmental fac- tors most strongly associated with cancer risk [21].
The few examples given here show the great potential of me- tabolomics for cancer research. Such studies have already con- tributed to elucidating the role of unexpected metabolic pathways of major importance in carcinogenesis. Investigation of metabolic processes has enabled the identification of nov- el biomarkers for cancer diagnosis. The application of metabolomics to molecular epidemiology also shows great potential to identify novel risk factors for cancer.
Metabolomics, among all “omics” technologies, is the one most recently introduced into biological and medi- cal research. Techniques for metabo- lomics, although not yet fully mature, have considerably improved over the past few years, and there is now little doubt that metabolomics will emerge as an essential tool, complementary to the well-established genomics, transcriptomics, and proteomics ap- proaches, to identify novel biomark- ers for cancer and to better under- stand the etiology of the disease.
Chapter 3.5 • Metabolic change and metabolomics 227
3 B
IO LO
G Y
C H
A P
T E
R 3
.5
References 1. Koppenol WH, Bounds PL, Dang CV
(2011). Otto Warburg’s contributions to current concepts of cancer metabolism. Nat Rev Cancer, 11:325–337. http://dx.doi. org/10.1038/nrc3038 PMID:21508971
2. DeBerardinis RJ (2008). Is cancer a disease of abnormal cellular metabolism? New angles on an old idea. Genet Med, 10:767–777. http://dx.doi.org/10.1097/GIM. 0b013e31818b0d9b PMID:18941420
3. Schafer ZT, Grassian AR, Song L et al. (2009). Antioxidant and oncogene rescue of metabolic defects caused by loss of matrix attachment. Nature, 461:109–113. http://dx.doi.org/10.1038/nature08268 PMID:19693011
4. Possemato R, Marks KM, Shaul YD et al. (2011). Functional genomics reveal that the serine synthesis pathway is essential in breast cancer. Nature, 476:346–350. http://dx.doi.org/10.1038/nature10350 PMID:21760589
5. Hanahan D, Weinberg RA (2011). Hallmarks of cancer: the next generation. Cell, 144:646–674. http://dx.doi.org/10.1016/j. cell.2011.02.013 PMID:21376230
6. Dang NH, Singla AK, Mackay EM et al. (2013). Targeted cancer therapeutics: biosynthetic and energetic pathways char- acterized by metabolomics and the inter- play with key cancer regulatory factors. Curr Pharm Des, [epub ahead of print]. PMID:23859615
7. Khandekar MJ, Cohen P, Spiegelman BM (2011). Molecular mechanisms of cancer development in obesity. Nat Rev Cancer, 11:886–895. http://dx.doi.org/10.1038/nrc 3174 PMID:22113164
8. Jain M, Nilsson R, Sharma S et al. (2012). Metabolite profiling identifies a key role for glycine in rapid cancer cell prolif- eration. Science, 336:1040 –1044. http:// d x .d o i.o r g /10.112 6 /s c i e n c e.1218 5 9 5 PMID:22628656
9. Zhang WC, Shyh-Chang N, Yang H et al. (2012). Glycine decarboxylase activity drives non-small cell lung cancer tumor- initiating cells and tumorigenesis. Cell, 148:259–272. http://dx.doi.org/10.1016/j. cell.2011.11.050 PMID:22225612
10. Ritchie SA, Ahiahonu PWK, Jayasinghe D et al. (2010). Reduced levels of hydrox- ylated, polyunsaturated ultra long-chain fatty acids in the serum of colorectal cancer patients: implications for early screening and detection. BMC Med, 8:13. http://dx.doi.org/10.1186/1741-7015-8-13 PMID:20156336
11. Chen J, Zhang X, Cao R et al. (2011). Serum 27-nor-5β-cholestane-3,7,12,24,25 pentol glucuronide discovered by metabolomics as potential diagnostic biomarker for epi- thelium ovarian cancer. J Proteome Res, 10:2625–2632. http://dx.doi.org/10.1021/ pr200173q PMID:21456628
12. Wang X, Zhang A, Sun H (2013). Power of metabolomics in diagnosis and biomarker discovery of hepatocellular carcinoma. Hepatology, 57:2072–2077. http://dx.doi. org/10.1002/hep.26130 PMID:23150189
13. Miyagi Y, Higashiyama M, Gochi A et al. (2011). Plasma free amino acid profiling of five types of cancer patients and its ap- plication for early detection. PLoS One, 6:e24143. http://dx.doi.org/10.1371/journal. pone.0024143 PMID:21915291
14. Ng DJY, Pasikanti KK, Chan ECY (2011). Trend analysis of metabonomics and sys- tematic review of metabonomics-derived cancer marker metabolites. Metabolomics, 7:155–178. http://dx.doi.org/10.1007/s113 06-010-0250-7
15. Sreekumar A, Poisson LM, Rajendiran TM et al. (2009). Metabolomic profiles delin- eate potential role for sarcosine in prostate cancer progression. Nature, 457:910–914. http://dx.doi.org/10.1038/nature07762 PMID:19212411
16. Jentzmik F, Stephan C, Miller K et al. (2010). Sarcosine in urine after digital rectal exami- nation fails as a marker in prostate cancer detection and identification of aggressive tumours. Eur Urol, 58:12–18, discussion 20–21. http://dx.doi.org/10.1016/j.eururo. 2010.01.035 PMID:20117878
17. Assfalg M, Bertini I, Colangiuli D et al. (2008). Evidence of different metabolic phenotypes in humans. Proc Natl Acad Sci U S A, 105:1420–1424. http:// dx.doi.org/10.1073/pnas.0705 6 85105 PMID:18230739
18. Nicholson G, Rantalainen M, Maher AD et al.; The MolPAGE Consortium (2011). Human metabolic profiles are stably con- trolled by genetic and environmental vari- ation. Mol Syst Biol, 7:525. http://dx.doi. org/10.1038/msb.2011.57 PMID:21878913
19. Wang TJ, Larson MG, Vasan RS et al. (2011). Metabolite profiles and the risk of developing diabetes. Nat Med, 17:448– 453. http://dx.doi.org/10.1038/nm.2307 PMID:21423183
20. Wild CP (2005). Complementing the genome with an “exposome”: the outstanding chal- lenge of environmental exposure measure- ment in molecular epidemiology. Cancer Epidemiol Biomarkers Prev, 14:1847–1850. http://dx.doi.org/10.1158/1055-9965.EPI- 05-0456 PMID:16103423
21. Stella C, Beckwith-Hall B, Cloarec O et al. (2006). Susceptibility of human meta- bolic phenotypes to dietary modulation. J Proteome Res, 5:2780–2788. http://dx.doi. org/10.1021/pr060265y PMID:17022649
22. Holmes E, Loo RL, Stamler J et al. (2008). Human metabolic phenotype diversity and its association with diet and blood pres- sure. Nature, 453:396– 400. http://dx.doi. org/10.1038/nature06882 PMID:18425110
23. Lloyd AJ, Favé G, Beckmann M et al. (2011). Use of mass spectrometry fingerprinting to identify urinary metabolites after con- sumption of specific foods. Am J Clin Nutr, 94:981–991. http://dx.doi.org/10.3945/ ajcn.111.017921 PMID:21865330
228
Summary • Stem cells are defined by their
capacity to self-renew, which is essential for the maintenance of a stem cell pool, and to differen- tiate according to different line- ages, as required for the integ- rity and functioning of particular tissues.
• Because of their long half-life, stem or progenitor cells may be particularly affected by genetic and epigenetic changes, and may thereby contribute to cancer development.
• “Cancer stem cell” is an op- erational term used to define a distinct subpopulation of tumour cells functionally, with reference to aberrant renewal potential and the capacity to confer tumour heterogeneity.
• Although the origin of cancer stem cells has not been estab- lished, cancer stem cells as cur- rently identified share many key properties with embryonic stem cells, including unlimited prolif- erative potential and the capacity to migrate.
• Research on stem cells and can- cer stem cells may indicate novel approaches to cancer therapy.
3.6 3 BIOLOGY
Stem cells and cancer stem cells Zdenko Herceg Hector Hernandez Vargas
Gianpaolo Papaccio (reviewer)
The pace of discovery in the field of stem cells has drawn academic, political, and public attention to stem cells as tools for innovation in medi- cine and biomedical research. Stem cells are found in all multicellular or- ganisms and are likely to be present as a discrete population in most tis- sues. Stem cells can be grown in culture and differentiated into spe- cialized cells that exhibit properties specific to various tissues. The land- mark discovery by Takahashi and Yamanaka that induced pluripotent stem cells, which have properties in common with embryonic stem cells, could be derived from differentiated cells – notably, skin fibroblasts [1] – led to the award of the Nobel Prize in Physiology or Medicine 2012 and has been reproduced using a variety of human cell types [2]. Furthermore, several studies have described the generation of induced pluripotent stem cell lines from individuals har- bouring both simple and complex ge- netic diseases [3]. Therefore, stem cells have been seen as an essential resource in cloning and regenerative medicine.
Stem cells share important char- acteristics with malignant cells. A bet- ter understanding of the behaviour and properties of stem cells could be exploited to devise a range of novel agents for and/or approaches to cancer therapy. Although technical
advances that enable the isolation and manipulation of embryonic stem cells, and hence the possibility of human cloning, have provoked in- tense ethical debate, these concerns should not detract from a recognition of the tremendous potential of stem cells for the treatment of various hu- man diseases, such as neurodegen- erative disorders and cancer [4,5].
Embryonic and tissue- specific stem cells All cells in the body are descended from a single cell: the fertilized egg or zygote. From the moment of fer- tilization of an egg until death, an organism passes through several developmental stages (Fig. 3.6.1). Multiplication of cells from the fer- tilized egg gives rise to different cell types, as evident in complex organisms. This process involves populations of stem cells that can be propagated indefinitely in culture under adequate conditions. These cells are designated embryonic stem cells and can give rise to any cell type and reconstitute the entire embryo. In addition, many adult tis- sues contain a discrete population of undifferentiated cells with prop- erties of stem cells. These cells are described as tissue-specific stem cells, also called somatic stem cells or adult stem cells. Haematopoietic stem cells are the best-characterized
Chapter 3.6 • Stem cells and cancer stem cells 229
3 B
IO LO
G Y
C H
A P
T E
R 3
.6
tissue-specific stem cells and can generate all blood lineages and ma- ture blood elements. Adult stem cells have been identified in many other tissues, such as the brain, skin, and liver. Although only a few types of tissue-specific stem cells have un- dergone rigorous identification and characterization, stem cells are likely present in any tissue that is able to undergo renewal.
Tissue-specific stem cells also have the capacity to perpetuate themselves through self-renewal and to produce the various mature cells of a particular tissue through differentiation [6]. Tissue-specific stem cells constitute a very minor cell population in adult tissues, but this population is essential for the maintenance of tissue homeostasis. Dysregulation of tissue-specific stem cells may initiate diseases, most notably cancer.
The two characteristics of stem cells that distinguish them from all other cells are self-renewal and multi- potency. Self-renewal is the capacity of a cell to divide and produce identi- cal daughter cells over a long time period. This crucial property allows
stem cells to persist for the lifetime of an organism. Multipotency is the capacity of stem cells to differentiate into many highly specialized cells,
such as neurons, muscle fibres, and blood elements. Multipotency is necessary for the maintenance of integrity and functionality of many
Fig. 3.6.1. Embryonic stem cells, derived from the inner cell mass of the blastocyst, are pluripotent and can give rise to all cell types of the body. Somatic stem cells, sometimes termed adult stem cells, are also capable of self-renewal and, with appropriate signals, differentiate into various cell types of the organ from which they are derived. The extent to which somatic stem cells are capable of differentiating into cell types from alternative lineages is controversial.
Fig. 3.6.2. Embryonic skin-derived progenitor cells. Cancer stem cells share many properties with embryonic stem cells.
230
tissues. Stem cells are essential for the development of each organism and thus can be considered to be a critical resource from the beginning of embryonic life until death. Given the potential of stem cells, their function and proliferation are sub- ject to a range of control processes. Dysregulation of the surveillance mechanisms for proliferation and dif- ferentiation of stem cells may initiate a shift in the balance between self- renewal and differentiation, leading to either stem cell loss, which is associ- ated with degenerative disorders, or abnormal proliferation of stem cells, which may be a source of malignant cells.
Cancer stem cells Stem cells were discovered more than 30 years ago and have been ex- ploited extensively for the generation
of genetically modified animal mod- els, such as knockout mice, which are essential in cancer research. However, the first identification of human stem cells and, in particular,
so-called cancer stem cells resulted in the cancer research community giving almost unprecedented atten- tion to these entities. Tumours have long been considered to be derived
Stem cells in cancer: determinants of clinical outcome? John E. Dick
The cellular and molecular basis for intratumour heterogeneity is poorly understood. Tumour cells can be genetically diverse due to mutations and clonal evolution, resulting in in- tratumour functional heterogeneity. Cancer stem cell models postulate that tumours are cellular hierarchies sustained by cancer stem cell het- erogeneity due to epigenetic differ- ences; that is, all cells present after long-term tumour growth are derived from cancer stem cells. There is strong evidence for the cancer stem cell model in acute myeloid leukae- mia (AML). Examination of the gene signatures specific to either AML stem cells or normal haematopoietic stem cells indicates that they share a set of genes that define a common stemness programme. Only these stem cell-related gene signatures were significant independent predic- tors of patient survival in large clini- cal databases. Thus, determinants of stemness influence clinical outcome of AML, establishing that leukaemia stem cells are clinically relevant and not artefacts of xenotransplantation.
In an effort to determine whether the clonal evolution and stem cell mod- els of cancer can be unified, two different studies were performed examining cancer stem cells at both the functional and the genetic level.
In Philadelphia chromosome- positive B-cell acute lymphoblastic leukaemia (B-ALL), diagnostic pa- tient samples had extensive sub- clonal genetic diversity, and through the use of xenotransplant assays, this diversity was established to have originated from within the leukaemia- initiating cells. Reconstruction of their genetic ancestry showed that multiple leukaemia-initiating cell sub- clones were related through a com- plex branching evolutionary process, indicating that genetic and functional heterogeneity are closely connected. This research highlights the need to develop effective therapies to eradi- cate all genetic subclones to prevent further evolution and recurrence.
In a related functional and genet- ic study, a highly reliable xenograft assay was used for primary colo- rectal cancer. With a combination of
DNA copy number alteration profil- ing, targeted and exome sequenc- ing, and lentiviral lineage tracing, the repopulation dynamics of many sin- gle lentivirus-marked lineages from colorectal cancers were followed through serial xenograft passages. The xenografts were genetically sta- ble on serial transplantation. Despite this genetic stability, the prolifera- tion, persistence, and chemother- apy tolerance of lentivirus-marked lineages were variable within each clone. Thus, apart from genetic di- versity, tumour cells display inher- ent functional variability in tumour propagation potential, a mechanism that maximizes tumour fitness, con- tributing to both tumour survival and therapy tolerance.
References Notta F et al. (2011). Nature, 469:362–367. ht t p: //d x .d o i.o r g /10.10 3 8 /n at u r e 0 973 3 PMID:21248843
Kreso A et al. (2013). Science, 339:543–548. http://dx.doi.org/10.1126/science.1227670 PMID:23239622
Fig. 3.6.3. Genetic and epigenetic changes in susceptible cells may be an early event in the development of cancer and give rise to cancer stem cells and contribute to tumour heterogeneity. The origin of cancer stem cells may be early stem or progenitor cells, or differentiated cells distinguished by abnormal expression and function of a set of genes that may contribute to reprogramming into a pluripotent state.
Chapter 3.6 • Stem cells and cancer stem cells 231
3 B
IO LO
G Y
C H
A P
T E
R 3
.6
from a single cell that has been sub- ject to transformation into a cancer- initiating cell through a series of ge- netic and epigenetic lesions. These initial events allow for the expansion of transformed cells and the forma- tion of a population of altered cells, or a clone, with the capacity to grow and divide in defiance of normal cel- lular control. Continuing selection of “fitter” and more aggressive cells re- sults in a generation of cancer clones capable of invading and destroying neighbouring tissues and migrating to distant organs to form metastatic tumours. The identity of that original “target cell” is not known (Fig. 3.6.3). However, recent studies indicate that many genetic and epigenetic chang- es underlying the aggressive and destructive behaviour of cancer are orchestrated by a discrete population of cancer cells with stem cell proper- ties. These cells are known as cancer stem cells.
The cancer stem cell hypothesis states that cancer clones are main- tained exclusively by a fraction of rare cells with stem cell properties [7]. A hierarchical organization has been established for several human cancers. In those cases, tumours are organized into undifferentiated cells that can drive disease progression
and differentiated cells with less ca- pacity to drive disease progression, consistent with the cancer stem cell model [8]. Many cancers have been found to contain cells with proper- ties of stem cells. However, in most cases the existence of cancer stem cells has been documented function- ally: the presence of cancer stem cells in the bulk of cancer cells is discerned by their capacity to form tumours after transplantation into an immunocompromised animal host, most commonly a mouse. These assays reveal that only a small frac- tion of cancer cells are capable of forming new tumours in the host. Importantly, these cells not only can form tumours upon transplantation but also can recapitulate tumour het- erogeneity [9]. However, until very recently, the isolation of a cancer stem cell population using molecular signatures, cell surface markers, or mutation profiles has proven to be extremely difficult [7,10,11].
Recently, numerous studies have shown that cancer stem cells can be enriched via different cell surface markers, such as CD133, CD44, CD90, and EpCAM (CD326) (Table 3.6.1) [12–30]. However, al- though some markers are common to different types of cancer stem cells,
so far no single marker (or combi- nation of markers) has been shown to be able to distinguish these cells across all cancer types. This lack of universal expression of surface markers limits their use. In addition, surface markers should be interpret- ed carefully, considering the tissue context and functional assessments (e.g. tumorigenic, clonogenic, and sphere formation assays) and co- expression with stemness markers [11]. In spite of these weaknesses, many studies have taken advantage of surface markers to shed light on the biology of putative cancer stem cells. Some markers are common to several malignancies (e.g. CD44, CD133) and, interestingly, some markers are common to normal stem or progenitor cells (e.g. CD133, EpCAM). Indeed, the cell surface ex- pression of the human prominin-1 (CD133) antigen is common to hae- matopoietic progenitors, and it is among the most frequently studied markers of cancer stem cells [31].
In addition to surface markers, cancer stem cells may share many key properties with embryonic stem cells. These properties include in- finite proliferation potential and the capacity to invade tissues and or- gans and to promote formation of blood vessels for their own oxygen- ation. Therefore, cancer stem cells may contribute to the heterogeneity of most human cancers [8]. Although important progress has been made in the identification of cancer stem cells, their origin remains obscure. Cancer stem cells are believed to arise in different ways. First, cancer stem cells can be derived from nor- mal tissue-specific stem cells as a result of specific genetic and epige- netic changes that abrogate prolif- eration control in the normal cells. Second, differentiated cells that normally have a limited lifespan can regain stem cell properties, that is de-differentiate and become can- cer stem cells. Third, normal stem cells may fuse with various differ- entiated cells and the resulting hy- brid cells may be cancer-initiating cells with stem cell properties [32]. These hypotheses are not mutually
Table 3.6.1. Surface marker phenotypes commonly used to identify cancer stem cells
Tumour type [source of data] Cancer stem cell marker
Acute myeloid leukaemia [12] CD34+/CD38 –
Breast cancer [13,14] CD44+/CD24lo, CD133+
Gall bladder cancer [15] CD44+/CD133+
Gastric cancer [16] CD44+
Glioblastoma [17,18] CD133+
Head and neck cancer [19] CD44+
Hepatocellular carcinoma [20–23] CD133+, CD13+, CD90+, EpCAM+
Lung cancer [24] CD133+
Melanoma [25] CD20+
Oesophageal cancer [26,27] CD90+, CD44+
Ovarian cancer [28] CD133+
Pancreatic cancer [29] CD133+
Prostate cancer [30] CD133+/CD44+/α2β1hi
232
exclusive, and the genesis of cancer stem cells may involve more than one mechanism.
Genomic determination of stem cell identity Although the features that distin- guish stem cells from differentiated cells have been known for many decades, the genetic basis of stem cell identity only began to be under- stood recently. The development of powerful methodologies in genomics for genome-wide screening enabled the identification of genes and gene networks that maintain stem cells in a special state. With these methods, a limited number of genes have been discovered that are necessary and sufficient to maintain self-renewal and pluripotency – the two distin- guishing features of stem cells. These genes have been referred to as “masters of stemness”. The genes Oct4, Sox2, and Nanog belong to this exclusive category [33]. These genes encode transcription factors that control the transcription of other genes (Fig. 3.6.4). Such signal trans- duction pathways maintain stem cell behaviour and identity. When the rel- evant genes are inactivated or mu- tated, stem cells may differentiate into specialized cells and the stem
cell pool may be rapidly depleted. Depletion of the stem cell pool can impede regeneration and compro- mise the integrity of normal tissue, leading to degenerative disease [34].
The discovery of stem cell mas- ter genes has enabled another, even more insightful branch of research: the transition of differentiated cells into immature pluripotent (stem) cells. This phenotypic reversal or de- differentiation essentially involves enabling specialized cells such as neurons or muscle fibres to regain stem cell properties, which would allow the generation of practically any type of cell. Such a scenario may solve important ethical issues associated with the use of embryos as a source of stem cells, as demon- strated by recent studies [3]. Several laboratories have demonstrated that the introduction of as few as four master genes into either human or murine differentiated cells mediates the emergence of stem cells. This remarkable phenomenon provides support for the argument that differ- entiated cells can be reprogrammed and that the course of events that normally characterizes differentia- tion can be reversed. An important implication of these findings is that
Fig. 3.6.4. Cell reprogramming factors can induce pluripotent stem cells, with potential applications in therapy. iPSCs, induced pluripotent stem cells.
Fig. 3.6.5. Reprogramming can alter any cell of the body so that it may function as a pluripotent stem cell. Shown are human induced pluripotent stem cells derived from dermal fibroblasts. Immunofluorescence analysis with an antibody against Nanog protein (red) is used to indicate that Nanog (a master transcription factor critically involved in stem cells and a widely used marker of the pluripotent state) is expressed and that cells have been successfully reprogrammed.
Chapter 3.6 • Stem cells and cancer stem cells 233
3 B
IO LO
G Y
C H
A P
T E
R 3
.6
cancer stem cells may also arise by non-genetic changes that confer certain features on differentiated cells. This hypothesis is supported by the knowledge that all cells, in- cluding stem cells, in any given or- ganism share an identical genome.
Experimental evidence from vari- ous recent studies suggests a simi- lar conclusion to that just described: cell differentiation is not unidirec- tional, and different factors may lead to reversal of cell differentiation un- der either physiological or pathologi- cal conditions. In the same way that tumorigenesis may be identified with the acquisition of self-renewal and pluripotency in vivo, manipula- tion of differentiated cells in vitro may lead to a controlled version
of the same process. Conversely, cells with established stemness features may acquire characteris- tics of partial differentiation. The balance between stemness and a differentiated character appears to be different during critical periods of life, such as embryonic develop- ment. An extrapolation of the same idea is evidenced by the increasing interest being accorded to the ef- fects of early-life exposures on the development of later disease. In this sense, early in utero life represents a critical period where pluripotency and differentiation networks interact dynamically and where environmen- tal exposures may have an impact beyond that initially perceived.
Stem cells and cancer therapy Much current interest in stem cells and cancer stem cells is predicated on the realization that this tiny, yet critical, population of cancer cells represents an opportunity to devise novel strategies for cancer therapy [35]. Many currently used protocols for cancer therapy are now under- stood to fail – marked by the reap- pearance of disease – due to an inability to eradicate cancer stem cells. Classic chemotherapy regimes have been developed to rapidly shrink tumours. However, these changes are often transient and are followed by recurrence (Fig. 3.6.6). Such a clinical course suggests that the ther- apy used specifically targets rapidly growing tumour cells, whereas more slowly growing cancer stem cells may be spared. Thus, major challenges will be (i) to discover efficient ways to identify and isolate tissue-specific stem cells and cancer stem cells, (ii) to gain insights into the mechanisms of self-renewal and pluripotency of normal stem cells and cancer stem cells, and (iii) to identify genetic and epigenetic events that are critical to cancer-initiating reprogramming and cancer stem cell development (see “Stem cells in cancer: determinants of clinical outcome?”).
The ever-increasing research ef- forts in the field of stem cells have the potential to identify processes that are currently unknown and yet key in cancer development, which will ultimately have a major impact on the development of novel strategies to control and perhaps eliminate cancer.
Fig. 3.6.6. Conventional therapies may reduce tumour size by killing mainly differenti- ated tumour cells. If the putative cancer stem cells are less sensitive to these therapies, then cancer stem cells will remain viable after therapy and re-establish the tumour. By contrast, if therapies can be targeted against cancer stem cells, then they might more effectively kill the cancer stem cells, rendering the tumours unable to maintain them- selves or grow. Thus, even if cancer stem cell-directed therapies do not shrink tumours initially, they may eventually lead to cures. CSC, cancer stem cell.
234
References
1. Takahashi K, Yamanaka S (2006). Induction of pluripotent stem cells from mouse em- bryonic and adult fibroblast cultures by defined factors. Cell, 126:663–676. http:// dx.doi.org/10.1016/j.cell. 20 0 6.07.024 PMID:16904174
2. Yamanaka S (2007). Strategies and new developments in the generation of patient- specific pluripotent stem cells. Cell Stem Cell, 1:39– 49. http://dx.doi.org/10.1016/j. stem.2007.05.012 PMID:18371333
3. Bellin M, Marchetto MC, Gage FH, Mummery CL (2012). Induced pluripo- tent stem cells: the new patient? Nat Rev Mol Cell Biol, 13:713–726. http://dx.doi. org/10.1038/nrm3448 PMID:23034453
4. Polyak K, Hahn WC (2006). Roots and stems: stem cells in cancer. Nat Med, 12:296–300. http://dx.doi.org/10.1038/nm 1379 PMID:16520777
5. Baumann M, Krause M, Hill R (2008). Exploring the role of cancer stem cells in radioresistance. Nat Rev Cancer, 8:545 –554. http://dx.doi.org/10.1038/nrc2419 PMID:18511937
6. Rossi DJ, Jamieson CH, Weissman IL (2008). Stems cells and the pathways to aging and cancer. Cell, 132:681–696. http://dx.doi.org/10.1016/j.cell.2008.01.036 PMID:18295583
7. Valent P, Bonnet D, De Maria R et al. (2012). Cancer stem cell definitions and terminology: the devil is in the details. Nat Rev Cancer, 12:767–775. http://dx.doi. org/10.1038/nrc3368 PMID:23051844
8. Magee JA, Piskounova E, Morrison SJ (2012). Cancer stem cells: impact, het- erogeneity, and uncertainty. Cancer Cell, 21:283–296. http://dx.doi.org/10.1016/j.ccr. 2012.03.003 PMID:22439924
9. Visvader JE, Lindeman GJ (2008). Cancer stem cells in solid tumours: accumulating evidence and unresolved questions. Nat Rev Cancer, 8:755–768. http://dx.doi. org/10.1038/nrc2499 PMID:18784658
10. Tirino V, Desiderio V, Paino F et al. (2011). Human primary bone sarcomas contain CD133+ cancer stem cells displaying high tumorigenicity in vivo. FASEB J, 25: 2022–2030. http://dx.doi.org/10.1096/fj.10- 179036 PMID:21385990
11. Tirino V, Desiderio V, Paino F et al. (2013). Cancer stem cells in solid tumors: an over- view and new approaches for their isolation and characterization. FASEB J, 27:13–24. http://dx.doi.org/10.109 6/f j.12-218222 PMID:23024375
12. Lapidot T, Sirard C, Vormoor J et al. (1994). A cell initiating human acute myeloid leu- kaemia after transplantation into SCID mice. Nature, 367:645–648. http://dx.doi. org/10.1038/367645a0 PMID:7509044
13. Al-Hajj M, Wicha MS, Benito-Hernandez A et al. (2003). Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A, 100:3983–3988. http://dx.doi.org/10.1073/pnas.0530291100 PMID:12629218
14. Wright MH, Calcagno AM, Salcido CD et al. (2008). Brca1 breast tumors contain distinct CD44+/CD24- and CD133+ cells with cancer stem cell characteristics. Breast Cancer Res, 10:R10. http://dx.doi. org/10.1186/bcr1855 PMID:18241344
15. Shi C, Tian R, Wang M et al. (2010). CD44+CD133+ population exhibits cancer stem cell-like characteristics in human gallbladder carcinoma. Cancer Biol Ther, 10:1182–1190. http://dx.doi.org/10.4161/ cbt.10.11.13664 PMID:20948317
16. Takaishi S, Okumura T, Tu S et al. (2009). Identification of gastric cancer stem cells using the cell surface marker CD44. Stem Cells, 27:1006–1020. http://dx.doi. org/10.1002/stem.30 PMID:19415765
17. Singh SK, Clarke ID, Terasaki M et al. (2003). Identification of a cancer stem cell in human brain tumors. Cancer Res, 63:5821–5828. PMID:14522905
18. Singh SK, Hawkins C, Clarke ID et al. (2004). Identification of human brain tu- mour initiating cells. Nature, 432:396– 401. http://dx.doi.org/10.1038/nature03128 PMID:15549107
19. Prince ME, Sivanandan R, Kaczorowski A et al. (2007). Identification of a sub- population of cells with cancer stem cell properties in head and neck squamous cell carcinoma. Proc Natl Acad Sci U S A, 104:973–978. http://dx.doi.org/10.1073/ pnas.0610117104 PMID:17210912
20. Yin S, Li J, Hu C et al. (2007). CD133 positive hepatocellular carcinoma cells possess high capacity for tumorigenicity. Int J Cancer, 120:1444–1450. http://dx.doi. org/10.1002/ijc.22476 PMID:17205516
21. Haraguchi N, Ishii H, Mimori K et al. (2010). CD13 is a therapeutic target in human liver cancer stem cells. J Clin Invest, 120:3326– 3339. http://dx.doi.org/10.1172/JCI42550 PMID:20697159
22. Yang ZF, Ho DW, Ng MN et al. (2008). Significance of CD90+ cancer stem cells in human liver cancer. Cancer Cell, 13:153–166. http://dx.doi.org/10.1016/j.ccr. 2008.01.013 PMID:18242515
23. Yamashita T, Ji J, Budhu A et al. (2009). EpCAM-positive hepatocellular carcinoma cells are tumor-initiating cells with stem/ progenitor cell features. Gastroenterology, 136:1012–1024, e4. http://dx.doi.org/10. 1053/j.gastro.2008.12.004 PMID:19150350
24. Eramo A, Lotti F, Sette G et al. (2008). Identification and expansion of the tumorigenic lung cancer stem cell popu- lation. Cell Death Differ, 15:504–514. http://dx.doi.org/10.1038/sj.cdd.4402283 PMID:18049477
25. Fang D, Nguyen TK, Leishear K et al. (2005). A tumorigenic subpopulation with stem cell properties in melanomas. Cancer Res, 65:9328–9337. http://dx.doi. org/10.1158/0008 - 5 472.CAN - 05 -1343 PMID:16230395
26. Tang KH, Dai YD, Tong M et al. (2013). A CD90+ tumor-initiating cell population with an aggressive signature and meta- static capacity in esophageal cancer. Cancer Res, 73:2322–2332. http://dx.doi. org/10.1158/0 0 08 - 5 472.CAN -12-2991 PMID:23382045
27. Zhao JS, Li WJ, Ge D et al. (2011). Tumor initiating cells in esophageal squamous cell carcinomas express high levels of CD44. PLoS One, 6:e21419. http://dx.doi. o r g / 1 0 .13 71 / j o u r n a l . p o n e . 0 0 2 141 9 PMID:21731740
28. Baba T, Convery PA, Matsumura N et al. (2009). Epigenetic regulation of CD133 and tumorigenicity of CD133+ ovarian cancer cells. Oncogene, 28:209–218. http://dx.doi.org/10.1038/onc.2008.374 PMID:18836486
29. Hermann PC, Huber SL, Herrler T et al. (2007). Distinct populations of cancer stem cells determine tumor growth and metastatic activity in human pancreatic cancer. Cell Stem Cell, 1:313–323. http:// dx.doi.org/10.1016/j.stem.2007.06.002 PMID:18371365
30. Collins AT, Berry PA, Hyde C et al. (2005). Prospective identification of tumorigenic prostate cancer stem cells. Cancer Res, 65:10946–10951. http://dx.doi.org/10.1158/0008-5472.CAN- 05-2018 PMID:16322242
31. Yin AH, Miraglia S, Zanjani ED et al. (1997). AC133, a novel marker for human hematopoietic stem and progenitor cells. Blood, 90:5002–5012. PMID:9389720
32. Bjerkvig R, Tysnes BB, Aboody KS et al. (2005). The origin of the cancer stem cell: current controversies and new insights. Nat Rev Cancer, 5:899–904. http://dx.doi. org/10.1038/nrc1740 PMID:16327766
Chapter 3.6 • Stem cells and cancer stem cells 235
3 B
IO LO
G Y
C H
A P
T E
R 3
.6
33. Boiani M, Schöler HR (2005). Regulatory networks in embryo-derived pluripotent stem cells. Nat Rev Mol Cell Biol, 6:87 2–884. http://dx.doi.org/10.1038/nrm1744 PMID:16227977
34. Erlandsson A, Morshead CM (2006). Exploiting the properties of adult stem cells for the treatment of disease. Curr Opin Mol Ther, 8:331–337. PMID:16955696
35. Visvader JE, Lindeman GJ (2012). Cancer stem cells: current status and evolving complexities. Cell Stem Cell, 10: 717–728. http://dx.doi.org/10.1016/j.stem. 2012.05.007 PMID:22704512
236
Summary • The tumour microenvironment
refers to the biochemical and cellular composition of the tis- sue encompassing a tumour and comprises a heterogeneous population of cancer cells, a va- riety of non-cancer cells, soluble proteins, blood vessels, peritu- moural lymphatic vessels, and a supporting structural matrix.
• Bidirectional signalling between cancer cells and stromal cells leads to enhanced cancer cell division, suppressed immune cell function, and resistance to therapeutic intervention, all of which result in poor clinical outcome.
• Cancer-associated inflammation parallels disease progression, and tumours have been charac- terized as “wounds that do not heal”.
• Accumulating evidence sug- gests that combined therapeutic regimens that both target cancer cells (e.g. chemotherapy) and modify the tumour microenvi- ronment (i.e. anti-angiogenic therapy) are superior to treat- ments directed against only cancer cells.
3.7 3 BIOLOGY
Tumour microenvironment Robert R. Langley Chris Paraskeva (reviewer)
Theresa L. Whiteside (reviewer) Jiri Zavadil (reviewer)
Over the past several years, it has become increasingly apparent that the reciprocal signalling that takes place between cancer cells and the cellular and molecular components of the surrounding tissue plays a decisive role in determining tumour growth, metastasis, and responsive- ness to therapy. Tumour growth is accompanied by an ever-increasing recruitment and activation of previ- ously quiescent cells, whose cumu- lative response is one that perpetu- ates disease progression (Fig. 3.7.1). Treatment efforts are hindered by a unique set of physiological param- eters present in the tumour micro- environment. Consequently, much effort is currently directed towards the development of therapeutic strategies that both target cancer cells and modify the tumour micro- environment. Several key aspects of the tumour microenvironment are described in this chapter.
Tumour vasculature The proximity of a cancer cell to a blood vessel determines its fate; cancer cells located within 75 μm of a blood vessel enter the cell di- vision cycle, whereas cancer cells located farther than 150 μm from a vessel undergo programmed cell death [1]. Tumour growth to a size beyond a few millimetres in diam- eter leads to hypoxia and nutrient
deprivation, which activate the “an- giogenic switch” and allow tumours to progress [2]. Cancer cells pro- duce cytokines and growth factors that act in an autocrine manner to promote their own expansion, and in a paracrine fashion to convey infor- mation to quiescent adjacent cells. The initial target cells for signal-re- leasing cancer cells are resident mi- crovascular endothelial cells, which respond by activating programmes that culminate in the formation of new vascular networks (i.e. angio- genesis) (Fig. 3.7.2). Angiogenesis diminishes the metabolic pressures associated with unrestricted cancer cell division and increases the likeli- hood of cancer cell dissemination. The tumour microenvironment be- comes enriched in vascular en- dothelial growth factor (VEGF), a primary mediator of pathological angiogenesis.
Angiogenesis is most pronounced at the interface between tumour and normal tissue, although the intensity of the neovascularization response varies between different types of tu- mours (Fig. 3.7.3). Angiogenesis is greatest in glioblastoma and renal cell carcinoma and least in tumours of the prostate and lung [3]. Some cancer cells (e.g. melanoma cells) may forgo angiogenesis altogether and satisfy their metabolic demands by prolifer- ating along the length of pre-existing
Chapter 3.7 • Tumour microenvironment 237
3 B
IO LO
G Y
C H
A P
T E
R 3
.7
blood vessels in a process referred to as vessel co-option. Cancer cells may co-opt blood vessels when chal- lenged with anti-angiogenic agents.
Pharmacological suppression of angiogenesis as a means to control tumour growth is an area of intense investigation. Several inhibitors of the VEGF signalling pathway are currently in clinical use; these in- clude bevacizumab, a monoclonal antibody that neutralizes VEGF, and small-molecule inhibitors of the ty- rosine kinase receptor VEGFR2 (e.g. sorafenib, sunitinib). The ad- dition of bevacizumab to standard therapy was found to increase
overall survival in patients with advanced malignancies, including non-small cell lung cancer and met- astatic colorectal cancer. However, many tumours are inherently resis- tant to anti-angiogenic agents, and those tumours that initially respond to treatment eventually become re- fractory to therapy and recur.
Independent investigations con- cluded that cancer cells circumvent anti-VEGF therapies by upregulat- ing alternative angiogenic proteins, such as basic fibroblast growth fac- tor and interleukin-8. Recent data suggest that stromal cells also contribute to resistance to anti-
VEGF agents. Indeed, preclinical studies using cross-species hybrid- ization of microarrays concluded that most of the gene expression changes associated with acquired resistance of non-small cell lung cancers to bevacizumab occur in stromal cells [4]. The epidermal growth factor receptor (EGFR) gene was identified as a network hub activated on stromal cells of bevacizumab-resistant non-small cell lung tumours. Co-localization studies revealed a 10-fold increase in the number of pericytes express- ing the activated form of EGFR in experimental bevacizumab-
Fig. 3.7.1. Illustration depicting the complexity of the interplay between cancer cells and stromal cells. (A) Transforming growth factor β (TGF-β) produced from initiated epithelial cells activates tissue fibroblasts. Activated fibroblasts (myofibroblasts) generate hepatocyte growth factor (HGF), which binds to the MET receptor on epithelial cells. Activation of MET signalling has been shown to stimulate pathways that promote cell division, motility, and anti-apoptotic activity. (B) Cancer cells produce the pro-angiogenic protein vascular endothelial growth factor (VEGF), in response to declining oxygen tension. (C) VEGF is chemotactic for macrophages, which contribute to the pool of pro-angiogenic proteins. (D) Myofibroblast-derived HGF activates invasive and migratory programmes in cancer cells, allowing them to penetrate structurally deficient angiogenic vessels and spread to distal tissues. (E) Interleukin-1 (IL-1) produced by myofibroblasts leads to the recruitment of inflammatory cells. (F) Tumour-associated macrophages generate VEGF-C, which signals for lymphangiogenesis, increasing the likelihood of lymph node metastasis. (G) Macrophages and myofibroblasts secrete matrix metalloproteinases (MMPs), which remodel the extracellular matrix. (H) Cancer cell chemokines including chemokine ligand 2 (CCL2) and CCL5 direct circulating monocytes to the tumour. Macrophage colony-stimulating factor (MCSF) promotes the differentiation of monocytes to macrophages. (I) Cancer cell-derived IL-6 and IL-10 polarize macrophages to an M2 phenotype with pro-angiogenic and immunosuppressive properties.
238
resistant tumours compared with control tumours. Dual inhibition of VEGFR and EGFR pathways re- duced pericyte coverage of tumour blood vessels and, moreover, de- layed emergence of the resistant phenotype [4].
In other animal models of can- cer, tumour refractoriness to anti- VEGF therapy was dependent on the tumour’s ability to prime and re- cruit CD11b+Gr1+ granulocyte-mac- rophage cells [5]. Cancer cells and tumour-associated fibroblasts syn- thesize and secrete granulocyte- macrophage colony-stimulating factor and stromal cell-derived factor 1, which stimulate myeloid recruitment into the tumour micro- environment. In addition to provid- ing a rich reserve of angiogenic proteins, CD11b+Gr1+ cells also suppress the immune response of CD4+ and CD8+ T cells through several different mechanisms, in- cluding upregulation of reactive oxygen species, nitric oxide, and immunosuppressive cytokines.
Physiological parameters of the tumour microenvironment Tumour blood vessels are distinct from normal vessels in that they are dilated, tortuous, and heteroge- neously distributed, and comprise constitutively activated endothe- lial cells. VEGF renders the tumour vascular bed hyperpermeable to macromolecules, and extravasated proteinaceous fluid accumulates in the tumour matrix due to a lack of functional intratumour lymphatic vessels. Consequently, tumour in- terstitial colloid osmotic pressures are significantly greater than those of normal tissues and most tumours have grossly elevated interstitial fluid pressures. The elevated interstitial fluid pressure creates a barrier to transcapillary transport, which lim- its the uptake of therapeutic agents in tumours. Interventions that lower interstitial fluid pressure, such as anti-VEGF therapies, are reported to normalize the tumour vasculature and improve drug uptake.
Chemotherapy and radiotherapy require optimal tumour blood flow to
achieve maximal benefit, whereas anti-angiogenic therapies and vas- cular disrupting agents seek to completely abolish neoplastic blood flow. Blood flow to tumours is het- erogeneous, and marked variability exists even in tumours with a simi- lar histological classification growing in the same anatomical region [6]. Elevated interstitial fluid pressure may compress tumour blood ves- sels and diminish tumour blood flow. Blood flow fluctuations (and vascular stasis) occur frequently in the tumour circulation and are associated with periods of hypoxia. Fluctuations in erythrocyte flux in tumour blood vessels predispose tumour cells to periods of transient hypoxia and reoxygenation.
Estimates indicate that 50–60% of advanced tumours contain het- erogeneously distributed hypoxic regions. There are two forms of hyp- oxia in human tumours. Diffusion- limited hypoxia refers to limitations in the diffusion distance of oxygen from a blood vessel. The diffusion dis- tance of oxygen from tumour blood
Fig. 3.7.2. Morphological classification of tumour vasculature in hepatocellular carcinoma. Three human hepatocellular carcinomas were stained with anti-CD34 antibody to show their vasculature. The vasculature can be (A) capillary-like microvessels, (B) a mixture of capillary-like and sinusoid-like vasculature, or (C) sinusoid-like vasculature. The prognosis of patients harbouring these three vasculature types differs; capillary-like microvessels correspond to the best prognosis and sinusoid-like tumour vasculature the worst.
A B C
Chapter 3.7 • Tumour microenvironment 239
3 B
IO LO
G Y
C H
A P
T E
R 3
.7
vessels has been measured and is approximately 120 μm. Perfusion- limited hypoxia results from non-uni- form blood flow patterns caused by structural and functional alterations in tumour vessels. For example, a rapidly expanding tumour may im- pinge on its vasculature and tem- porarily or permanently block blood flow, thereby creating areas deficient in oxygen. Arteriovenous shunting is also a common feature in tumour circulation; approximately one third of arterial blood can pass through tumours without taking part in the microcirculatory exchange process.
The oxygenation status of tu- mours has important clinical im- plications. Radiotherapy and che- motherapy are significantly more effective in an aerobic environment. Aerobic cells are 3 times as sensi- tive as hypoxic cells to the effects of radiation. Hypoxia acts as a physiological selective pressure for variants with diminished apoptotic potential, enhanced genetic instabil- ity, and increased invasive capacity.
Oxygen deprivation is associated with increased metastasis and poor outcomes in patients with head and neck carcinomas and cervical cancer.
Studies evaluating the genetic stability of the stromal compart- ment have yielded conflicting results. Examinations of formalin-fixed, par- affin-embedded archival tumours reported loss of heterozygosity (LOH) and changes in copy number of certain loci in cancer-associated fibroblasts, whereas studies per- formed on fresh-frozen tissues failed to detect chromosomal changes in cancer-associated fibroblasts. A re- cent genome-wide copy number and LOH analysis of fresh-frozen breast and ovarian cancers using array analysis for hundreds of thousands of single-nucleotide polymorphisms reported that LOH and copy num- ber alterations are extremely rare in cancer-associated fibroblasts [7].
There is no debate about the im- portance of inflammation to carcino- genesis and neoplastic progression
(see “Inflammation and cancer” in Chapter 2.4). Epidemiological re- ports indicate that chronic inflamma- tion in select organs increases the risk for development of cancer and that non-steroidal anti-inflammatory drugs reduce the incidence of, and mortality from, several tumour types [8]. An inflammatory component is also present in the microenvironment of tumours that are not epidemio- logically related to inflammation. In general, cancer-associated inflam- mation parallels disease progres- sion and contributes to cancer cell division and survival, angiogenesis, metastasis, subversion of adaptive immunity, and reduced response to therapy [8].
Tumour-associated macrophages Macrophages are among the most represented innate immune cells in the tumour microenvironment. A correlation exists between high mac- rophage content and poor prognosis in many human tumours. Circulating monocytes traffic to tumours in re- sponse to chemokines including chemokine ligand 2 (CCL2) and CCL5, growth factors such as VEGF, and vasoactive peptides including endothelins. Cancer cell-derived fac- tors mediate the differentiation and orientation of incoming monocytes. Macrophage colony-stimulating fac- tor promotes the differentiation of monocytes to resident macrophag- es, whereas interleukin-6 (IL-6) and IL-10 polarize macrophages towards an M2 phenotype [9]. Results from clinical and experimental studies suggest that the functional proper- ties of M2 macrophages favour tu- mour progression. M2 macrophages possess limited antigen-presenting capacity and suppress Th1 adaptive immunity while actively promoting angiogenesis and tissue remodel- ling processes.
Tumour-associated macrophag- es are also a major source of matrix metalloproteinases (MMPs), a family of zinc-containing endopeptidases, which play a critical role in tissue remodelling processes. MMPs are
Fig. 3.7.3. Novel assays for measuring surrogate markers of tumour angiogenesis activity. The complex biology of angiogenesis inhibitors has accentuated the need to develop technologies that can be used to assess the effects of biological markers. A compilation of data from multiple assays, including measuring angiogenic factors in serum, plasma, and urine; tumour biopsy analysis; radiological imaging; and, recently, ex vivo analyses of isolated peripheral blood cells (labelled circulating endothelial cells) may facilitate defining the optimal biological dose for subsequent clinical studies of angiogenesis inhibitors.
240
initially expressed as inactive zymo- gens, and activation requires cleav- age of an autoinhibitory pro-domain. MMP-mediated proteolytic destruc- tion of the extracellular matrix results in the release of matrix-embedded bioactive fragments and growth fac- tors that encourage cancer cell divi- sion, angiogenesis, and neutrophil recruitment processes. Expression and activity levels of MMPs are up- regulated in most human cancers and correlate with advanced tumour stage, increased invasion and me- tastasis, and shortened survival [10].
Macrophages express an exten- sive array of pro-angiogenic proteins sufficient to trigger the angiogenic switch or amplify an ongoing neovas- cularization response. Conditions that enhance angiogenic activity in macrophages include hypoxia and
elevated levels of pyruvate, lactate, and hydrogen ions, all of which are characteristic of the tumour micro- environment. Macrophage-produced VEGF and prostaglandins diminish endothelial cell barrier function by in- creasing the permeability of tumour vasculature. Macrophage-derived VEGF-C and VEGF-D ligands signal for peritumoural lymphangiogenesis, which may increase the spread of cancer cells to regional lymph nodes.
Microenvironment in metastasis Most deaths due to cancer result from the progressive growth of me- tastases that are resistant to con- ventional therapies. The process of metastasis involves a complex series of interrelated steps that begins when cancer cells detach
from the primary mass and invade the surrounding stroma (Fig. 3.7.4). Carcinoma cells at the invasive front adopt morphological features and patterns of gene expression that are characteristic of mesenchymal cells during a process referred to as epithelial–mesenchymal tran- sition (EMT). EMT plays a critical role during embryonic development and contributes to tissue repair pro- cesses in the adult. Disseminating carcinoma cells reactivate this de- velopmental programme and re- press epithelial proteins E-cadherin, β-catenin, and cytokeratin and up- regulate mesenchymal-associated proteins N-cadherin, vimentin, and fibronectin. Recent studies suggest that cancer cells undergoing EMT may also acquire stem cell-like properties and become resistant to
Fig. 3.7.4. The process of metastasis consists of sequential, interlinked, and selective steps. Each step of the metastatic process is considered rate-limiting in that failure of a tumour cell to complete any step effectively terminates the process. The formation of clinically relevant metastases represents the survival and growth of a unique subpopulation of cells that pre-exist in primary tumours. See the text for a detailed description of the individual steps of the metastatic process.
Chapter 3.7 • Tumour microenvironment 241
3 B
IO LO
G Y
C H
A P
T E
R 3
.7
anticancer therapies [11]. Reciprocal signalling between cancer cells and cancer-associated fibroblasts is re- ported to stimulate EMT in some ex- perimental models. Invading cancer cells gain access to the systemic cir- culation by penetrating thin-walled blood vessels and lymphatic chan- nels. Circulating cancer cells are re- tained in target organs by adhering to microvascular endothelial cells or the sub-endothelial basement membrane. Extravasated cancer cells are thought to undergo mes- enchymal–epithelial transition in the secondary tissue, by a yet-unknown mechanism. Cancer cells that leave the cell cycle or are unable to trig- ger the angiogenic switch may en- ter a period of dormancy or undergo apoptosis.
The process of metastasis is ex- tremely inefficient in that less than 0.1% of circulating tumour cells eventually form metastases. That certain tumours have a tendency to form metastases in specific organs provided the foundation for the “seed and soil” hypothesis, which was pro- posed well over a century ago. In essence, Paget’s hypothesis states that tumour cells (“seed”) grow pref- erentially in the microenvironment of select organs (“soil”). In the con- cluding sections of this chapter, the discussion focuses on the microen- vironment of the brain and bone, two frequent target organs of metastasis.
Brain microenvironment Estimates indicate that approximate- ly 200 000 cases of brain metasta- ses occur each year in the USA, and that 20–40% of patients with sys- temic cancers develop brain metas- tases during the course of their dis- ease. Brain metastasis has several unmet clinical needs; there are few clinically relevant tumour models, no targeted therapies exist that are specific for brain metastases, and the median survival for untreated pa- tients is 5 weeks [12]. Most brain me- tastases arise from primary tumours that originate in the lung (40–50%), breast (15–20%), or skin (5–10%). Patients tend to present with multiple
brain metastases, 80% of which are located in the cerebral hemispheres.
Comparative genome expres- sion analysis was used to study the molecular mechanisms that medi- ate breast cancer cell arrest and extravasation in the brain. Cyclo- oxygenase 2 (COX-2), heparin- binding EGF-like growth factor (HB- EGF), and α-2,6-sialyltransferase 5 (ST6GALNAC5) were identified as key regulatory molecules that control the passage of breast cancer cells through the blood–brain barrier [13]. ST6GALNAC5 enhances the affin- ity of breast cancer cells for brain endothelial cells, whereas COX-2 increases the permeability of brain vasculature. Autocrine activation of breast cancer cell EGFR by HB- EGF signals for invasion of the brain parenchyma.
Results from real-time imag- ing studies suggest that survival of cancer cells in the central nervous system is dependent on their abil- ity to communicate with vascular endothelial cells. Failure to stimu- late angiogenesis leads to regres- sion of lung adenocarcinoma cell brain metastases, whereas the in- ability to locate cerebral vessels for co-option leads to activation of cell death programmes in melanoma brain metastases [14]. Brain en- dothelial cells also protect brain me- tastases from the cytotoxic effects of chemotherapy. Unlike endothelial cells from other regional circulations, brain endothelial cells constitutively express the P-glycoprotein trans- porter on their apical membrane, which mediates the efflux of several chemotherapeutic agents, including vinblastine, methotrexate, vincris- tine, doxorubicin, and etoposide. P-glycoprotein also limits the cen- tral nervous system uptake of sev- eral small-molecule targeted thera- pies, such as erlotinib, gefitinib, and imatinib.
One hallmark of human brain metastases is reactive astroglio- sis, the process in which astro- cytes alter their patterns of gene expression and undergo changes in cell morphology. Astrogliosis is characterized by upregulation of
the intermediate filament protein called glial fibrillary acidic protein, and hypertrophy of astrocyte cellular processes. Astrocytes encircle and infiltrate brain metastases, and the magnitude of astrogliosis parallels tumour growth. The precise mecha- nism responsible for astrogliosis remains unclear. Results obtained from co-culture systems using as- trocytes and human cancer cells demonstrated that astrocytes protect cancer cells from chemotherapeutic agents through gap junction-medi- ated communication that results in upregulation of GSTA5, BCL2L1, and TWIST1 survival genes in cancer cells [15].
Bone microenvironment Bone metastasis is also a signifi- cant public health concern; it oc- curs in up to 70% of individuals with advanced cancer of the breast or prostate and in approximately 40% of patients with carcinomas of the lung or kidney. The pathophysiology of bone metastasis involves several different cell populations and a va- riety of regulatory proteins. Bone metastases are classified as either osteoblastic or osteolytic, depending on whether the pathology involves bone formation or bone destruc- tion. Pathological bone remodelling results in significant skeletal compli- cations, including pain, hypercalcae- mia, fractures, spinal cord compres- sion, and immobility.
The highly vascularized me- taphyseal bone found at the ends of long bones, ribs, and vertebrae is a target site for bone-homing tu- mour cells. The bone is enriched in several cytokines and growth fac- tors that act as homing signals for cancer cells. Studies suggest that stromal cell-derived factor 1 and its receptor, CXCR4, may direct the traf- ficking of a variety of cancer cells to bone. Stromal cell-derived factor 1 is constitutively expressed in the bone microenvironment by fibroblasts, endothelial cells, and osteoblasts. Neutralizing antibodies and synthetic peptide antagonists directed against CXCR4 expressed on prostate and
242
breast cancer cells were found to re- duce the number of bone metasta- ses in experimental models. CXCR4 was identified as one of a small set of genes that mediate breast cancer metastasis to bone. Activation of CXCR4 on cancer cells transduces Src-dependent survival signals and increases their affinity for microvas- cular endothelial cells.
Several factors have been im- plicated in the skeletal remodelling that accompanies metastatic bone cancer. One of the most studied
mediators of bone metastasis is parathyroid hormone-related peptide (PTHrP). More than 90% of breast cancer bone metastases are PTHrP- positive, whereas only 20% of breast cancers in non-bone metastatic sites express PTHrP. PTHrP produced by breast cancer cells binds to parathy- roid hormone receptor 1 on osteo- blasts and stimulates upregulation of receptor activator of nuclear fac- tor κB ligand (RANKL). RANKL then binds to its receptor on osteoclasts and promotes cell differentiation and
activation. Bone resorption liberates from the bone matrix transforming growth factor β, which then binds to its receptor on cancer cells and ac- tivates a positive feedback loop by signalling for increased production of PTHrP from cancer cells. Agents that target RANKL and interfere with osteoclastogenesis decrease the incidence of skeletal complica- tions and are the current standard of care for patients with bone me- tastasis [16].
Fig. 3.7.5. The role of platelet-derived growth factor receptor β (PDGFR-β) in the pathogenesis of hormone-refractory prostate cancer bone metastasis. (A) Circulating prostate cancer cells ensure their retention in the bone by exploiting their sialyl LewisX antigen to adhere to E-selectin expressed on bone endothelial cells. (B) Extravasating tumour cells express interleukin-6 (IL-6), which is chemotactic for monocytes and imparts M2 functional properties to macrophages. Tumour necrosis factor α (TNFα) produced by macrophages stimulates prostate cancer cells to increase their production of IL-6. (C) TNFα activates osteoclasts and signals for activation of the bone remodelling process. (D) Transforming growth factor β (TGF-β) released from the bone matrix binds to its receptor on prostate cancer cells, where it stimulates production of platelet-derived growth factor (PDGF). (E) PDGF acts in an autocrine manner by binding to PDGFR-β expressed on cancer cells and stimulating cell division. (F) Phosphorylation of PDGFR-β on tumour-associated endothelial cells promotes activation of angiogenic programmes in resident endothelial cells.
Chapter 3.7 • Tumour microenvironment 243
3 B
IO LO
G Y
C H
A P
T E
R 3
.7
Mortality from prostate cancer usually results from metastasis of hormone-refractory prostate cancer cells. The spread of pros- tate cancer to the bone is thought to involve both anatomical com- ponents (i.e. the Batson venous plexus) and site-specific molecular adhesive interactions between the tetrasaccharide sialyl LewisX anti- gen on prostate cancer cells and its receptor E-selectin, which is constitutively expressed on bone endothelial cells. Preclinical studies of androgen-independent prostate cancer cells implanted into the bone of mice implicated platelet-derived growth factor receptor β (PDGFR-β) in metastatic progression in that in- hibition of PDGFR-β activation with imatinib therapy produced apoptosis
of both cancer cells and the tumour- associated blood vessels. PDGFR-β signalling promotes survival of bone endothelial cells, and the receptor is expressed in its phosphorylated form on tumour-associated endothe- lial cells and cancer cells in clinical bone marrow samples from prostate cancer patients. PDGFR-β was also identified as one of the five genes that predicted recurrence after pros- tatectomy. The role of PDGFR-β in hormone-refractory prostate cancer bone metastasis is illustrated in Fig. 3.7.5.
Conclusion Over the past 5 years, the number of investigations into the cellular inter- actions that take place in the tumour microenvironment has increased
dramatically. During this same pe- riod, both the United States National Cancer Institute and the European Commission have established major research initiatives focused exclu- sively on an improved understand- ing of the cellular communication that takes place between cancer cells and the stromal compartment. One consequence of these efforts has been the realization that thera- peutic targeting of both cancer cells and components of the microenvi- ronment leads to improved clinical outcomes. Continued study of the complex cross-talk between cancer cells and the microenvironment of target organs of metastasis is likely to yield new opportunities for thera- peutic intervention.
References
1. Fidler IJ, Yano S, Zhang RD et al. (2002). The seed and soil hypothesis: vascularisation and brain metastases. Lancet Oncol, 3:53– 57. http://dx.doi.org/10.1016/S1470-2045 (01)00622-2 PMID:11905606
2. Weis SM, Cheresh DA (2011). Tumor angiogenesis: molecular pathways and therapeutic targets. Nat Med, 17:1359– 1370. http://dx.doi.org/10.1038/nm.2537 PMID:22064426
3. Eberhard A, Kahlert S, Goede V et al. (2000). Heterogeneity of angiogenesis and blood vessel maturation in human tumors: implications for antiangiogenic tumor therapies. Cancer Res, 60:1388– 1393. PMID:10728704
4. Cascone T, Herynk MH, Xu L et al. (2011). Upregulated stromal EGFR and vascular remodeling in mouse xenograft models of angiogenesis inhibitor-resistant hu- man lung adenocarcinoma. J Clin Invest, 121:1313–1328. http://dx.doi.org/10.1172/ JCI42405 PMID:21436589
5. Shojaei F, Wu X, Malik AK et al. (2007). Tumor refractoriness to anti-VEGF treatment is mediated by CD11b+Gr1+ myeloid cells. Nat Biotechnol, 25:911– 920. http://dx.doi.org/10.1038/nbt1323 PMID:17664940
6. Vaupel P, Kallinowski F, Okunieff P (1989). Blood flow, oxygen and nutrient supply, and metabolic microenvironment of human tumors: a review. Cancer Res, 49:6449– 6465. PMID:2684393
7. Qiu W, Hu M, Sridhar A et al. (2008). No evidence of clonal somatic genetic alter- ations in cancer-associated fibroblasts from human breast and ovarian carcino- mas. Nat Genet, 40:650–655. http://dx.doi. org/10.1038/ng.117 PMID:18408720
8. Colotta F, Allavena P, Sica A et al. (2009). Cancer-related inflammation, the seventh hallmark of cancer: links to genetic instability. Carcinogenesis, 30:1073–1081. http://dx.doi.org/10.1093/carcin/bgp127 PMID:19468060
9. Sica A, Allavena P, Mantovani A (2008). Cancer related inflammation: the mac- rophage connection. Cancer Lett, 267: 204–215. http://dx.doi.org/10.1016/j.canlet. 2008.03.028 PMID:18448242
10. Page-McCaw A, Ewald AJ, Werb Z (2007). Matrix metalloproteinases and the regula- tion of tissue remodelling. Nat Rev Mol Cell Biol, 8:221–233. http://dx.doi.org/10.1038/ nrm2125 PMID:17318226
11. Thiery JP, Acloque H, Huang RY, Nieto MA (2009). Epithelial-mesenchymal tran- sitions in development and disease. Cell, 139:871–890. http://dx.doi.org/10.1016/j. cell.2009.11.007 PMID:19945376
12. Langley RR, Fidler IJ (2013). The biol- ogy of brain metastasis. Clin Chem, 59:180–189. http://dx.doi.org/10.1373/clin chem.2012.193342 PMID:23115057
13. Bos PD, Zhang XH, Nadal C et al. (2009). Genes that mediate breast cancer metas- tasis to the brain. Nature, 459:1005–1009. http://dx.doi.org/10.1038/nature08021 PMID:19421193
14. Kienast Y, von Baumgarten L, Fuhrmann M et al. (2010). Real-time imaging reveals the single steps of brain metastasis forma- tion. Nat Med, 16:116–122. http://dx.doi. org/10.1038/nm.2072 PMID:20023634
15. Kim SJ, Kim JS, Park ES et al. (2011). Astrocytes upregulate survival genes in tumor cells and induce protection from chemotherapy. Neoplasia, 13:286–298. PMID:21390191
16. Weilbaecher KN, Guise TA, McCauley LK (2011). Cancer to bone: a fatal attraction. Nat Rev Cancer, 11:411– 425. http://dx.doi. org/10.1038/nrc3055 PMID:21593787
Websites Tumor Microenvironment Network: http://tmen.nci.nih.gov/
MicroEnviMet: Seventh Framework Programme: http://www.microenvimet.eu/
244
Summary • Signalling pathways regulate a
plethora of biological processes in all cells of complex organisms.
• Consistent with mediating cell survival, proliferation, response of growth factors, and related bi- ological processes, components of such signalling pathways are often variously modified by mu- tation or epigenetic effects in different types of cancer.
• A range of small-molecule inhib- itors and monoclonal antibod- ies are categorized as targeted therapies because the benefits these agents provide depend upon perturbation of particular signalling pathways in malignant cells.
• One particularly well-studied signalling network is the epi- dermal growth factor receptor (EGFR) pathway. Due to the role of this pathway in maintaining the transformed phenotype in tumours such as non-small cell lung cancer and squamous cell carcinoma of the head and neck, the EGFR pathway is a promi- nent target for pharmacological inhibition.
• A major problem with therapies aimed at inhibiting signalling
3.8 3 BIOLOGY
Signal transduction and targeted therapy Fabio Savarese Martin Holcmann Maria Sibilia
Zdenko Herceg (reviewer) Michael P. Brown (contributor) Nikki Burdett (contributor)
pathways is that tumour cells manifest a wide array of strate- gies, which render them refrac- tory to drug treatment.
• Modern whole-genome se- quencing techniques offer great potential for the provision of op- timal targeted therapy on an in- dividual basis, as demonstrated by their initial successful appli- cation in the clinic.
Responsiveness to extracellular signals is a fundamental attribute of all living organisms. For instance, bacteria sense nutrition gradients through signalling cascades medi- ated by transmembrane receptors, which result in bacterial motion to- wards nutrients [1]. This paradigm of receptor-mediated signal trans- duction reaches its highest level of complexity in mammals, where hun- dreds of different cell types commu- nicate with each other – both locally, as involving the paracrine system, and over relatively vast distances, as identified by the endocrine system. Virtually every biological process is at least co-regulated by the integra- tion of extracellular signals from cells and organs. This chapter focuses on the molecular events that mediate signal transduction, how these sig- nalling cascades are dysregulated
in cancer, and how targeting these signalling networks may increase the efficiency of anticancer therapy.
Receptor tyrosine kinases – molecular biology and functions In humans, various signalling cas- cades with distinct molecular proper- ties play key roles in health and dis- ease. The major signalling pathways perturbed in various tumour types (Table 3.8.1) mediate organ devel- opment and homeostasis in normal tissue. The receptor tyrosine kinases identify a specific group of receptor molecules [2]. In humans, 58 differ- ent receptor tyrosine kinases have been identified, grouped into 20 subfamilies. Members of the ERBB receptor subfamily, namely ERBB1 (also named epidermal growth factor receptor [EGFR]), ERBB2, ERBB3, and ERBB4, all play important roles in development and homeostasis but are dysregulated in various human tumours and therefore constitute therapeutic targets in the treatment of cancer [3].
Generally, receptor tyrosine kinas- es function as homodimers and het- erodimers, including the ERBB family members. Each monomer consists of an extracellular N-terminal domain, which primarily serves as a binding site for various ligands; a hydropho- bic transmembrane domain; and an
Chapter 3.8 • Signal transduction and targeted therapy 245
3 B
IO LO
G Y
C H
A P
T E
R 3
.8
Table 3.8.1. Features of the major signalling pathways and representative cancers in which these pathways play a role
Pathway Molecular features of signal transduction Examples of associated cancers
RTKs Receptor tyrosine kinases (RTKs) initiate phosphorylation cascades that involve adaptor proteins and various mitogen-activated protein kinases (MAPKs), ultimately modulating transcription of genes involved in proliferation and differentiation.
Most cancers
PI3K Activation by RTKs leads to the chemical modification of components of the cell membrane, which through a multistep process leads to increased cell growth and survival.
Ovarian cancer
JAK-STAT Activation by receptors, like cytokine receptors or the RTK family. Directly activates proteins called signal transducer and activator of transcription (STATs).
Lymphoma
NF-κB Cytokine receptors activate a complex protein network, which through proteolysis of inhibitory factors ultimately leads to the nuclear translocation and activation of transcription factors belonging to the NF-κB family.
Lymphoma, breast cancer
Notch Activation of the transmembrane receptor Notch leads to the proteolysis of its intracellular domain, which shuttles to the nucleus and ultimately serves as a transcription factor.
Pancreatic cancer
Hedgehog Ligand binding to the receptor Patched leads to its inactivation and inhibits the repressive role of Patched on another receptor, Smoothened. Activated Smoothened initiates a signalling cascade that results in changes in gene expression.
Basal cell carcinoma, medulloblastoma
WNT In a very complex signalling cascade, activation of various receptors by WNT ligands inhibits the activity of an important cytoplasmic kinase, GSK3B, which in turn stabilizes β-catenin protein, which subsequently shuttles to the nucleus and serves as a transcription factor.
Colon cancer
TGF-β Activation of the TGF-β receptors phosphorylates proteins called SMADs, which subsequently enter the nucleus and regulate expression of genes required for processes such as proliferation or cell motility.
Breast cancer, colon cancer
NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; PI3K, phosphatidylinositol 3-kinase; TGF-β, transforming growth factor β.
intracellular C-terminal region con- taining the catalytic kinase domains (Fig. 3.8.1). Signal transduction is initiated by binding of a ligand to the extracellular domain of a receptor molecule, inducing homodimeriza- tion when a second identical receptor molecule is involved, or heterodimer- ization, which involves a different re- ceptor molecule, usually of the same family, with subsequent conforma- tional changes. Dimerization causes activation of the intracellular kinase domain within the receptor. This leads to phosphorylation (usually in trans) of various tyrosine residues, which are located in the C-terminal part of these proteins (Fig. 3.8.1). In biochemical terms, the phosphate group (PO4) of ATP, the phosphate donor in most biochemical reactions, is covalently linked to a tyrosine residue of the re- ceptor by the catalytic activity of its kinase domain. As is discussed later, these two steps, namely receptor di- merization and ATP catalysis, under- pin therapeutic strategies targeting signalling pathways [4].
Once receptor phosphorylation has occurred, multiple proteins en- sure that the original signal is ampli- fied and transduced within the cell to different compartments. Adaptor
proteins bind to the phosphorylated residues of the receptor tyrosine ki- nases and propagate the stimulus by protein–protein interactions, many of which result in further phosphoryla- tion events. Particular propagators of the signalling cascade initiated by ERBB receptors are the protein SOS and the small GTPase RAS, which are frequently mutated in various cancers. Cellular processes affected by ERBB signalling include (i) the reg- ulation of gene expression, mediated mainly by mitogen-activated protein kinase (MAPK)-regulated activity of the activator protein 1 (AP1) tran- scription factors; (ii) AKT-mediated cell survival; and (iii) changes in cell motility, controlled by RHO and RAC proteins. This signalling cascade is kept in a dynamic state as multiple molecular mechanisms mediate the turnover of activated receptor ty- rosine kinases. These mechanisms range from the dephosphorylation of the cytoplasmic domains by phos- phatases to the internalization and ubiquitin-mediated degradation of receptor complexes, all intended to terminate the activating signal. Mutations in genes that are required for turning off activated receptor
tyrosine kinases may contribute to cancer development.
Functions of ErbB signalling in health and disease Analyses of mice carrying mutated alleles of the genes encoding the various ErbB receptors have great- ly contributed to understanding the role of these proteins in various tis- sues and organs. In general, the ge- netic inactivation of ErbB2, ErbB3, or ErbB4 leads to early embryonic lethality in mice [5]. Likewise, mice lacking Egfr show embryonic lethal- ity or die during the first weeks af- ter birth, depending on the genetic background, with defects in sev- eral epithelia and in the brain [6]. Conditional inactivation of the Egfr allele has contributed to overcoming this early lethality and therefore to an improved understanding of the function of Egfr signalling in adult tis- sues [7]. The analysis of animals in which Egfr was specifically mutated in distinct cell types has revealed roles in epithelial development and differentiation as well as in liver regeneration and proper brain pat- terning and function. Egfr signal- ling in these tissues induces (i) the
246
suppression of differentiation at the expense of proliferation and (ii) pro- survival signalling. Increased pro- liferation and aberrant cell survival contribute to cancer development, and indeed various studies in mice have revealed the role of aberrant Egfr signalling in different models of murine tumorigenesis.
Multiple components of the ERBB signalling network may be mutated in cancer. EGFR stimulates basic cellular functions such as pro- liferation and cell survival, and muta- tions that lead to enhanced EGFR signalling are frequently found in tu- mour cells [8]. Several different mu- tations may result in pro-tumorigenic signalling through EGFR; the most notable involve (i) overexpression, (ii) activating mutations, (iii) hetero- dimerization with other ERBB family members, (iv) autocrine ligand pro- duction, and (v) reduced downregu- lation of ERBB signalling (Fig. 3.8.2). Interestingly, different types of muta- tions are found in different tumours. For example, EGFR overexpression is frequently found in brain, liver, and colon cancers, whereas activating mutations are found in non-small cell
lung cancer (e.g. EGFRL858R) and glioblastomas (e.g. EGFRvIII: dele- tion of exons 2–7). Mechanistically, overexpression may be caused by higher transcriptional activity of the corresponding genes or by gene amplification. Often, in individual tumours of patients with a particular type of cancer, increased transcrip- tional activity as well as gene dupli- cation can be observed, alone or together with activating mutations, such as in colorectal cancer or non- small cell lung cancer.
Role of EGFR in distinct types of cancer EGFR and glioblastoma Both gene amplification and activat- ing mutations of EGFR can be found in patients with glioblastoma multi- forme. Two subtypes of glioblastoma can be distinguished, based on their ontogeny: primary glioblastomas arise de novo, whereas secondary glioblastomas develop from low- grade astrocytoma. Interestingly, 97% of primary glioblastomas are known to contain additional cop- ies of the EGFR gene but show no
mutations in the tumour suppres- sor p53 protein encoded by TP53. In addition to EGFR amplification, the EGFRvIII mutation is frequently encountered in glioblastoma, and overall more than 60% of all types of glioblastoma show dysregulated EGFR expression or function. In contrast, secondary glioblastomas are associated with frequent mu- tations in TP53, while EGFR am- plifications or mutations are rare. Unfortunately, targeted anti-EGFR therapies have been shown to have very limited success in the treatment of glioblastomas [9].
EGFR, skin cancers, and tumours of the head and neck Skin cancer is one of the most fre- quent diseases in humans worldwide. One of its most common forms is squamous cell carcinoma, and EGFR is frequently amplified and EGFR overexpressed in squamous cell car- cinoma. In melanoma, a frequently metastatic skin cancer, EGFR is often amplified in late-stage tumours.
In addition to squamous cells in skin tumours, EGFR is also fre- quently overexpressed in head and neck squamous cell carcinomas (HNSCCs), which arise from mu- cosal epithelia of the head and neck (lips, oral and nasal cavities, larynx, or pharynx). High EGFR levels are found in 90% of all HNSCCs and in- dicate a poor prognosis. More than 80% of HNSCC patients with nor- mal EGFR expression survive longer than 5 years, compared with 25% of patients in whom EGFR overexpres- sion has been detected. Currently, the most frequently used therapeu- tic approach in HNSCC patients is a combination of radiotherapy and targeted anti-EGFR therapies [10].
EGFR and non-small cell lung carcinoma Non-small cell lung carcinoma refers to any type of epithelial lung cancer other than small cell lung carcinoma. Lung cancers account for about one third of all cancer deaths worldwide, and more than 80% of all lung tu- mours can be classified as non-small
Fig. 3.8.1. (A,B) Structural change in the transmembrane protein epidermal growth factor receptor (EGFR) caused by binding of its ligand. (A) EGFR (blue) is a transmembrane receptor with an extracellular ligand-binding domain (the ligand EGF is shown in red), a transmembrane domain (the membrane is drawn in grey), and an intracellular domain. (B) Binding of EGF leads to receptor dimerization and activation of the intracellular kinase domains. (C) Principles of signal transduction via ERBB receptors: (1) ligand binding leads to receptor dimerization and the activation of the intracellular kinase domain; (2) various adaptor proteins and kinases lead to the transduction of the signal into the nucleus; (3) activation of transcription factors results in induction of gene expression.
Chapter 3.8 • Signal transduction and targeted therapy 247
3 B
IO LO
G Y
C H
A P
T E
R 3
.8
cell lung carcinomas, which are main- ly adenocarcinoma and squamous cell carcinoma. The most common genetic mutations found in non-small cell lung carcinoma affect the genes that express EGFR and K-RAS, a signalling protein downstream of re- ceptor tyrosine kinases, demonstrat- ing the role of dysregulated ERBB signalling in this disease. EGFR and K-RAS mutations are mutually exclu- sive, indicating that a constitutively active downstream mediator (RAS) of the ERBB cascade does not require increased receptor levels for onco- genicity. Non-small cell lung carcino- mas frequently display specific EGFR mutations, which are very rarely ob- served in other cancers, and which result in expression of mutated EGFR protein with constitutively active ki- nase domains. Anti-EGFR targeted therapies are singularly effective in patients harbouring these mutations, although ultimately hampered by the development of resistance [11].
Therapeutic approaches that target signal transduction Various molecular strategies aimed at the functional inactivation of onco- genes are currently being developed and tested in numerous clinical trials [3]. At present, however, all types of anticancer drugs that inhibit recep- tor tyrosine kinases either belong to the class of small-molecule inhibi- tors or are recombinant antibodies that bind and inhibit the specific receptors (Table 3.8.2). Advances in resolving protein structures have facilitated the design of small mole- cules targeting oncoproteins specifi- cally. For a protein to be “druggable” by this approach, the protein must have an active domain shaped like a cleft or a pocket that is accessible to water. Small-molecule inhibitors of protein kinases, exemplified by the intracellular domains of the ERBB receptors, primarily affect enzymatic activity by inhibiting ATP catalysis occurring within the kinase domain (Fig. 3.8.3).
Several parameters influence the efficacy of small-molecule inhibitors. Drugs binding irreversibly to their
target permanently inhibit kinase function, as opposed to the activity of reversible inhibitors. Irreversible inhibition is more effective but much more toxic. Another param- eter is the specificity of the drug – either targeting the ATP-binding domain of a kinase, which usually leads to a highly specific drug, or targeting ATP itself. Recent studies demonstrate frequent interactions of protein kinase inhibitors, previ- ously believed to be specific, with multiple other protein kinases. These off-target effects present a major challenge since they might be the cause of potentially detrimental side- effects. However, in some cases this non-selectiveness can prove benefi- cial, as in the case of sorafenib [12]. Sorafenib was originally identified as a RAF inhibitor, but the clinical suc- cess of this agent is attributable to an inhibitory effect on the kinases of the vascular endothelial growth factor receptor 2 (VEGFR2), plate- let-derived growth factor receptor (PDGFR), and fibroblast growth fac- tor receptor (FGFR) types. Similarly, oral multikinase inhibitors such as sunitinib target PDFGRs, VEGFRs,
KIT, RET, CSF1R, and FLT3 and have been successfully used for the cure of renal cell carcinoma and imatinib-resistant gastrointestinal stromal tumours.
As well as small-molecule inhibi- tors, monoclonal antibodies to the extracellular domain of transmem- brane receptors have been success- fully established for the treatment of certain types of cancer [13]. An example is the chimeric antibody cetuximab, which inhibits hetero- dimerization and consequently the activation of the signalling cascade downstream of EGFR (Fig. 3.8.3). In addition, antibodies may act via the induction of an anti-tumour immune response by a process designated antibody-dependent cytotoxicity. The difference in size between the small-molecule inhibitor erlotinib and the antibody cetuximab is illustrated in Fig. 3.8.3.
Clinical application of targeted therapies In most cases, tumours from patients with non-small cell lung carcinoma exhibit hyperactive EGFR signalling. Hence, when drugs that selectively
Fig. 3.8.2. Various types of structural change that may result in alterations of ERBB expression and that have been described in cancer cells.
248
Table 3.8.2. Monoclonal antibodies and small-molecule inhibitors currently in clinical use that target diverse signal transduction pathways
Generic name Target EMA-approved indications FDA-approved indications
Monoclonal antibodies
Bevacizumab VEGF Metastatic colorectal cancer, breast cancer, non-small cell lung cancer, renal cell carcinoma, grade IV glioma; epithelial ovarian, fallopian tube, or primary peritoneal cancer
Metastatic colorectal cancer, non-small cell lung cancer, glioblastoma, metastatic renal cell carcinoma
Cetuximab EGFR Head and neck squamous cell carcinoma, EGFR-positive K-RAS wild- type colorectal cancer
Head and neck squamous cell carcinoma, EGFR-positive colorectal cancer
Ipilimumab CTLA-4 Unresectable or metastatic melanoma Unresectable or metastatic melanoma
Panitumumab EGFR K-RAS wild-type metastatic colorectal cancer (conditional)
K-RAS-negative metastatic colorectal cancer
Pertuzumab HER2 HER2-positive unresectable or metastatic breast cancer
HER2-positive metastatic breast cancer
Trastuzumab HER2 HER2-positive breast cancer; HER2- positive gastric or gastro-oesophageal junction adenocarcinoma
HER2-positive breast cancer, HER2- positive gastric or gastro-oesophageal junction cancer
Small-molecule inhibitors
Afatinib EGFR, HER2, HER4 Treatment of EGFR tyrosine kinase inhibitor-naive adult patients with locally advanced or metastatic non-small cell lung cancer with activating EGFR mutation(s)
First-line treatment of patients with metastatic non-small cell lung cancer whose tumours have EGFR exon 19 deletions or exon 21 (L858R) substitution mutations as detected by an FDA- approved test
Axitinib VEGFR, PDGFR Advanced renal cell carcinoma Advanced renal cell carcinoma
Cabozantinib VEGFR, MET, RET None Metastatic medullary thyroid cancer Crizotinib ALK, MET ALK-positive non-small cell lung cancer
(conditional) ALK-positive non-small cell lung cancer
Dabrafenib V600E BRAF None BRAF V600E-positive unresectable or metastatic melanoma
Erlotinib EGFR EGFR mutation-positive non-small cell lung cancer, pancreatic cancer
Non-small cell lung cancer, pancreatic cancer
Gefitinib EGFR Locally advanced or metastatic EGFR mutation-positive non-small cell lung cancer
Non-small cell lung cancer
inhibit EGFR, namely the small- molecule inhibitors gefitinib and erlotinib, first became available, efficient treatment of this disease was anticipated. However, from the outset, clinical trials indicated that only a small percentage of patients responded well to treatment [14]. It became apparent that the therapy was beneficial for patients carrying mutations in the kinase domain of EGFR but not for individuals with tumours that overexpress wild-type EGFR. However, secondary EGFR mutations arose, specifically in pa- tients initially responsive to gefitinib and erlotinib, and these mutations rendered the protein resistant to
inhibition by these compounds. In preclinical trials, treatment of such gefitinib/erlotinib-resistant tumour cells with other classes of drugs, such as the EGFR/ErbB2 inhibitor lapatinib or the irreversible kinase inhibitor afatinib, overcame resis- tance towards anti-EGFR therapy. However, due to toxicity, a com- binatorial treatment has yet to be adopted clinically.
In HNSCC patients in whom ab- normally activated EGFR signalling has been detected, a combination of radiotherapy and the EGFR-targeting antibody cetuximab is widely used at present. Cetuximab is a humanized mouse antibody, and binding of this
antibody to the extracellular part of EGFR not only inhibits the activation of the downstream signalling cas- cade but also inhibits receptor turn- over and renders tumour cells more immunogenic, leading to increased levels of antibody-dependent cellular toxicity. However, similarly to non- small cell lung carcinoma patients treated with gefitinib and erlotinib, HNSCC patients quickly develop resistance to cetuximab [15]. In gen- eral, the use of EGFR-targeted drugs currently leads to prolonged survival of patients diagnosed with cancer, but not to reduced mortality.
In summary, the use of drugs inhibiting the function of EGFR and
ALK, anaplastic lymphoma kinase; CSF1R, colony-stimulating factor 1 receptor; EGFR, epidermal growth factor receptor; EMA, European Medicines Agency; FDA, United States Food and Drug Administration; GnRH, gonadotropin-releasing hormone; mTOR, mammalian target of rapamycin; PDGFR, platelet-derived growth factor receptor; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor.
Chapter 3.8 • Signal transduction and targeted therapy 249
3 B
IO LO
G Y
C H
A P
T E
R 3
.8
Generic name Target EMA-approved indications FDA-approved indications
Imatinib PDGFR, KIT KIT-positive gastrointestinal stromal tumours, adjuvant or unresectable
KIT-positive gastrointestinal stromal tumours, adjuvant or unresectable
Lapatinib EGFR, HER2 HER2-positive breast cancer (conditional)
HER2-positive breast cancer
Pazopanib VEGFR, KIT, PDGFR Advanced renal cell carcinoma, advanced soft tissue sarcoma (conditional)
Advanced renal cell carcinoma, advanced soft tissue sarcoma
Regorafenib PDGFR, KIT, RET, VEGFR
None Metastatic colorectal cancer
Sorafenib VEGR, PDGFR, CRAF, FLT3
Hepatocellular carcinoma, advanced renal cell carcinoma
Unresectable hepatocellular carcinoma, advanced renal cell carcinoma
Sunitinib VEGFR, PDGFR, KIT, FLT3, RET, CSF1R
Advanced renal cell carcinoma, unresectable or metastatic malignant gastrointestinal stromal tumours, pancreatic neuroendocrine tumours
Advanced renal cell carcinoma, progressive gastrointestinal stromal tumours, pancreatic neuroendocrine tumours
Trametinib V600 BRAF None BRAF V600-positive unresectable or metastatic melanoma
Vandetanib VEGFR, EGFR, RET Medullary thyroid cancer (conditional) Medullary thyroid cancer
Vemurafenib V600 BRAF BRAF V600-positive unresectable or metastatic melanoma
BRAF V600E-positive unresectable or metastatic melanoma
Small-molecule allosteric inhibitors
Everolimus mTOR Advanced renal cell carcinoma; unresectable pancreatic neuroendocrine tumours; hormone-receptor-positive, HER2-negative postmenopausal breast cancer; subependymal giant cell astrocytoma (conditional)
Advanced renal cell carcinoma; unresectable pancreatic neuroendocrine tumours; subependymal giant cell astrocytoma; hormone-receptor-positive, HER2-negative postmenopausal breast cancer
Temsirolimus mTOR Advanced renal cell carcinoma Advanced renal cell carcinoma
Small-molecule Smoothened antagonist
Vismodegib Smoothened receptor None Metastatic or recurrent basal cell carcinoma
Recombinant decoy receptor
Ziv-aflibercept VEGFR Metastatic colorectal cancer Metastatic colorectal cancer
Antibody–drug conjugate
Ado-trastuzumab emtansine
HER2 None HER2-positive metastatic breast cancer
Hormonal agents
Abiraterone Androgen synthesis– CYP17A1 inhibition
Metastatic castration-resistant prostate cancer
Metastatic castration-resistant prostate cancer
Anastrazole Aromatase None Hormone-receptor-positive postmeno- pausal breast cancer, or hormone-recep- tor-negative disease after tamoxifen
Bicalutamide Androgen receptor None Metastatic prostate cancer
Degarelix GnRH receptor Advanced prostate cancer Advanced prostate cancer Enzalutamide Androgen receptor None Metastatic castration-resistant prostate
cancer Exemestane Aromatase None Estrogen-receptor-positive postmeno-
pausal breast cancer, or estrogen-recep- tor-negative disease after tamoxifen
Flutamide Androgen receptor None Prostate cancer Goserelin GnRH receptor None Prostate cancer, palliative treatment of
premenopausal breast cancer Letrozole Aromatase None Postmenopausal breast cancer
Leuprorelin acetate GnRH receptor None Palliative treatment of prostate cancer
Tamoxifen Estrogen receptor None Metastatic breast cancer
ALK, anaplastic lymphoma kinase; CSF1R, colony-stimulating factor 1 receptor; EGFR, epidermal growth factor receptor; EMA, European Medicines Agency; FDA, United States Food and Drug Administration; GnRH, gonadotropin-releasing hormone; mTOR, mammalian target of rapamycin; PDGFR, platelet-derived growth factor receptor; VEGF, vascular endothelial growth factor; VEGFR, vascular endothelial growth factor receptor.
Table 3.8.2. Monoclonal antibodies and small-molecule inhibitors currently in clinical use that target diverse signal transduction pathways (continued)
250
other receptor tyrosine kinases achieves an initial response, but the regrowth of tumours resistant to these drugs is commonly ob- served; such a scenario is attribut- able to inherent genetic instability, as evidenced by rapid mutation. Emergence of new mutations under the selective pressure of treatment with receptor tyrosine kinase inhibi- tors allows tumour cells to contin- ue growth despite drug treatment. Ideally, this scenario is resolved by the identification of secondary mu- tations and the development of new treatment strategies. Although great progress has been made in develop- ing targeted anticancer drugs to be applied specifically in preselected cancer patients, mechanisms of resistance have yet to be fully elucidated.
A better understanding of resis- tance mechanisms would lead to the identification of new targets against which new drugs could be designed and developed. Importantly, mecha- nisms mediating resistance to broad classes of drugs have to be identi- fied and to be distinguished from processes specific to particular agents. The search for biomarkers, which predict the response of certain tumours or patients to therapeutic treatment, has thus become a focus
of many pharmaceutical companies and research laboratories [16].
The catch of anticancer therapy: acquired drug resistance Research addressing mechanisms of resistance to therapy targeted against signalling pathways, includ- ing anti-EGFR drugs, has markedly increased in recent years. Currently, recognized resistance mechanisms to anti-EGFR drugs include (i) addi- tional mutations of EGFR; (ii) com- pensatory signalling via other recep- tors, such as VEGFR, insulin-like growth factor receptor (IGFR), and AXL; (iii) increased expres- sion of VEGF, which modulates angiogenesis and may reduce lo- calized drug concentration; and (iv) mutations in molecules down- stream of EGFR, such as RAS or RAF (Fig. 3.8.4). Such downstream activating mutations render the tu- mour cells independent of EGFR function and therefore refractory to treatment with drugs against EGFR [17].
Mutations in downstream media- tors of EGFR signalling commonly occur in tumours exhibiting neither EGFR mutation nor overexpression. An example involves metastatic melanoma, in which an activating
mutation in BRAF, V600E, is fre- quently found. Recently, treatment of metastatic melanoma with the BRAF inhibitor vemurafenib yield- ed promising initial results [18]. In one clinical study, more than 50% of patients receiving vemurafenib experienced prolonged survival compared with untreated pa- tients, and in an independent study more than 80% of treated patients showed partial to complete regres- sion of tumours compared with non-treated individuals. However, tumour regression lasted only 2–18 months, and subsequently vemurafenib-resistant tumours ap- peared. Furthermore, vemurafenib had severe side-effects, giving rise in some cases to cutaneous squa- mous cell carcinoma or keratoacan- thoma [19].
New insights gained from the genomic revolution The new millennium has seen the complete sequencing of the human genome and genomes of other im- portant organisms and pathogens. In particular, recent advances in whole-genome sequencing are revolutionizing modern medicine [20]. A pioneering study compar- ing the whole genomes of tumour and normal cells in a patient with acute myeloid leukaemia led to the discovery of various somatic mu- tations in malignant cells, one of them affecting the gene encoding the receptor tyrosine kinase FLT3 [21], demonstrating the power of modern genomic techniques. With such powerful tools available, a ma- jor goal for cancer therapy comes within reach: the identification of ge- netic mutations responsible for can- cer development as well as those occurring during cancer progres- sion. This would enable treatment strategies to be adapted based on the genetic changes during the course of the disease. In particu- lar, mutations in cellular processes that “drive” tumorigenesis present attractive targets for therapeutic in- tervention. Notably, more than half of the “tumour driver” gene classes
Fig. 3.8.3. (A) The action of tyrosine kinase inhibitors (TKIs) and inhibitory antibodies (cetuximab). TKIs inhibit signal transduction by blocking the activation of the receptor through inhibition of the kinase domain. In contrast, cetuximab impairs the dimerization of epidermal growth factor receptor (EGFR). (B) Size comparison between ATP and erlotinib, and between EGFR and cetuximab. Erlotinib is a small-molecule inhibitor and is approximately the same size as ATP, the molecule essential for the phosphorylation of the receptor after dimerization. Cetuximab and EGFR are approximately 400 times bigger, as indicated by the magnifying glass.
Chapter 3.8 • Signal transduction and targeted therapy 251
3 B
IO LO
G Y
C H
A P
T E
R 3
.8
are constituted by components of signalling pathways [22]. Whole- genome sequencing of cancer pa- tients leads to the discovery of se- quence variations that can be used to predict whether a patient or an individual tumour can respond to a certain drug, which is referred to as pharmacogenomics [23].
Prospects Ever-expanding knowledge about signal transduction networks un- derpins the design of novel agents and improved anticancer strategies. However, greater knowledge also re- veals another degree of complexity in relation to contemporary understand- ing of tumour biology. Consequently, for every drug that successfully tar- gets a specific pathway, multiple re- sistance mechanisms following from the initial treatment may be antici- pated. Such an outlook, if applicable universally, is daunting but may be offset by the following considerations.
First, the expanding repertoire of drugs targeting proteins mediating every step of a signalling cascade will also increase our repertoire for ameliorating secondary and subse- quent mutations. Perhaps even more importantly, modern screening tech- niques of gene expression patterns, or the identification of novel biomark- ers of individual tumours, will allow
more informed development of com- binations of therapeutics that have the highest chance of success in an individual patient. The introduction of massively parallel DNA sequencing technologies will soon provide com- prehensive data at low cost. The term “personalized medicine” will no longer connote response to a single
agent but will indicate therapy based on the biological characteristics of individual tumours, specified with reference to tumour progression in real time [24].
We thank Thomas Bauer for the artwork and design of the figures.
Fig. 3.8.4. Resistance mechanisms against drugs targeting the ERBB signalling network: (1) activating mutations in downstream effector molecules make the pathway completely independent of receptor function; (2) activating mutations in the kinase domain lead to constitutively active receptors; (3) mutations in the extracellular domain interfere with binding of cetuximab; (4) overproduction of vascular endothelial growth factor (VEGF) leads to altered angiogenesis and reduced exposure to drugs; (5) compensatory signal- ling via other receptors.
252
References
1. Falke JJ, Bass RB, Butler SL et al. (1997). The two-component signaling pathway of bacterial chemotaxis: a molecular view of signal transduction by receptors, kinas- es, and adaptation enzymes. Annu Rev Cell Dev Biol, 13:457–512. http://dx.doi. org /10.114 6/annurev.c ellbio.13.1.4 57 PMID:9442881
2. Lemmon MA, Schlessinger J (2010). Cell signaling by receptor tyrosine kinases. Cell, 141:1117–1134. http://dx.doi.org/10.1016/j. cell.2010.06.011 PMID:20602996
3. Levitzki A, Klein S (2010). Signal transduc- tion therapy of cancer. Mol Aspects Med, 31:287–329. http://dx.doi.org/10.1016/j. mam.2010.04.001 PMID:20451549
4. Schlessinger J (2002). Ligand-induced, receptor-mediated dimerization and ac- tivation of EGF receptor. Cell, 110:669– 672. http://dx.doi.org/10.1016/S0092-8674 (02)00966-2 PMID:12297041
5. Sibilia M, Kroismayr R, Lichtenberger BM et al. (2007). The epidermal growth factor receptor: from development to tu- morigenesis. Differentiation, 75:770–787. http://dx.doi.org/10.1111/j.1432-0436.2007. 00238.x PMID:17999740
6. Sibilia M, Wagner EF (1995). Strain- dependent epithelial defects in mice lacking the EGF receptor. Science, 269: 234–238. http://dx.doi.org/10.1126/science. 7618085 PMID:7618085
7. Wagner B, Natarajan A, Grünaug S et al. (2006). Neuronal survival depends on EGFR signaling in cortical but not midbrain astrocytes. EMBO J, 25:752–762. http:// dx.doi.org/10.1038/sj.emboj.76 0 0 9 88 PMID:16467848
8. Normanno N, De Luca A, Bianco C et al. (2006). Epidermal growth factor recep- tor (EGFR) signaling in cancer. Gene, 366:2–16. http://dx.doi.org/10.1016/j.gene. 2005.10.018 PMID:16377102
9. Mellinghoff IK, Wang MY, Vivanco I et al. (2005). Molecular determinants of the re- sponse of glioblastomas to EGFR kinase inhibitors. N Engl J Med, 353:2012–2024. http://dx.doi.org/10.1056/NEJMoa051918 PMID:16282176
10. Sundvall M, Karrila A, Nordberg J et al. (2010). EGFR targeting drugs in the treatment of head and neck squamous cell carcinoma. Expert Opin Emerg Drugs, 15:185–201. http:// dx.doi.org/10.1517/14728211003716442 PMID:20415599
11. da Cunha Santos G, Shepherd FA, Tsao MS (2011). EGFR mutations and lung cancer. Annu Rev Pathol, 6:49–69. http://dx.doi.org/10.1146/annurev-pathol- 011110-130206 PMID:20887192
12. Adnane L, Trail PA, Taylor I, Wilhelm SM (2006). Sorafenib (BAY 43-9006, Nexavar), a dual-action inhibitor that targets RAF/ MEK/ERK pathway in tumor cells and tyrosine kinases VEGFR/PDGFR in tumor vasculature. Methods Enzymol, 407:597– 612. http://dx.doi.org/10.1016/S0076-6879 (05)07047-3 PMID:16757355
13. Astsaturov I, Cohen RB, Harari P (2007). EGFR-targeting monoclonal antibod- ies in head and neck cancer. Curr Cancer Drug Targets, 7:650–665. http:// d x . d o i . o r g / 1 0 . 2 174 / 15 6 8 0 0 9 0 7 7 8 2 418365 PMID:18045070
14. Sridhar SS, Seymour L, Shepherd FA (2003). Inhibitors of epidermal-growth- factor receptors: a review of clinical re- search with a focus on non-small-cell lung cancer. Lancet Oncol, 4:397– 406. http:// dx.doi.org/10.1016/S1470-2045(03)01137- 9 PMID:12850190
15. Brand TM, Iida M, Wheeler DL (2011). Molecular mechanisms of resistance to the EGFR monoclonal antibody ce- tuximab. Cancer Biol Ther, 11:777–792. http://dx.doi.org/10.4161/cbt.11.9.15050 PMID:21293176
16. Brooks JD (2012). Translational geno- mics: the challenge of developing cancer biomarkers. Genome Res, 22:183–187. http://dx.doi.org/10.1101/gr.124347.111 PMID:22301132
17. Wheeler DL, Dunn EF, Harari PM (2010). Understanding resistance to EGFR inhibi- tors – impact on future treatment strate- gies. Nat Rev Clin Oncol, 7:493–507. http:// dx.doi.org/10.1038/nrclinonc. 2010.97 PMID:20551942
18. Chapman PB, Hauschild A, Robert C et al.; BRIM-3 Study Group (2011). Improved sur- vival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med, 364:2507–2516. http://dx.doi.org/10.1056/ NEJMoa1103782 PMID:21639808
19. Anforth R, Fernandez-Peñas P, Long GV (2013). Cutaneous toxicities of RAF inhibi- tors. Lancet Oncol, 14:e11–e18. http://dx. doi.org/10.1016/S1470-2045(12)70413-8 PMID:23276366
20. Kilpivaara O, Aaltonen LA (2013). Diagnostic cancer genome sequencing and the contribution of germline vari- ants. Science, 339:1559–1562. http:// d x .d o i.o r g /10.112 6 /s c i e n c e.12 3 3 8 9 9 PMID:23539595
21. Ley TJ, Mardis ER, Ding L et al. (2008). DNA sequencing of a cytogenetically normal acute myeloid leukaemia ge- nome. Nature, 456:66–72. http://dx.doi. org/10.1038/nature07485 PMID:18987736
22. Vogelstein B, Papadopoulos N, Velculescu VE et al. (2013). Cancer genome land- scapes. Science, 339:1546–1558. http:// d x .d o i.o r g /10 .112 6 /s c i e n c e.12 3 512 2 PMID:23539594
23. McLeod HL (2013). Cancer pharmacog- enomics: early promise, but concerted effort needed. Science, 339:1563–1566. http://dx.doi.org/10.1126/science.1234139 PMID:23539596
24. van’t Veer LJ, Bernards R (2008). Enabling personalized cancer medicine through analysis of gene-expression pat- terns. Nature, 452:564–570. http://dx.doi. org/10.1038/nature06915 PMID:18385730
Chapter 3.9 • Immunology and immunotherapy 253
3 B
IO LO
G Y
C H
A P
T E
R 3
.9Summary • Inflammation and immunity affect
cancer initiation, progression, dissemination, co-morbidities, and response to therapy. Cancer often originates in chronically inflamed tissue, due to infections or other causes. Tumour-elicited inflammation contributes to pro- gression and dissemination. Cancer incidence is reduced by eliminating the causes of inflam- mation or by prolonged use of non-steroidal anti-inflammatory drugs.
• Anti-tumour immunity is evident in established tumours and may affect progression negatively or positively. Histological evidence of a brisk immune response is correlated with more favourable prognosis. Tumours, however, use various mechanisms to escape the immune response.
• Tumour immunotherapy proto- cols have used pro-inflammatory therapies, cytokines, antibodies, adoptive T-cell transfer, and a variety of therapeutic vaccines, with mixed success. Antibodies that block immune checkpoints may prevent tumour immune evasion. When successful, im- munotherapy provides more du- rable remissions than cytotoxic
3.9 3 BIOLOGY
Immunology and immunotherapy Giorgio Trinchieri Jean-Pierre Abastado (reviewer)
or targeted therapies, suggest- ing that combined treatment protocols may prove efficacious.
Cancer cells initially maintain altered morphological and functional char- acteristics of their tissue of origin, representing almost a caricature of it. The organism does not ignore the growing tumour but reacts to it similarly to the response evoked by tissue damage and, as in wound re- pair, by establishing a symbiotic rela- tionship that favours tumour growth and dissemination. An adaptive im- mune response reminiscent of that against foreign pathogens is also often present, with variable effects on tumour growth. Inflammation and immunity likely have an instrumental role in cancer initiation, progression, and dissemination and are also re- sponsible for co-morbidities such as anorexia/cachexia [1] (Fig. 3.9.1).
The study of cancer etiopatho- genesis and the identification of ther- apeutic targets were initially focused on intrinsic cancer cell traits that af- fect the ability of malignant cells to proliferate and invade local or dis- tant tissues [2]. For a tumour to thus establish itself or form metastases, not only are the transformed cells (the “seed”) important, but equally so is the tissue that receives them (the “soil”) [3]. Hence, the tumour
microenvironment is key, not only as a participant in the neoplastic pro- cess but also as a possible thera- peutic target [2]. The tumour is a complex organized tissue formed by both the transformed clonal cells and stromal cells. The genetic and epigenetic evolution of successful malignant cells is dependent on their fitness to the microenvironment, re- sulting in optimal access to nutri- ents and tissue remodelling. The microenvironment of most tumours is infiltrated by innate and adaptive immune inflammatory cells that influ- ence the survival and adaptation of the transformed cells.
Inflammation and cancer Chronic inflammation related to bacterial, viral, or parasitic infec- tions has been associated with the etiopathogenesis of about one fifth of cancer cases worldwide, particu- larly in developing or less-developed countries [4]. Particular pathogens such as Helicobacter pylori, human papillomaviruses (HPV), Epstein– Barr virus, and hepatitis B and C viruses account for the majority of those cases. Some of these patho- gens, for example HPV or Epstein– Barr virus, directly transform the in- fected cells, whereas others, such as H. pylori, induce tissue inflamma- tion that favours cancer initiation and progression (Fig. 3.9.2). However,
254
the inflammatory response to onco- genic pathogens may also act as a tumour promoter. Cancer resulting from infection can be prevented by prophylactic vaccination, as is the case for HPV and hepatitis B virus, or by antibiotic treatment in the case of H. pylori. Early data indicate that HPV vaccines are efficient not only in preventing infections but also in preventing development of precan- cerous lesions in the cervix and at other anatomical locations [5]. A re- duction in the prevalence of H. pylori in Caucasian individuals in the USA and other areas of the world has been paralleled by a dramatic de- cline in stomach cancer incidence, although accompanied by increased incidence of oesophageal adenocar- cinoma, which may be associated with concurrent increases in obesity and gastro-oesophageal reflux or with bacterial dysbiosis in the upper digestive tract in the absence of H. pylori [6]. The chronic nature of the cancer-inducing infections and the persistence of pathogenic genes in malignant cells suggest the pos- sibility of targeting infection-driven
cancer with post-exposure vaccines to prevent cancer development or with therapeutic vaccine to treat established cancers.
Other cancers, not associated with infection, originate in chroni- cally inflamed tissues injured by chemical or physical agents, such as irradiation or lung-irritating particles, or are dependent on condi- tions resulting from mutation of di- gestive enzymes or of inflammatory genes. In the absence of pathogens, altered composition or interaction with the commensal microbiota at the body interfaces may result both locally and systemically in a distort- ed inflamed state that favours cancer initiation; an example is increased risk of colon cancer in patients with inflammatory bowel diseases char- acterized by intestinal dysbiosis and chronic inflammation. Intestinal dys- biosis may also mediate increased cancer incidence in obese popula- tions, the pathogenesis of cancer co-morbidities, and chemotherapy side-effects [1].
Infiltrating inflammatory and im- mune haematopoietic cells charac-
terize inflamed tissue, but epithelial, stromal, and endothelial cells also produce and respond to inflammatory mediators, affecting their proliferation and functions and attracting haema- topoietic cells. Activation of onco- genes such as Ras, Myc, Ret, and Src is associated with intrinsic inflam- mation through production of growth factors, cytokines, chemokines, and tissue remodelling enzymes acting on the transformed cells with an au- tocrine feedback loop, while simulta- neously reprogramming the tumour microenvironment and attracting hae- matopoietic cells and affecting their functional differentiation [7].
Chronic inflammation favours cancer initiation by multiple mech- anisms, including genetic and epigenetic instability mediated in part by reactive oxygen species that induce or amplify DNA dam- age after toxic insults. Inflammation also acts as a tumour promoter by providing growth factors and tissue remodelling factors and by support- ing angiogenesis. Once the tumour is established, cancer-induced inflammation, revealed by tumour-
Fig. 3.9.1. The multiple links between inflammation and cancer. With time (left to right), a tumour progresses in size and metastasizes (right). Inflammation (intrinsic in the tissue of tumour origin or induced by the growing tumour, purple) contributes to the mechanisms of tumour induction/progression, co-morbidities, or response to therapy (red). Tumour-specific immune response (blue) generally has an anti-tumour effect, which should be evaded or redirected for the tumour to progress.
Chapter 3.9 • Immunology and immunotherapy 255
3 B
IO LO
G Y
C H
A P
T E
R 3
.9
associated haematopoietic cells, is a constant in most tumours, including those not known to originate from an inflamed tissue. Tumour-elicited inflammation mediates promotion, facilitates tumour dissemination, and creates an immunosuppressive environment responsible for im- mune escape [8]. Thus, for exam- ple, the oncogene-driven secretion of granulocyte-macrophage colony- stimulating factor (GM-CSF) endows tumour-infiltrating myeloid cells with immunosuppressive activity.
Anti-inflammatory therapies and cancer prevention Regular use of aspirin or other non- steroidal anti-inflammatory drugs over extended periods decreases cancer incidence and specifically re- duces the risk of colorectal cancer or polyp recurrence [9]. Meta-analysis of randomized trials of cardiovas- cular disease prevention in which patients took aspirin daily for more than 7.5 years indicated that the 20-year risk of cancer death was reduced by 40% for gastrointesti- nal cancers and by approximately 20% for all solid cancers, including lung, brain, and prostate cancers and melanoma [10]. Non-steroidal anti-inflammatory drugs inhibit the
cyclooxygenases COX-1 and COX-2 that catalyse the production of pros- taglandins from fatty acids. Drugs like aspirin inhibit both COX-1 and COX-2 and thus are toxic to the stomach and intestinal lining, whereas COX-2 in- hibitors do not alter the gastrointesti- nal homeostasis, although their cardi- ovascular toxicity has severely limited their use, especially for prophylactic therapies. While cyclooxygenase in- hibitors are anti-inflammatory, their preventive effect on cancer may also be due to non-inflammation-related effects of prostaglandins on vaso- dilation, angiogenesis, DNA muta- tion rate, epithelial cell adhesion, or apoptosis. At the concentrations used in cardiovascular studies, aspirin in- hibits cyclooxygenases but has no effect on major pathways of cancer- related inflammation, such as those downstream of activation of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), which are inhibited only at much higher dos- ages. The inflammatory pathways hijacked by tumours to promote their own progression are also required for physiological tissue homeostasis, re- sistance to infections, and response to tissue damage. Targeting these molecular pathways without affect- ing their physiological roles may be difficult.
Anti-tumour immunity and immunosurveillance Tumour cell immunogenicity refers to the ability of tumour cells to express antigens recognized by the immune system. These tumour antigens can be: molecules encoded by cancer- germline genes that tend not to be efficiently presented by central tol- erance-inducing medullary thymic epithelial cells; tumour-specific an- tigenic polypeptides resulting from mutation of key regulatory genes, for example the products of acti- vated oncogenes; oncogenic viral proteins; and molecules selectively overexpressed by tumours, including wild-type p53 [11]. For recognition by T cells, antigenic peptides are pre- sented by major histocompatibility complex (MHC) molecules: class I for CD8 cytotoxic T cells and class II for CD4 helper T cells. Loss of MHC expression prevents tumour immu- nogenicity, although tumour antigens can be cross-presented by antigen- presenting cells, such as dendritic cells, in the tumour microenviron- ment. Tumour antigens exposed on the cell surface are recognized by antibodies that may have both anti-tumour and pro-tumour effects (see “Role of the innate and ac- quired immune systems in mam- mary development and breast cancer”). Similarly, anti-tumour T-cell-mediated immunity may have variable effects on tumour growth. Certain types of immune response, in particular T helper type 1 (Th1) responses, are generally considered to be anti-tumour, whereas Th2 re- sponses are pro-tumour growth or favourable to metastasis formation [12]. The more recently described Th17 response in different situations may be either pro- or anti-tumour.
Anticancer immunity is evi- denced by patients with paraneo- plastic neurological degeneration and smouldering tumours in whom neurological disorders are triggered by an antibody response to neuronal antigens expressed on tumour cells [13]. Historically, the presence of lymphocyte infiltrates in tumours has been considered a positive prognos- tic factor for several tumour types.
Fig. 3.9.2. Gastric corpus with chronic (lymphoplasmacellular) and active (neutrophilic) inflammation, due to Helicobacter pylori infection.
256
Role of the innate and acquired immune systems in mammary development and breast cancer Zena Werb
The tumour-promoting role of the immune system has long been rec- ognized. Postnatal development of the mammary gland occurs in an immunocompetent environment. In breast cancer, the immune environ- ment is compromised, allowing neo- plastic progression and metastasis to proceed. Activation of CD4+ T cells by antigen-presenting cells is evident within the developing mammary gland, and production of interferon-gamma by T helper type 1 (Th1) effector T cells negatively regulates postnatal mammary gland development.
First, the importance of tissue antigen-presenting cells was dem- onstrated, by depletion of CD11chigh/ MHCIIhigh mammary antigen-pre- senting cells in vivo in mice and ex vivo in three-dimensional primary or- ganotypic cultures, which increased
ductal invasion and epithelial branching. Second, depletion of α/β T cells, blocking antibodies to MHCII and CD4, and T-cell proliferation as- says all suggested that regulation of mammary postnatal organogen- esis by antigen-presenting cells and CD4+ T cells is antigen-mediated. The production of interferon-gamma by Th1 effector cells inhibited mam- mary postnatal organogenesis by acting directly on the mammary lu- minal epithelial cells. These results suggest a novel regulatory role for the adaptive immune system in nor- mal tissue remodelling, even in the absence of injury or inflammation.
As tumours develop, the inflam- matory compartment changes dra- matically. Tumour-associated mac- rophages and antigen-presenting cells and CD11b+Gr1+ cells increase markedly, starting very early before
malignant conversion. While the tu- mour antigen-presenting cells still interact with T cells, the T cells no longer are activated for anti-tumour functions. These myeloid cells also increase in tissues distant from the primary tumour, establishing sites for metastasis. The expansion of in- flammatory cells contributes to the events stimulating tumour growth and metastasis. Characterization of the molecular pathways that regulate these populations will lead to identification of pathways and molecules that could be targeted to prevent tumour progression and metastasis.
Reference Egeblad M et al. (2010). Dev Cell, 18:884–901. http://dx.doi.org/10.1016/j.devcel.2010.05.012 PMID:20627072
In human colon cancer, a particular immune contexture, characterized by a high density of Th1/cytotoxic memory T lymphocytes located both in the centre and at the invasive margin of the primary tumour, is positively correlated with disease- free and overall survival and lower incidence of relapse and metastasis [14], a finding that is being extended to other tumour types. This immu- noscore classification distinguishes high- and low-risk patients in both initial and advanced stages of the disease, and it is independent and more informative than conventional staging methods.
The historical concept of immuno- surveillance has been more recently defined by studies in experimental animals that have identified major phases of tumour growth regulated by the anti-tumour immune response: elimination of nascent tumours, equi- librium with dormant tumours, edit- ing of progressing tumours, and fi- nally escape and uncontrolled growth [15]. A role for immunosurveillance in
humans is indicated by increased cancer incidence in immunosup- pressed patients. Such kidney trans- plant recipients or HIV-infected pa- tients appear to have a strikingly increased risk of tumour types of possible viral etiology. However, epidemiological evidence of an in- creased risk of other tumour types – such as colon and breast tumours, which have an increased incidence in immunodeficient mice – is less compelling. Immunodeficient mice have increased infections and mi- crobial dysbiosis, and this may com- pound the effect of lack of immunity on tumour susceptibility. Long-term clinical cancer dormancy is often ob- served – for example, in survivors of breast cancer and melanoma – and transplantation of organs from these patients into immunosuppressed re- cipients results in break of dorman- cy and donor tumour progression. However, the relative contribution of an immune equilibrium compared with the contribution of other ho- meostatic mechanisms in clinical
cancer dormancy remains to be fully defined [16].
Tumour immune escape Even in the presence of immuni- ty, clinical tumours progress and disseminate; spontaneous regres- sions are rare. Tumour growth can be attributed in part to the induc- tion of inefficient types of immune response, such as Th2 rather than Th1, and by an immunosuppres- sive tumour microenvironment. The activity of both effector T and B cells, as well as that of antigen- presenting cells, is depressed in the tumour microenvironment. Tumour immunosuppression is maintained by immunoregulatory cell types, such as regulatory T and B cells, as well as myeloid-derived suppressor cells. Immunosuppressive or anti- inflammatory molecules, such as interleukin-10 (IL-10), vascular en- dothelial growth factor, transform- ing growth factor β, IL-4, IL-13, nitric oxide, arginase, and indoleamine
Chapter 3.9 • Immunology and immunotherapy 257
3 B
IO LO
G Y
C H
A P
T E
R 3
.9
2,3-dioxygenase, are produced in the tumour microenvironment [12].
One of the major mechanisms of tumour escape is the loss or mu- tation of tumour antigens, as dem- onstrated experimentally by the immmunoediting of tumour cells in immunocompetent animals, as well as by decreased or lost expression of MHC molecules on the cell sur- face [15].
Immunity is modulated by a series of co-stimulatory recep- tor–ligand pairs expressed on an- tigen-presenting cells and T cells that amplify the T-cell response to antigen–MHC complexes [17]. The CD28/B7 family represents the best example of co-stimulatory molecules. Other members of this family are co-inhibitor molecules – such as, for example, CTLA-4 – that downregulate immune re- sponses when no longer needed. Tumours may escape the immune response by using these recep- tors and mechanisms involved in the physiological downregulation of the immune response. The pro- grammed cell death 1 (PD1) recep- tor and its ligand PD-L1 (also called B7-H1) are particularly relevant because PD-L1, unlike B7-1/B7-2,
is selectively upregulated on many human tumour cells (Fig. 3.9.3).
Tumour immunotherapy Since the early attempts of Coley to cure solid tumours with bacterial preparations (“Coley toxin”), investi- gators have attempted to harness the immune system to reject cancer by stimulating innate resistance or using adaptive immunity for cancer therapy. Various bacterial preparations have been used to induce inflammation and innate resistance, with some positive results, specifically including the use of bacillus Calmette–Guérin (BCG) for the topical treatment of bladder cancer [18]. Granted the role of innate receptors in the inflamma- tory response to bacterial products, use of Toll-like receptor (TLR) ligands in cancer therapy has been investi- gated; the TLR7 ligand imiquimod is approved for skin cancer, and the TLR4 ligand monophosphoryl lipid A as an adjuvant in a HPV vaccine.
Many pro-inflammatory cytokines have shown promise in experimental animals but have proven ineffective or highly toxic in clinical trials. Type I interferon and IL-2 are approved as single agents for cancer therapy. Interferon is approved for adjuvant
therapy of melanoma and for the treatment of certain haematopoietic malignancies, of Kaposi sarcoma, and, in combination with anti-angi- ogenic compounds, of renal cancer. With the development of targeted therapies, the use of interferon for melanoma has been declining. IL-2 is approved for advanced melanoma and renal cell carcinoma. Other im- munoregulatory cytokines that af- fect T-cell expansion and activation, such as IL-7, IL-12, and IL-15, are being clinically tested, alone or as part of immune-enhancing therapy. Granulocyte colony-stimulating factor and GM-CSF are approved to short- en the time to neutrophil recovery and reduce the incidence of infections af- ter cancer chemotherapy.
Immunotherapy based on the adoptive transfer of either naturally occurring tumour-infiltrating T cells or T cells engineered to express genes encoding well-defined anti- tumour T-cell receptors or designed chimeric receptors mediate tumour regression in a large proportion of patients with metastatic cancer, particularly when infused after lym- phodepletion [19]. Current efforts are focused on increasing the accuracy of targeting tumour antigens and the associated vasculature, as well as
Fig. 3.9.3. Timeline and milestones of the development of therapies involving anti-CTLA-4 and anti-PD-1, from gene discovery to clinical development.
258
on identification and purification of T-cell subsets that are more efficient in expanding and mediating tumour eradication.
Since the development of mono- clonal antibody technology, great effort has been devoted to using antibodies in cancer therapy [20]. Therapeutic antibodies can be di- rected against surface antigens se- lectively expressed on tumour cells. These antibodies induce tumour cell death by antibody-dependent cell- mediated cytotoxicity, recruiting effector cells such as natural killer cells and macrophages via binding of the Fc portion of the immuno- globulin G antibodies to Fc receptors
expressed on effector cells. The cytotoxic effect can be amplified by conjugating radioisotopes, cytotox- ic drugs, or toxins to the antibody. Alternatively, cell-mediated cytotox- icity can be facilitated by chimeric antibodies combining specificity for tumour antigens with specificity for activating receptors on immune cells, thus cross-linking tumour and effector cells while activating cyto- toxic mechanisms. Unconjugated anti-CD20 and anti-CD5 antibodies and radioisotope- or ozogamicin- conjugated anti-CD20 and anti- CD33 antibodies are approved for haematopoietic malignancies, and unconjugated anti-HER2 antibodies
for breast cancer. Another mecha- nism of action of antibodies in cancer is the functional inhibition of growth factor receptors or pro-tumour fac- tors. Anti-epithelial growth factor re- ceptor and anti-vascular endothelial growth factor unconjugated antibod- ies are approved for colorectal can- cer and other solid tumours.
Unlike the success of prophylac- tic vaccines against oncogenic vi- ruses, efficient therapeutic vaccines able to induce an immune response against established tumours have been difficult to develop. Although knowledge of the mechanisms of tumour immunity has been exten- sively applied in both experimental and clinical trials, clinical responses to tested cancer vaccines have been modest, and T-cell-mediated anti- tumour immunity elicited by vaccines was followed by clinical response in only a minority of patients [19]. Different formulations of antigens and adjuvants, including gene-based vaccination, have been used to har- ness the immunostimulatory ability of dendritic cells, either by adop- tively transferring them after in vitro activation and exposure to tumour antigens or by targeting them in vivo with tumour antigen-conjugated an- tibodies directed against their sur- face receptors [21]. Sipuleucel-T, the only approved vaccine for metastatic hormone-independent prostate cancer, has in clinical trials significantly but modestly increased survival by approximately 4 months [19]. Sipuleucel-T is prepared by ex- posing the patient’s dendritic cells to a fusion protein containing GM- CSF and the antigen prostatic acid phosphatase.
The poor efficacy of cancer vac- cines is probably due to persistence of an immunosuppressive tumour microenvironment, which prevents the development of the anti-tumour response [18]. Immunosuppression might be reversed by developing better adjuvants or by blocking im- munosuppressive factors, such as transforming growth factor β, in- doleamine 2,3-dioxygenase, or IL-10. Use of antibodies that block co-inhibitory molecules has
Fig. 3.9.4. Generation and regulation of anti-tumour immunity. Active cancer immuno- therapy may target antigen presentation at the level of dendritic cells, promote generation of anticancer T cells, and reverse the immunosuppressive mechanisms of the tumour microenvironment. The dendritic cells that have captured tumour antigens from dead or dying tumour cells (A) migrate to the draining lymph nodes, where they present captured antigenic peptides on major histocompatibility complex (MHC) class II and class I molecules to T cells (B), inducing an anticancer effector T-cell response if immunogenic maturation stimuli are present (C) or alternatively, in their absence, inducing tolerance and immunosuppression (D). Depending on the maturation stimulus, dendritic cells will differentially express co-stimulatory molecules and produce immunoregulatory cytokines, affecting the quality and the class of the T-cell response. Interaction of CD28 or OX40 with B7-1/B7-2 or OX40L favours protective T-cell responses, whereas interaction of CTLA-4 with B7-1/B7-2 or of PD-1 with PD-L1/PD-L2 will suppress T-cell responses. Effector T cells will re-enter the tumour microenvironment (E), where their anticancer ability will be limited by immunosuppressive mechanisms that include upregulation of PD-L1/PD-L2 on the cancer cell surface and release of IL-10, vascular endothelial growth factor, transforming growth factor β, IL-4, IL-13, nitric oxide, arginase, and indoleamine 2,3-dioxygenase. NK, natural killer cell; Treg, regulatory T cell.
Chapter 3.9 • Immunology and immunotherapy 259
3 B
IO LO
G Y
C H
A P
T E
R 3
.9
promise. The monoclonal antibody to CTLA-4 (ipilimumab) is approved for use in patients with metastatic melanoma; increased survival is reasonably attributable to amplifica- tion of the endogenous anti-tumour T-cell response (Fig. 3.9.4). In sev- eral clinical trials, antibodies against PD1 receptor or its ligand PD-L1 are showing significant responses in patients with melanoma or other solid tumours, such as lung cancer [17]. These new drugs may amplify the anti-tumour immune response elicited by cancer vaccines or the endogenous immune response in- duced after tumour destruction by certain therapies, resulting in clinical benefit. A downside of any therapy that increases anti-tumour immu- nity by reversing physiological im- mune checkpoints is that regulation of self-tolerance and antimicrobial resistance are also affected, with possible autoimmune and inflam- matory side-effects.
Combination of chemo- therapy, targeted therapy, and immunotherapy Cytotoxic agents have underpinned cancer therapy. Increasingly, agents that specifically target molecular pathways underlying malignant transformation have been developed (see Chapter 3.8). Both cytotoxic and targeted drugs may adversely affect the anticancer immune response. However, in experimental animals, certain drugs, including some widely used chemotherapy compounds, in- duce immunogenic cell death, which favours generation of a tumour-spe- cific immune response able to eradi- cate or delay tumour progression. Immunogenic cell death is charac- terized by expression of endogenous activators of inflammation – surface exposure of the endoplasmic reticu- lum protein calreticulin, and release of ATP and high-mobility group box 1 protein – resulting in effec- tive cross-presentation of tumour antigens and protective anticancer
immunity [22] (see “Premortem autophagy and endoplasmic reticu- lum stress as immunogenic signals in cancer therapy”). In addition, certain chemotherapeutic agents induce the intratumour expression of T-cell-attracting chemokines [23].
Combination of non-specific or targeted tumour therapies with immunotherapy approaches may provide successful cancer therapy. Although cytotoxic and targeted therapies are effective, clinical responses are often temporary. Experimental and clinical evidence suggests that immunotherapy, when successful, provides durable responses or even tumour eradica- tion [18]. Immunity may be able to destroy cancer stem cells that are unresponsive to standard or tar- geted therapies. While cytotoxic or targeted therapy may select tumour cells that became insensitive or had mutations in the targeted genes, strong anticancer immunity may de- stroy a small proportion of mutated
Premortem autophagy and endoplasmic reticulum stress as immunogenic signals in cancer therapy Guido Kroemer
The ultimate goal of anticancer therapy is the induction of tumour cell death. Physiological cell death, which occurs as a continuous by- product of cellular turnover, is non- immunogenic or even tolerogenic, thereby avoiding autoimmunity. However, cancer cell death elicited by radiotherapy and some chemo- therapeutic agents, such as an- thracyclines and oxaliplatin, can be immunogenic. Immunogenic death involves changes in the composition of the cell surface, as well as the release of soluble immunogenic sig- nals that occur in a defined temporal sequence. This “key” then operates on a series of receptors expressed by dendritic cells (the “lock”) to allow for the presentation of tumour anti- gens to T cells and for the initiation of a productive immune response. Immunogenic cell death is char- acterized by the early cell-surface
exposure of calreticulin, which deter- mines the uptake of tumour antigens by dendritic cells. The late release of the high-mobility group box 1 pro- tein, which acts on Toll-like recep- tor 4, is required for the presenta- tion of antigens from dying tumour cells. In addition, the release of ATP from dying cells causes the P2RX7 purinergic receptor-dependent acti- vation of the NLRP3 inflammasome in dendritic cells, thereby allowing them to release interleukin-1β and to polarize tumour antigen-specific CD8 T cells towards a Tc1 cytokine pattern.
It may be hypothesized that the immune system determines the long-term success of anticancer therapies and that this immune re- sponse is dictated by immunogenic tumour cell death. Thus, therapeu- tic failure can result from failure to undergo immunogenic cell death
(rather than cell death as such). Agents that fail to induce immu- nogenic cell death cannot yield a long-term success in cancer therapy. Moreover, tumours that are intrinsically unable to undergo immunogenic cell death are incur- able. Importantly, it appears that mitochondrial events determine whether cancer cells die in response to chemotherapy, while an endoplas- mic reticulum stress response com- bined with autophagy determines whether this cell death is perceived as immunogenic. Strategies may be developed to restore the immuno- genicity of cell death in the context of deficient autophagy or endoplasmic reticulum stress.
Reference Senovilla L et al. (2012). Science, 337: 1678–1684. http://dx.doi.org/10.1126/science. 1224922 PMID:23019653
260
cells present within the antigenic tumours. Anti-tumour immunity not only mediates direct cytotoxicity of antigenic tumour cells but also ac- tivates humoral and cellular inflam-
matory mechanisms that damage tumour cells as well as their microen- vironment, stroma, and vasculature by cross-presentation of tumour an- tigen on stromal antigen-presenting
cells. Progress in understanding these mechanisms will define opti- mal clinical protocols that combine different therapeutic approaches with immunotherapy.
References 1. Trinchieri G (2012). Cancer and inflamma-
tion: an old intuition with rapidly evolving new concepts. Annu Rev Immunol, 30:677– 706. http://dx.doi.org/10.1146/annurev- immunol-020711-075008 PMID:22224761
2. Hanahan D, Weinberg RA (2011). Hallmarks of cancer: the next generation. Cell, 144:646–674. http://dx.doi.org/10.1016/j. cell.2011.02.013 PMID:21376230
3. Fidler IJ (2003). The pathogenesis of cancer metastasis: the ‘seed and soil’ hypothesis revisited. Nat Rev Cancer, 3:453– 458. http://dx.doi.org/10.1038/nrc 1098 PMID:12778135
4. de Martel C, Ferlay J, Franceschi S et al. (2012). Global burden of cancers at- tributable to infections in 2008: a review and synthetic analysis. Lancet Oncol, 13:607–615. http://dx.doi.org/10.1016/S14 70-2045(12)70137-7 PMID:22575588
5. Lowy DR, Schiller JT (2012). Reducing HPV-associated cancer globally. Cancer Prev Res (Phila), 5:18–23. http://dx.doi. org/10.1158/1940 - 6207.CAPR-11- 0542 PMID:22219162
6. Atherton JC, Blaser MJ (2009). Coadaptation of Helicobacter pylori and humans: ancient history, modern impli- cations. J Clin Invest, 119:2475–2487. h t t p: / / d x . d o i . o r g / 10 .117 2 /J C I 3 8 6 0 5 PMID:19729845
7. Mantovani A, Allavena P, Sica A, Balkwill F (2008). Cancer-related inflammation. Nature, 454:436– 444. http://dx.doi.org/ 10.1038/nature07205 PMID:18650914
8. Grivennikov SI, Greten FR, Karin M (2010). Immunity, inflammation, and cancer. Cell, 140:883–899. http://dx.doi.org/10.1016/j. cell.2010.01.025 PMID:20303878
9. Wang D, Dubois RN (2010). Eicosanoids and cancer. Nat Rev Cancer, 10:181– 193. http://dx.doi.org/10.1038/nrc2809 PMID:20168319
10. Rothwell PM (2013). Aspirin in preven- tion of sporadic colorectal cancer: cur- rent clinical evidence and overall balance of risks and benefits. Recent Results Cancer Res, 191:121–142. http://dx.doi. o r g / 1 0 .1 0 0 7/ 9 7 8 - 3 - 6 4 2 - 3 0 3 31 - 9 _ 7 PMID:22893203
11. Blankenstein T, Coulie PG, Gilboa E, Jaffee EM (2012). The determinants of tumour immunogenicity. Nat Rev Cancer, 12:307–313. http://dx.doi.org/10.1038/nrc 3246 PMID:22378190
12. Coussens LM, Zitvogel L, Palucka AK (2013). Neutralizing tumor-promoting chronic inflammation: a magic bullet? Science, 339:286–291. http://dx.doi.org/ 10.1126/science.1232227 PMID:23329041
13. Albert ML, Darnell RB (2004). Paraneoplastic neurological degenerations: keys to tumour immunity. Nat Rev Cancer, 4:36–44. http://dx.doi.org/10.1038/nrc1255 PMID:14708025
14. Fridman WH, Pagès F, Sautès-Fridman C, Galon J (2012). The immune contex- ture in human tumours: impact on clini- cal outcome. Nat Rev Cancer, 12:298– 306. http://dx.doi.org/10.1038/nrc3245 PMID:22419253
15. Schreiber RD, Old LJ, Smyth MJ (2011). Cancer immunoediting: integrating im- munity’s roles in cancer suppression and promotion. Science, 331:1565–1570. http://dx.doi.org/10.1126/science.1203486 PMID:21436444
16. Uhr JW, Pantel K (2011). Controversies in clinical cancer dormancy. Proc Natl Acad Sci U S A, 108:12396–12400. http:// dx.doi.org /10.1073 /pnas.110 6 61310 8 PMID:21746894
17. Pardoll DM (2012). Immunology beats can- cer: a blueprint for successful translation. Nat Immunol, 13:1129–1132. http://dx.doi. org/10.1038/ni.2392 PMID:23160205
18. Mellman I, Coukos G, Dranoff G (2011). Cancer immunotherapy comes of age. Nature, 480:480– 489. http://dx.doi.org/ 10.1038/nature10673 PMID:22193102
19. Restifo NP, Dudley ME, Rosenberg SA (2012). Adoptive immunotherapy for can- cer: harnessing the T cell response. Nat Rev Immunol, 12:269–281. http://dx.doi. org/10.1038/nri3191 PMID:22437939
20. Weiner LM, Surana R, Wang S (2010). Monoclonal antibodies: versatile plat- forms for cancer immunotherapy. Nat Rev Immunol, 10:317–327. http://dx.doi. org/10.1038/nri2744 PMID:20414205
21. Palucka K, Banchereau J (2012). Cancer immunotherapy via dendritic cells. Nat Rev Cancer, 12:265–277. http://dx.doi. org/10.1038/nrc3258 PMID:22437871
22. Zitvogel L, Kepp O, Kroemer G (2011). Immune parameters affecting the ef- ficacy of chemotherapeutic regimens. Nat Rev Clin Oncol, 8:151–160. http://dx. d o i .o r g /10 .10 3 8 /n r c l i n o n c . 2 010 . 2 2 3 PMID:21364688
23. Abastado JP (2012). The next challenge in cancer immunotherapy: controlling T-cell traffic to the tumor. Cancer Res, 72:2159– 2161. http://dx.doi.org/10.1158/0008-5472. CAN-11-3538 PMID:22549945
Summary This Perspective summarizes sets of epidemiological data on early childhood tumours favouring their etiology as a result of interactions between specific host-cell chromo- somal modifications and infectious events. This is discussed mainly for acute lymphoblastic leukaemia but also for neuroblastomas and brain tumours. In contrast to the major- ity of previous publications, this re- port underlines evidence in support of a prenatal infection, resulting in immune tolerance for the infecting agent. Episomally persisting single- stranded DNA viruses with a circular genome are discussed as potential candidates. Prenatal infections have been documented for torque teno (TT) viruses. They are widespread in all human populations, replicate in cells of bone marrow and the periph- eral blood, and respond to interferon treatment. Thus far, in part due to the type heterogeneity of this virus family, no follow-up studies are avail- able demonstrating induction of im- mune tolerance in cases of prenatal infections.
If we include the approximately 10% of gastric cancers linked to Epstein–Barr virus infections and
Harald zur Hausen has devoted
his scientific life to investigating
to what extent infectious agents
contribute to human cancer. He
was awarded the Nobel Prize in
Physiology or Medicine 2008 for
his research demonstrating that
human papillomavirus (HPV)
causes cervical cancer and for the
identification of the high-risk types
of HPV. His work has resulted in
improved screening and treatment
of cervical cancer and has paved the
way for the development of vaccines
against high-risk HPV. Dr zur Hausen
studied medicine at the Universities
of Bonn, Hamburg, and Düsseldorf.
He was a researcher and professor
at the University of Würzburg, the
University of Erlangen-Nürnberg,
and the University of Freiburg, and
he served as the scientific director of
the German Cancer Research Center
(Deutsches Krebsforschungszentrum)
from 1983 to 2003. Dr zur Hausen’s
contributions to the field of virology
have informed our understanding of
the connections between infections
and chronic diseases.
the increasing percentage of head and neck cancers linked to high- risk human papillomavirus (HPV) infections, then slightly more than 20% of the global cancer incidence can be linked to infectious events [1]. Many of the existing links were originally identified based on epide- miological observations. This poses an interesting question: can we identify malignant disorders where epidemiological data may point to an infectious origin, although no convincing supportive evidence has thus far been obtained by different approaches? Cancers that could be considered from this perspective have been discussed before [2].
Particular childhood malignancies This Perspective will concentrate on a specific set of childhood malignan- cies: acute lymphoblastic leukaemia (ALL), neuroblastomas, and central nervous system (CNS) cancers, with some sidestepping to acute my- eloid leukaemia (AML) and Hodgkin disease. None of the malignancies discussed here has been consistently connected to infectious disorders, ex- cept for Hodgkin disease, where 25– 35% of cases in developed countries have been linked to Epstein–Barr virus infections. Since the major- ity of epidemiological studies have
Prenatal infections with subsequent immune tolerance could explain the epidemiology of common childhood cancers 261
Prenatal infections with subsequent immune tolerance could explain the epidemiology of common childhood cancers
Harald zur Hausen in collaboration with Ethel-Michele de Villiers
analysed ALL data, this cancer will be specifically emphasized.
At least two events contribute to ALL development, according to pre- vailing views: chromosomal modifica- tions acquired during the fetal period and an additional postnatal event, commonly discussed as an unspeci- fied infection [3–6]. Specific chromo- somal translocations, for example TEL-AML and t(11;19)(q23;p13), have been documented for the prenatal pe- riod [7] (reviewed in [8]). They occur, however, “at a substantially higher frequency (~100×) before birth than the cumulative incidence or risk of disease, reflecting the requirement for complementary and secondary genetic events that occur postnatally” [8], and they are thus not sufficient to cause leukaemia [9].
This Perspective summarizes sets of epidemiological data favour- ing the interaction of specific molecu- lar modifications with an infectious event. In contrast to the majority of previous publications, it underlines evidence in support of a prenatal in- fection, resulting in immune tolerance for the infecting agent.
Infections that mediate protection During the past decades, several re- ports have described a protective ef- fect of multiple infections during early childhood for ALL (reviewed in [10], [11]), neuroblastomas [12–14], CNS cancers [15], and Hodgkin disease [16,17]. Only a few studies have failed to confirm these observations for leu- kaemias (reviewed in [10], [11]). Some publications referred to a higher risk of ALL development after maternal in- fections during pregnancy (reviewed in [10]). Early immunization against Haemophilus influenzae type b seems to reduce the risk of ALL [18,19]. Even human leukocyte antigen (HLA) hap- lotype may influence ALL susceptibil- ity [20,21]. Interestingly, about 2% of childhood cancers are present at the time of delivery. Neuroblastomas are the most frequent connatal malig- nancy among the tumours discussed here, followed by CNS tumours and leukaemias (reviewed in [22]).
Attendance at day-care centres, number of siblings, and less breast- feeding during the first year of life have been implicated as surrogate markers for multiple infections and for the likelihood of childhood infection. Breastfeeding for more than 6 months has been repeatedly reported as a protective factor for ALL development (see reviews in [10], [11]). The protec- tive role of communicative contacts is well documented by the lower risk for subsequent siblings compared with the first-born child in a pooled analy- sis from five states in the USA [13]. In addition, allergies acquired during the first year of life have been identified as protective factors. Most studies to date have reported an inverse association between allergies and childhood leu- kaemia (reviewed in [23]).
Thus far, protective factors for ALL, and to a more limited degree for neuroblastomas and CNS tumours, have been discussed. Several risk factors can also be clearly identified; they concern, in particular, a high so- cioeconomic status and a protected environment. The rate of ALL is clear- ly higher in countries with high stan- dards of living than in poorer socie- ties [24]. As an overriding risk factor, however, were considered molecular changes, specifically chromosomal translocations, occurring during the intrauterine phase of life [4–8,25]. These observations soon led to the conclusion that ALL arises in utero already, although it has quickly been recognized that genetic modifications are not sufficient for leukaemia de- velopment [8,9]. It became a com- mon postulate that, as the regular course, first a genetic event occurs during the fetal period, followed by a second one, possibly an infection, during the postnatal phase.
Specific predictions If infections are involved in the eti- ology of those childhood cancers discussed here, the occurrence of connatal neuroblastomas, CNS tumours, and ALL would be an in- dication for prenatal (fetal) infec- tions with a high likelihood of inter- action with specific chromosomal
modifications. Prenatal infections with subsequent postnatal persis- tence will result in acceptance of the respective antigens by the im- mune system of the carrier, and thus in immune tolerance [26–28]. Anticipating prenatal infections and a resulting immune tolerance of the persisting agent, several predic- tions can be made: • The respective tumours should not
increase under conditions of immu- nosuppression (in organ allograft recipients and persons with HIV infection);
• The agent should retain its suscep- tibility to antiviral cytokines (e.g. interferons);
• Interferon-inducing nonspecific in- fections should negatively interfere with the suspected infection;
• Interferon-inducing allergic condi- tions should also negatively inter- fere with the suspected infection;
• Interferon that is present in breast milk should also have protective properties;
• Conditions that avoid multiple in- fections in early childhood should increase the risk for these tumours;
• The cancer risk for the first-born child (with a reduced risk for early childhood infections) should be higher than that for subsequent siblings (with a higher risk for com- municative infections);
• Prenatal chromosomal modifica- tions have been identified as risk factors but not as sufficient for can- cer induction; thus, there should be a syncarcinogenic interaction be- tween these modifications and the respective infection.
Immune tolerance has not yet been proven for antigens in the tu- mours discussed here. In case it exists for antigens of potentially on- cogenic agents, it is anticipated that immunosuppression should not fa- vour an enhanced outgrowth of the respective tumour type. However, antiviral cytokines (e.g. interferon- gamma), whose synthesis is not exclusively controlled by the lym- phatic system, should retain a broad function in antiviral defence [29]. Similarly, nonspecific infec- tions leading to interferon synthesis
262
should interact protectively. In this respect it seems to be interesting that enhanced interferon-gamma synthesis has been described for allergic conditions [30–33] and was also demonstrated in human breast milk for extended breastfeeding pe- riods [34–37].
The major risk factors are obvi- ously specific chromosomal trans- locations and gene modifications, which, as discussed before, clearly emerge as risk factors but are not sufficient to trigger malignant growth. The other major predisposing factor for the tumour types analysed here appears to be the absence of mul- tiple infections during the first year of life, lack of many communicative contacts, and an environment not exposing the infant to conditions of poor hygiene.
If we take prenatal infection and resulting immune tolerance for granted, all these data are readily compatible with a model we pro- posed in 2005 [38]. It is shown, with some minor modifications, in Fig. P3.1. If “Peri- and postnatal res- piratory infections” is replaced by “Interferon-gamma production due to allergic conditions or prolonged breastfeeding”, the basic features of the scheme remain unaltered.
Prenatal infections implicated Do prenatal infections exist that would fit into this picture? Obviously sev- eral infections have been identified as occurring during fetal life, among them cytomegalovirus and rubella infections, in part with grave conse- quences for the developing fetus. If we analyse, however, those infections that would fit into the criteria outlined here, very few of the identified agents would remain. Besides vertical transmission, additional properties should be subsequent development
of immune tolerance, replication or effective persistence within the pre- cursor cells of ALL and possibly the other cancers discussed here, and finally responsiveness to interferon. Polyomaviruses and anelloviruses (torque teno [TT] viruses) emerge as potential candidates [11,38].
Particularly anelloviruses seem to represent promising candidates. Vertical transmission of TT virus DNA has been demonstrated by detection in umbilical cord blood [39–41]. These viruses replicate in mononuclear cells of bone marrow and the peripheral blood [42–44] and, as shown by interferon treat- ment of hepatitis C virus carriers, they respond well to interferon treatment [45–47]. The question of whether these viruses induce immune tolerance af ter prena- tal infections remains open and is
currently difficult to study in view of their remarkable heterogeneity. Their analysis is being complicated by observations of frequent intrage- nomic rearrangements, resulting in novel open reading frames [48,49]. A synergistic function with specific cellular genetic modifications still requires further investigation.
It would be difficult, if not impos- sible, to reconcile a chain of mu- tational events with the published epidemiological data available con- cerning the etiology of the tumours discussed here. A prenatal infection with resulting immune tolerance, however, acting syncarcinogenical- ly with cellular gene modifications, seems to be in line with all epide- miological observations thus far re- ported. Virological studies, however, have not yet led to the identification of responsible infectious factors.
Fig. P3.1. The target cell conditioning model based on prenatal or perinatal torque teno (TT) virus-like infections. Transplacental as well as perinatal infections should lead to virus persistence and an increasing viral load. Multiple intermittent infections with interferon-inducing agents should reduce the viral load substantially and thus decrease the risk for virus-induced chromosomal modifications.
Prenatal infection
Peri- and postnatal
respiratory infections
A protected environment should result in a high load and high multiplicities of
infection, increasing the risk for infection of cells with specific chromosomal
modifications
Intermittent infections cause interferon induction
Thus, frequent infections should lower the load of a persisting agent and reduce the risk for infection of cells with specific
chromosomal translocations
Viral load
Time of delivery
Prenatal period
Clonal expansion
Time
Perinatal infection
263 Prenatal infections with subsequent immune tolerance could explain the epidemiology
of common childhood cancers
References
1. zur Hausen H (2006). Infections Causing Human Cancer. Weinheim, Germany: Wiley. http://dx.doi.org/10.1002/3527609318
2. zur Hausen H (2009). The search for infectious causes of human cancers: where and why. Virology, 392:1–10. http:// dx.doi.org /10.1016/ j.virol. 20 0 9.0 6.0 01 PMID:19720205
3. Greaves MF, Alexander FE (1993). An infectious etiology for common acute lymphoblastic leukemia in childhood? Leukemia, 7:349–360. PMID:8445941
4. Greaves M (1999). Molecular genetics, natural history and the demise of childhood leukaemia [Review]. Eur J Cancer, 35:1941–1953. http://dx.doi.org/10.1016/ S0959-8049(99)00296-8 PMID:10711237
5. Taub JW, Ge Y (2004). The prenatal origin of childhood acute lymphoblastic leukemia. Leuk Lymphoma, 45:19–25. http://dx.doi. o r g / 10 .10 8 0 / 10 4 2 819 0 310 0 0 14 9 4 0 3 PMID:15061193
6. Kim AS, Eastmond DA, Preston RJ (2006). Childhood acute lymphocytic leukemia and perspectives on risk assessment of early-life stage exposures. Mutat Res, 613:138 –160. http://dx.doi.org/10.1016/j. mrrev.2006.09.001 PMID:17049456
7. Mahmoud HH, Ridge SA, Behm FG et al. (1995). Intrauterine monoclonal origin of neonatal concordant acute lymphoblastic leukemia in monozygotic twins. Med Pediatr Oncol, 24:77–81. http:// d x .d o i .o r g /10 .10 0 2 /m p o. 2 9 5 0 24 0 2 0 3 PMID:7990767
8. Greaves M (2005). In utero origins of childhood leukaemia. Early Hum Dev, 81:123 –129. http://dx.doi.org/10.1016/j. earlhumdev.2004.10.004 PMID:15707724
9. Greaves MF, Wiemels J (2003). Origins of chromosome translocations in childhood leukaemia. Nat Rev Cancer, 3:639– 6 49. ht tp://dx.doi.org/10.1038/nrc116 4 PMID:12951583
10. O’Connor SM, Boneva RS (2007). Infectious etiologies of childhood leukemia: plausibility and challenges to proof. Environ Health Perspect, 115:146–150. http://dx.doi. org/10.1289/ehp.9024 PMID:17366835
11. zur Hausen H (2009). Childhood leukemias and other hematopoietic malignancies: interdependence between an infectious event and chromosomal modifications. Int J Cancer, 125:1764–1770. http://dx.doi. org/10.1002/ijc.24365 PMID:19330827
12. Menegaux F, Olshan AF, Neglia JP et al. (2004). Day care, childhood infections, and risk of neuroblastoma. Am J Epidemiol, 159:843–851. http://dx.doi.org/10.1093/aje/ kwh111 PMID:15105177
13. Von Behren J, Spector LG, Mueller BA et al. (2011). Birth order and risk of childhood cancer: a pooled analysis from five US States. Int J Cancer, 128:2709– 2716. http://dx.doi.org/10.1002/ijc.25593 PMID:20715170
14. Maule MM, Zuccolo L, Magnani C et al. (2006). Bayesian methods for early detection of changes in childhood cancer incidence: trends for acute lymphoblastic leukaemia are consistent with an infectious aetiology. Eur J Cancer, 42:78–83. http:// dx.doi.org /10.1016/ j.ejc a. 20 0 5.07.028 PMID:16324832
15. Altieri A, Castro F, Bermejo JL, Hemminki K (2006). Association between number of siblings and nervous system tumors suggests an infectious etiology. Neurology, 67:1979–1983. http://dx.doi. org/10.1212/01.wnl.0000247036.98444.38 PMID:17159104
16. Paffenbarger RS Jr, Wing AL, Hyde RT (1977). Characteristics in youth indicative of adult-onset Hodgkin’s disease. J Natl Cancer Inst, 58:1489–1491. PMID:857036
17. Roman E, Ansell P, Bull D (1997). Leukaemia and non-Hodgkin’s lymphoma in children and young adults: are prenatal and neonatal factors important determinants of disease? Br J Cancer, 76:406 – 415. http://dx.doi. org/10.1038/bjc.1997.399 PMID:9252212
18. Groves FD, Gridley G, Wacholder S et al. (1999). Infant vaccinations and risk of childhood acute lymphoblastic leukaemia in the USA. Br J Cancer, 81:175–178. http://dx.doi.org/10.1038/sj.bjc.6690668 PMID:10487630
19. Groves F, Auvinen A, Hakulinen T (2000). Haemophilus influenzae type b vaccination and risk of childhood leukemia in a vaccine trial in Finland [Abstract]. Ann Epidemiol, 10:474. http://dx.doi.org/10.1016/S1047- 2797(00)00110-1 PMID:11018411
20. Dorak MT, Oguz FS, Yalman N et al. (2002). A male-specific increase in the HLA-DRB4 (DR53) frequency in high-risk and relapsed childhood ALL. Leuk Res, 26:651– 656. ht tp://dx.doi.org/10.1016/ S0145-2126(01)00189-8 PMID:12008082
21. Taylor GM, Dearden S, Payne N et al. (1998). Evidence that an HLA-DQA1– DQB1 haplotype influences susceptibility to childhood common acute lymphoblastic leukaemia in boys provides further support for an infection-related aetiology. Br J Cancer, 78:561–565. http://dx.doi. org/10.1038/bjc.1998.540 PMID:9744491
22. Moore SW, Satgé D, Sasco AJ et al. (2003). The epidemiology of neonatal tumours. Report of an international working group. Pediatr Surg Int, 19:509–519. http:// dx.doi.org/10.1007/s00383 - 003 -1048 - 8 PMID:14523568
23. Chang JS, Wiemels JL, Buffler PA (2009). Allergies and childhood leukemia. Blood Cells Mol Dis, 42:99–104. http:// dx.doi.org/10.1016/j.bcmd.2008.10.003 PMID:19049852
24. Pisani P, Parkin DM, Bray F, Ferlay J (1999). Estimates of the worldwide mortality from 25 cancers in 1990. Int J Cancer, 8 3 :18 – 2 9 . h t t p: //d x . d o i . o r g /10 .10 0 2 / (SICI)1097-0215(19990924)83:1<18::AID- IJC5>3.0.CO;2-M PMID:10449602
25. Gruhn B, Taub JW, Ge Y et al. (2008). Prenatal origin of childhood acute lymphoblastic leukemia, association with birth weight and hyperdiploidy. Leukemia, 22:1692–1697. http://dx.doi.org/10.1038/ leu.2008.152 PMID:18548099
26. Jamieson BD, Ahmed R (1988). T-cell tolerance: exposure to virus in utero does not cause a permanent deletion of specific T cells. Proc Natl Acad Sci U S A, 85:2265–2268. http://dx.doi.org/10.1073/ pnas.85.7.2265 PMID:3258424
27. Dietert RR (2009). Developmental immunotoxicity (DIT), postnatal immune dysfunction and childhood leukemia [Review]. Blood Cells Mol Dis, 42:108–112. http://dx.doi.org/10.1016/j.bcmd.2008.10.005 PMID:19019708
28. Malhotra I, Dent A, Mungai P et al. (2009). Can prenatal malaria exposure produce an immune tolerant phenotype? A prospective birth cohort study in Kenya. PLoS Med, 6:e1000116. http://dx.doi.org/10.1371/journal. pmed.1000116 PMID:19636353
29. Hara T, Ohashi S, Yamashita Y et al. (1996). Human V delta 2+ gamma delta T-cell tolerance to foreign antigens of Toxoplasma gondii. Proc Natl Acad Sci U S A, 93:5136–5140. http://dx.doi.org/10.1073/ pnas.93.10.5136 PMID:8643541
30. Smart JM, Horak E, Kemp AS et al. (2002). Polyclonal and allergen-induced cytokine responses in adults with asthma: resolution of asthma is associated with normalization of IFN-gamma responses. J Allergy Clin Immunol, 110:450–456. http://dx.doi.org/10.1067/mai.2002.127283 PMID:12209093
31. Brown V, Warke TJ, Shields MD, Ennis M (2003). T cell cytokine profiles in childhood asthma. Thorax, 58:311–316. http://dx.doi.org/10.1136/thorax.58.4.311 PMID:12668793
32. Friedlander SL, Jackson DJ, Gangnon RE et al. (2005). Viral infections, cytokine dysregulation and the origins of childhood asthma and allergic diseases. Pediatr Infect Dis J, 24 Suppl:S170–S176, discussion S174–S175. PMID:16378042
264
33. Simon D, Braathen LR, Simon HU (2007). Increased lipopolysaccharide-induced tumour necrosis factor-alpha, interferon- gamma and interleukin-10 production in atopic dermatitis. Br J Dermatol, 157:583– 5 8 6. ht t p: //dx.doi.o r g /10.1111/ j.13 6 5 - 2133.2007.08050.x PMID:17596153
34. Daniels JL, Olshan AF, Pollock BH et al. (2002). Breast-feeding and neuro- blastoma, USA and Canada. Cancer Causes Control, 13:401– 405. http:// d x .d o i.o r g /10.10 2 3 /A :101574 67019 2 2 PMID:12146844
35. Kwan ML, Buffler PA, Abrams B, Kiley VA (2004). Breastfeeding and the risk of childhood leukemia: a meta-analysis. Public Health Rep, 119:521–535. http:// d x .d o i.o r g /10.1016 / j.p h r. 2 0 0 4. 0 9. 0 0 2 PMID:15504444
36. Martin RM, Gunnell D, Owen CG, Smith GD (2005). Breast-feeding and childhood cancer: a systematic review with metaanalysis. Int J Cancer, 117:1020– 1031. http://dx.doi.org/10.1002/ijc.21274 PMID:15986434
37. Tomicić S, Johansson G, Voor T et al. (2010). Breast milk cytokine and IgA composition differ in Estonian and Swedish mothers– relationship to microbial pressure and infant allergy. Pediatr Res, 68:330–334. http:// dx.doi.org/10.1203/PDR.0b013e3181ee049d PMID:20581738
38. zur Hausen H, de Villiers EM (2005). Virus target cell conditioning model to explain some epidemiologic characteristics of childhood leukemias and lymphomas. Int J Cancer, 115:1–5. http://dx.doi.org/10.1002/ ijc.20905 PMID:15688417
39. Gerner P, Oettinger R, Gerner W et al. (2000). Mother-to-infant transmission of TT virus: prevalence, extent and mechanism of vertical transmission. Pediatr Infect Dis J, 19:1074–1077. http://dx.doi. org/10.1097/00006454-200011000-00009 PMID:11099089
40. Goto K, Sugiyama K, Ando T et al. (2000). Detection rates of TT virus DNA in serum of umbilical cord blood, breast milk and saliva. Tohoku J Exp Med, 191:203–207. ht tp://dx.doi.org /10.1620/tjem.191. 203 PMID:11038012
41. Martínez-Guinó L, Kekarainen T, Segalés J (2009). Evidence of Torque teno virus (TTV) vertical transmission in swine. Theriogenology, 71:1390–1395. http://dx.doi. org/10.1016/j.theriogenology.2009.01.010 PMID:19249089
42. Okamoto H, Takahashi M, Nishizawa T et al. (2000). Replicative forms of TT virus DNA in bone marrow cells. Biochem Biophys Res Commun, 270:657–662. http://dx.doi.org/10.1006/bbrc.2000.2481 PMID:10753679
43. Mariscal LF, López-Alcorocho JM, Rodríguez- Iñigo E et al. (2002). TT virus replicates in stimulated but not in nonstimulated peripheral blood mononuclear cells. Virology, 301:121– 129. http://dx.doi.org/10.1006/viro.2002.1545 PMID:12359452
44. Kakkola L, Hedman K, Qiu J et al. (2009). Replication of and protein synthesis by TT viruses [Review]. Curr Top Microbiol Immunol, 331:53–64. http://dx.doi.org/10. 1007/978-3-540-70972-5_4 PMID:19230557
45. Maggi F, Pistello M, Vatteroni M et al. (2001). Dynamics of persistent TT virus infection, as determined in patients treated with alpha interferon for concomitant hepatitis C virus infection. J Virol, 75:11999–12004. http://dx.doi. org/10.1128/JVI.75.24.11999-12004.2001 PMID:11711590
46. Lai YC, Hu RT, Yang SS, Wu CH (2002). Coinfection of TT virus and response to interferon therapy in patients with chronic hepatitis B or C. World J Gastroenterol, 8:567–570. PMID:12046094
47. Moreno J, Moraleda G, Barcena R et al. (2004). Response of TT virus to IFN plus ribavirin treatment in patients with chronic hepatitis C. World J Gastroenterol, 10:143–146. PMID:14695786
48. de Villiers EM, Kimmel R, Leppik L, Gunst K (2009). Intragenomic rearrangement in TT viruses: a possible role in the pathogenesis of disease. Curr Top Microbiol Immunol, 331:91–107. http:// dx.doi.org/10.1007/978-3-540-70972-5_6 PMID:19230559
49. de Villiers EM, Borkosky SS, Kimmel R et al. (2011). The diversity of torque teno viruses: in vitro replication leads to the formation of additional replication- competent subviral molecules. J Virol, 85:7284–7295. http://dx.doi.org/10.1128/ JVI.02472-10 PMID:21593173
265 Prenatal infections with subsequent immune tolerance could explain the epidemiology
of common childhood cancers
This page intentionally left blank
Preventing cancer by avoiding exposure to particular carcinogens follows from knowledge of causation. Priorities for cancer prevention differ on national, and sometimes local, scales. Certain individual life- style choices are also identified as key to reducing cancer incidence. Tobacco-related mortality has prompted measures from supporting smoking cessa- tion to adopting an international treaty. In communi- ties where cancers are clearly attributable to alcohol drinking and/or avoidable sunlight exposure, corre- sponding behaviour change is warranted. Beyond exposure to specific carcinogens, various cancer types are attributable to the joint effect of diet, obe- sity, and physical inactivity, providing imperatives for prevention of diabetes, heart disease, and cancer, particularly in high-income countries. Primary pre- vention of some infection-caused cancers is achiev- able by vaccination, and regulatory measures have been proven to reduce or eliminate cancer caused by workplace or environmental pollutants. Secondary prevention by detecting premalignant or early stages of disease can reduce morbidity and mortality from particular tumours.
Cancer prevention
4
268
Summary • Tobacco use, particularly ciga-
rette smoking, remains a major cancer control priority as well as being a priority for other ar- eas of disease prevention.
• The WHO Framework Conven- tion on Tobacco Control is at the centre of international ef- forts to reduce tobacco-related harms.
• Uptake and continuation of to- bacco use by individuals is in- fluenced by personal and soci- etal/environmental factors. The practice of tobacco companies to actively market their prod- ucts is a major influence.
• Government-supported tobac- co control policies are neces- sary to constrain the activities of the tobacco industry, to deter tobacco use, and to encourage existing users to quit.
• A comprehensive tobacco con- trol programme requires poli- cies that adequately regulate the practices of tobacco com- panies, by eliminating their ca- pacity to promote their products; do all that is possible to reduce the harmfulness and allure of tobacco products; create a dis- incentive to use these products
through high taxes; protect peo- ple from involuntary smoking; ensure that the public is ad- equately informed; and support services to help those unable to quit unassisted.
Cigarette smoking is the most fre- quent form of tobacco use and a major cause of cancer as well as many other diseases, but stopping use can reverse some of the dam- age and prevent other risks from continuing to rise [1]. Other forms of tobacco are also harmful. Most of the harms to health are caused by the delivery of carcinogens and ad- ditional toxic substances, not by the nicotine, which is the main psycho- active substance in tobacco and the core reason people use and contin- ue to use tobacco. Tobacco use is also partly sustained by beliefs that users hold and by the roles that to- bacco use plays in society, both of which are influenced by tobacco in- dustry marketing. Quitting smoking can be facilitated by an environment where non-smoking has become the norm, but this is not sufficient for many smokers. Non-smokers need to be protected from involun- tary smoking. The importance of tobacco control to health and the need for coordinated international action has been recognized by the
international community through the WHO Framework Convention on Tobacco Control (FCTC) [2].
Prevention of initiation and pro- motion of cessation are the two main objectives guiding efforts to- wards reducing smoking. Along with protecting others from invol- untary smoking and reducing the
4.1 4 PREVENTION
Changing behaviours – tobacco control Ron Borland Maria E. Leon
Frank J. Chaloupka (reviewer) Hana Ross (reviewer) Melanie Wakefield (reviewer)
Fig. 4.1.1. Campaign poster from the French association Droits des Non- Fumeurs (Non-Smokers’ Rights): “The tobacco industry also has a social com- mitment. Every year, we solve the prob- lems of thousands of French pensioners. The tobacco industry kills 73 000 people every year. Keep trusting them.”
Chapter 4.1 • Changing behaviours – tobacco control 269
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.1
harmfulness of tobacco products, these form the four pillars of a com- prehensive approach to tobacco control. The ideal is to prevent up- take, but this is never likely to be universal, so other strategies are needed as well. This chapter focus- es on reducing cigarette smoking, although much of the discussion can be generalized to the control of other forms of tobacco use.
Tobacco use: the scope of the challenge The prevalence of cigarette smok- ing varies greatly across the world and also varies greatly by sex [3] but remains unacceptably high every- where. In several countries, more than 50% of men smoke cigarettes; these include China, Indonesia, and the Russian Federation, three of the largest countries. In some African countries, less than 10% of men smoke. Overall, far fewer women than men smoke. In some small countries, more than 40% of wom- en smoke, but the prevalence is less than 5% in many countries, in- cluding China. Use of other smoked products, notably bidis, is very high in India and other parts of South Asia. Use of a variety of smokeless tobacco products is most prevalent in South Asia and parts of Africa, where smokeless tobacco products are often the main form of tobacco use among women, and in parts of Scandinavia among men.
Tobacco use is addictive, due to the psychoactive effects of the nicotine. However, the addiction is more to the way in which the nico- tine is delivered than to the drug itself. Cigarettes are the most ad- dictive form of nicotine delivery, whereas nicotine provided in trans- dermal patches to facilitate cessa- tion appears to have no addiction potential. There is evidence that it is easier to quit smokeless to- bacco use than cigarette smoking [4]. Hence, the development and maintenance of addiction or depen- dence is a biopsychosocial phe- nomenon [5,6]. Rituals associated with tobacco use, sustained in part
by social conventions, become part of what is hard to give up. Tobacco dependence develops as the cumu- lative effect of tobacco use, usu- ally in adolescence, and becomes a barrier – physiological or/and psy- chological – to stopping use [7].
The majority of smokers start using tobacco as teenagers, with a younger age of initiation in high-in- come countries compared with mid- dle- and low-income countries [8]. The stages in the development of the habit in a tobacco user include never-use, trial, experimentation, and established use (with varying degrees of dependence), followed by an often-repeated cycle of quit attempts and relapse, leading to either permanent cessation or con- tinued use [9]. Preventing initial trial or progression of use after initial trial are the most important goals of tobacco control policies, as their achievement effectively eliminates the health risk, but while a popula- tion of smokers persists, additional efforts are required.
Tobacco control is grounded in the harms of the behaviour. Smoking behaviour is maintained by two sets of factors: the experi- enced effects of tobacco use, and the ways in which smoking has become socially useful. As smok- ing became more prevalent, ritu- als associated with it became part of the prevailing culture of coun- tries, as illustrated by the offering of cigarettes to others to facilitate socialization and the giving of cig- arettes as gifts. These factors not only help sustain use but also pro- vide reasons for use and reasons to oppose efforts to control use. The denormalization of smoking involves eliminating as far as possi- ble these social benefits of tobacco use, leaving only the experienced effects to be overcome for those motivated to quit. Denormalization takes time, and much of it occurs as a side-effect of strong public education campaigns and the in- troduction of smoke-free places. As smoking becomes less valued socially, policy-makers may more easily enact stronger laws to further
constrain use, thus creating a vir- tuous cycle of activity. Changing the social conditions or normative context to discourage tobacco use reduces the incentives to take up or continue smoking and should make quitting easier. However, part of the addiction is caused by factors within the individual; thus, norma- tive change, while helping to drive down the prevalence of smoking, is not enough to assist many smokers who remain unable to quit [10].
Measurement and evaluation Central to any systematic, scientifi- cally grounded approach to reduc- ing the tobacco epidemic is regular monitoring of smoking prevalence and high-quality evaluation of the impacts of control efforts. There are now coordinated international efforts to monitor use [11]. The key instruments are the Global Adult Tobacco Survey [12] and the Global Youth Tobacco Survey [13]. The Global Adult Tobacco Survey is a more recent innovation and has not extended to many countries yet. The use of the Global Youth Tobacco Survey is more widespread (Figure 4.1.2). There are similar pre-exist- ing surveys of both adults and chil- dren in most developed countries.
Trends indicated by survey data provide key indicators of progress. Surveys should be sufficiently large as to provide data about important subgroups, for example to indicate whether progress is equivalent across levels of socioeconomic sta- tus or by sex. Youth surveys are de- terminative in monitoring efforts to prevent uptake.
Volume 12 of the IARC Hand- books of Cancer Prevention, on methods for evaluating tobacco control policies, presents an exten- sive set of measures or variables, collected from multiple surveys, to monitor the natural history of tobacco use [14]. However, cross- sectional repeat surveys, while an excellent method for assessing prevalence and changes in preva- lence, lack specificity to reveal the mechanisms by which any observed changes have taken place. Cohort
270
ANTIGUA & BARBUDA
BARBADOS
ST LUCIA
TRINIDAD & TOBAGO
GUYANA
ST KITTS & NEVIS MONTSERRAT
ST VINCENT & THE GRENADINES DOMINICA
PUERTO RICO (associated)
BAHAMAS
CUBA
GRENADA
BR. VIRGIN IS. US. VIRGIN IS.
JAMAICA
DOMINICAN REP.HAITI
BOLIVARIAN REP.
VENEZUELA
TURKEY
RUS. FED.
CROATIA
REP. MOLDOVA
UKRAINE
FYR MACEDONIA
LITHUANIA LATVIA
ESTONIA
ALBANIA REP. MONT.
HUNGARY
BULGARIA
ROMANIA
GREECE
POLAND
SLOVENIA
B-H
BELARUS
RUSSIAN FED.
SLOVAKIA CZECH REP.
SERBIA KOSOVO
Tashkent
Havana
Shanghai
Macau SAR
Vientiane
Islamabad
Kabul
Abuja
Bangui
Kinshasa
Atlantic Lit. Addis Ababa
Constantine
Banjul
Monrovia
Western Area
Bissau
Maputo City
Lusaka
Arusha
Harare
Somaliland
Mexico City Port-au-Prince
La Paz
Rio de Janeiro
Bogotá
Tegucigalpa
Managua
Quito
Abidjan
Santiago
Baghdad
Gaza Strip
West Bank
REP. KOREA
TIMOR-LESTE
ZIMB.
UGANDA
UZBEKISTAN
TAJIKISTAN
KYRGYZSTANGEORGIA
ARMENIA
MADAGASCAR
MOZAMBIQUE
SRI LANKA
M A L A Y S I A
SAUDI ARABIA
C H I N A
M O N G O L I A
CAMBODIA
VIET NAM
LAO PDR
THAILAND
I N D I A
BHUTAN
BANGLADESH
ISLAMIC REP. IRAN
PAKISTAN
AFGHANISTAN
TURKEY
I R A Q
CYPRUS SYRIAN ARAB REP. LEBANON
JORDAN
BAHRAIN QATAR
UAE
OMAN
KUWAIT
YEMEN
NEPAL
K A Z A K H S T A N
ALGERIA
CHAD SUDAN
NIGERIA
NIGER
CENTRAL AFRICAN REP.
DEM. REP. CONGO
ETHIOPIA
EGYPT
MOROCCO LIBYAN ARAB
JAMAHIRIYA
MAURITANIA
SENEGAL
GAMBIA
GUINEA-BISSAU
CAPE VERDE
LIBERIA
CÔTE D’IVOIRE
BURKINA FASO
G H
A N
A
B EN
IN
CAMEROONEQUAT. GUINEA
CONGO
NAMIBIA BOTSWANA
SOUTH AFRICA
MALAWI
ZAMBIA
UNITED REP. OF TANZANIA
BURUNDI
RWANDA
KENYA
DJIBOUTI
SOMALIA
TO G
O
SIERRA LEONE
SWAZILAND
MALI
LESOTHO
ERITREA GUATEMALA
EL SALVADOR
MEXICO
BOLIVIA
B R A Z I L
U S A
COLOMBIA
HOND.
NIC.
COSTA RICA PANAMA
ECUADOR
PERU
SURINAME
BELIZE
ARGENTINA
URUGUAY
PARAGUAY CHILE
PHILIPPINES
NEW ZEALAND
PAPUA NEW
GUINEA I N D O N E S I A
R U S S I A N F E D E R A T I O N
MAURITIUS
MALDIVES
COMOROS
SEYCHELLES
SINGAPORE
MYANMAR
TUNISIA
PALAU
NORTHERN MARIANA IS.
VANUATU
SAMOA
TUVALU
FIJI
FED. STATES MICRONESIA
GUAM
COOK ISLANDS
AMERICAN SAMOA
30.0% or more
20.0% – 29.9%
10.0% – 19.9%
fewer than 10.0%
no data
sub-national data site
CIGARETTE SMOKING Percentage of boys aged 13–15 years who currently smoke cigarettes National or sub- national data latest available 1999–2008
SUSCEPTIBILITY TO INITIATE CIGARETTE SMOKING
Percentage of boys who have never smoked but consider themselves
susceptible to starting 1999–2008 AFR EMR EUR SEAR WPR TOTAL
17%
20%
26%
AMR
23%
16% 15%
19%
No di�erence was observed in cigarette smoking prevalence between boys and girls in 58% of the countries. Boys had higher rates than girls in 39% of the countries. Girls had higher rates than boys in 3% of the countries.
ANTIGUA & BARBUDA
BARBADOS
ST LUCIA
TRINIDAD & TOBAGO
GUYANA
ST KITTS & NEVIS
ST VINCENT & THE GRENADINES DOMINICA
PUERTO RICO (associated)
BAHAMAS
CUBA
GRENADA
BR. VIRGIN IS. US. VIRGIN IS.
JAMAICA
DOMINICAN REP.HAITI
BOLIVARIAN REP.
VENEZUELA
MONTSERRAT
REP. KOREA
TIMOR-LESTE
ZIMB.
UGANDA
UZBEKISTAN
TAJIKISTAN
KYRGYZSTANGEORGIA
ARMENIA
MADAGASCAR
MOZAMBIQUE
SRI LANKA
M A L A Y S I A
SAUDI ARABIA
C H I N A
M O N G O L I A
CAMBODIA
VIET NAM
LAO PDR
THAILAND
I N D I A
BHUTAN
BANGLADESH
ISLAMIC REP. IRAN
PAKISTAN
AFGHANISTAN
I R A Q
CYPRUS SYRIAN ARAB REP. LEBANON
JORDAN
BAHRAIN QATAR
UAE
OMAN
KUWAIT
YEMEN
NEPAL
K A Z A K H S T A N
ALGERIA
NIGER SUDANCHAD
NIGERIA
CENTRAL AFRICAN REP.
DEM. REP. CONGO
ETHIOPIA
EGYPT
MOROCCO LIBYAN ARAB
JAMAHIRIYA
MAURITANIA
SENEGAL
GAMBIA
GUINEA-BISSAU
CAPE VERDE
LIBERIA
CÔTE D’IVOIRE
BURKINA FASO
G H
A N
A
B EN
IN
CAMEROONEQUAT. GUINEA
CONGO
NAMIBIA BOTSWANA
SOUTH AFRICA
MALAWI
ZAMBIA
UNITED REP. OF TANZANIA
BURUNDI
RWANDA
KENYA
DJIBOUTI
SOMALIA
TO G
O
SIERRA LEONE
SWAZILAND
MALI
LESOTHO
ERITREA GUATEMALA
EL SALVADOR
MEXICO
BOLIVIA
B R A Z I L
U S A
COLOMBIA
HOND.
NIC.
COSTA RICA PANAMA
ECUADOR
PERU
SURINAME
BELIZE
ARGENTINA
URUGUAY
PARAGUAY CHILE
PHILIPPINES
NEW ZEALAND
PAPUA NEW
GUINEA I N D O N E S I A
R U S S I A N F E D E R A T I O N
MAURITIUS
MALDIVES
COMOROS
SEYCHELLES
SINGAPORE
MYANMAR
TUNISIA
PALAU
TURKEY
Gaza Strip
West Bank
NORTHERN MARIANA IS.
VANUATU
SAMOA
TUVALU
FIJI
FED. STATES MICRONESIA
GUAM
COOK ISLANDS
AMERICAN SAMOA
Tashkent
Havana
Shanghai
Macau SAR
Vientiane
Islamabad
Kabul
Abuja
Bangui
Kinshasa
Atlantic Lit. Addis Ababa
Constantine
Banjul Bissau
Monrovia
Maputo City
Lusaka
Arusha
Harare
Somaliland
Mexico City Port-au-Prince
La Paz
Rio de Janeiro
Bogotá
Tegucigalpa
Managua
Quito
Abidjan
Santiago
Baghdad
Western Area
TURKEY
RUS. FED.
CROATIA
REP. MOLDOVA
UKRAINE
FYR MACEDONIA
LITHUANIA LATVIA
ESTONIA
ALBANIA REP. MONT.
HUNGARY
BULGARIA
ROMANIA
GREECE
POLAND
SLOVENIA
B-H
BELARUS
RUSSIAN FED.
SLOVAKIA CZECH REP.
SERBIA KOSOVO
Half of sites indicated
girls’ cigarette smoking was equal to
or greater than that of adult
females
30.0% or more
20.0% – 29.9%
10.0% – 19.9%
fewer than 10.0%
no data
CIGARETTE SMOKING Percentage of girls aged 13–15 years who currently smoke cigarettes National or sub-national data latest available 1999–2008
sub-national data site
SUSCEPTIBILITY TO INITIATE CIGARETTE SMOKING
Percentage of girls who have never smoked but consider themselves
susceptible to starting 1999–2008
14% 14%
33%
13% 13%
19%
AFR EMR EUR
25%
AMR SEAR WPR TOTAL
Fig. 4.1.2. Global Youth Tobacco Survey (GYTS): monitoring of cigarette smoking in (A) boys and (B) girls. Regions: AFR, Africa; AMR, the Americas; EMR, Eastern Mediterranean; EUR, Europe; SEAR, South-East Asia; WPR, Western Pacific.
A
B
Chapter 4.1 • Changing behaviours – tobacco control 271
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.1
or panel studies are much more powerful in this regard. A good ex- ample of high-quality evaluation is the International Tobacco Control Policy Evaluation project, which not only uses cohorts of smokers within countries but also attempts to measure all of the main theorized mediators of policy effects as well as indicators of policy reach and proximal impact. These data are generated using parallel surveys in different countries, which permit results in a country that implements a policy to be compared with those in control countries that do not, and also allow for exploration of pos- sible differences in policy impact because of cultural traditions or the presence of complementary poli- cies and the history of past efforts. These methods can also help to tease out interactions between mul- tiple policies. This kind of approach is realizing the intention of the WHO FCTC to be an evidence-grounded treaty, by generating valid scientific data that are used to inform policy.
Tobacco control interventions The two main domains and the key elements of tobacco control that are considered in this section are outlined in Fig. 4.1.3 [14]. The first is tobacco industry control, which is directed at regulating the mar- keting resources of the industry and targets the product, its price, its availability, and its packaging. The second is tobacco use control, which focuses more directly on us- ers or on individuals vulnerable to become users, and includes restric- tions on where people can smoke, educational interventions, and pro- vision of support for cessation. Both types of controls are shown in Table 4.1.1.
Price and taxation Price is a factor that theoretically acts to reduce consumption inde- pendently of whether people be- lieve that use is harmful. Imposition of taxes to increase the retail price of tobacco reduces the prevalence and frequency of tobacco use in
the population. Similarly, decreas- ing prices are associated with in- creased use. Taxation is considered the single most effective interven- tion to change tobacco behaviours [15,16], but this may be because its impact is easy to quantify as it typically occurs immediately after the price changes, whereas other strategies, for example educational interventions, have rather more dif- fuse and longer-term effects.
Because price increases re- duce consumption and smoking prevalence, ensuring that the price is as high as possible is a key to- bacco control strategy. A significant proportion of the retail price of a to- bacco product involves taxes, and a change in the tax rate is the main way that governments can influence the retail price. Although WHO rec- ommends that excise taxes should account for a minimum of 70% of the retail price of tobacco prod- ucts, this level has been achieved by only a few countries in the world [17,18] and sometimes only for
some tobacco products. Apart from taxes, other ways that governments can influence retail prices include banning gifts with cigarette pur- chases, price controls exemplified by minimum price laws, and strong enforcement of penalties for evad- ing tobacco taxes. These options necessitate registration of manu- facturers and monitoring of national production as well as efforts to min- imize smuggling [16]. There is now a Protocol to the WHO FCTC – the Protocol to Eliminate Illicit Trade in Tobacco Products – that, when im- plemented, should help internation- al coordination to reduce the illegal trade in tobacco products, which is currently a major problem in some countries [2].
The impact of tobacco taxes on prices can also be lessened by le- gal responses to the tax increase [16]. For instance, in anticipation of or in response to a tax increase, tobacco companies can limit price increases by reducing profit mar- gins, sometimes targeted at those
Fig. 4.1.3. Schematic overview of tobacco control interventions and how they relate to tobacco products, users, and potential users.
272
products most used by young peo- ple and/or lower-income smokers. Companies can also turn a sud- den tax increase into a series of gradual increases in the hope that consumers will not notice and thus the impact on consumption will be less. Where products are differen- tially taxed, consumers can shift to products that are taxed less and are thus cheaper.
The demand for tobacco prod- ucts is inversely related to price. The magnitude of this response is measured by the price elasticity of the demand for tobacco products,
usually expressed in terms of the percentage change in use that results from a 10% increase in price. The more price-elastic the demand, the more effective is a tobacco price increase in lowering use. The price elasticity of demand can be estimated by using aggre- gate data or with household-level or individual-level data, based on surveys. Individual-level data allow the effect of price on consumption to be assessed by sex, age group, socioeconomic status, and other attributes of the population, poten- tially revealing differences between
subgroups in sensitivity to price changes. Aggregate data do not allow the effects on prevalence to be separated, independent of con- sumption per user.
There is an inverse relation- ship between cigarette prices and cigarette consumption across all countries. Data from the USA and the United Kingdom, where most studies have been conducted, in- dicate that the magnitude of the price elasticity of cigarette demand is about –0.4 (range, –0.2 to –0.6); that is, for a 10% increase in price a 4% decrease in consumption is
Table 4.1.1. Tobacco control interventions
Intervention Aims Examples
Tobacco industry controls – targeted at producers
Product controls To reduce use, and to reduce harmfulness of products Rules preventing the establishment of new products (prohibition of marketing and sales of snus in the European Union, with the exception of Sweden); rules limiting the appeal of products (bans on the use of flavourings or imposing limits on constituents); rules restricting the levels of constituents and emissions (upper limits on tar, nicotine, and carbon monoxide)
Price controls To reduce use in current users, and to deter use in never-users and former users
Increase in excise taxes passed onto the retail price of tobacco products; price policies that include a minimum price floor to counteract industry’s marketing discounts
Promotion controls
To reduce use and appeal of products Bans on paid advertisement, sponsorship, and place- ment of products, including restrictions on packaging (generic packaging); bans on product price-reducing promotions
Availability controls
To reduce availability of products and discourage use in certain venues
Restrictions on the number and types of outlets; restrictions about to whom products can be sold (age limits and vending machines); ban on sales in bars
Packaging controls
To reduce cues to use, and to make the product less attractive
Rules on descriptors of package content; ban on sales of single cigarettes; establishment of a minimum pack size
Form and content of warnings
To discourage use and inform the public of adverse health effects
Inclusion of facts about the harms of tobacco use or the health benefits of quitting use; information about toxin levels
Provisions to detect compliance with tax directives
To limit tobacco tax avoidance and evasion Use of tax stamps and of systems for tracking and tracing tobacco products
Tobacco use controls – targeted at users or potential users
Rules about use To reduce use, and to protect non-users Smoke-free policies that protect non-smokers and also have effects on users (reduction in consumption)
Public communication campaigns
To disseminate to as many people as possible the messages of anti-tobacco campaigns
Mass media campaigns to promote programmes designed to inform people while minimizing the risk of awakening the desires of experimenting with tobacco
Programmes to disseminate the availability of tobacco cessation services
To increase the level of information about cessation resources in place to increase their use
Rules regulating cessation medications, availability of services, and existence of subsidies on medications and services (among these: quitlines, “quit to win” campaigns, brief advice from physicians, nurses, and/or other health-care providers)
Chapter 4.1 • Changing behaviours – tobacco control 273
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.1
observed [8]. Elasticity estimates from other high-income countries are more varied. Greater elastic- ity is often observed in low- and middle-income countries, with esti- mates between –0.2 and –0.8 [16]. In those environments with elas- ticity estimates that are very low, cigarettes are very cheap or their affordability has risen significantly over time (due to increased earn- ings), rendering tax increases less effective.
A smaller number of studies have assessed the effect of increas- es in cigarette prices on smoking cessation in adults. Higher taxes and prices reduce the duration of smoking, increase quit attempts, and increase the number of smok- ers who successfully quit [16,19]. About half of the impact of the price increase is on adult prevalence and about half is on the number of ciga- rettes consumed by adult smokers [16]. Limited evidence indicates
that all other tobacco products are similarly price-elastic. Some stud- ies have documented substitution among these products in response to increases in their relative prices, specifically between cigarettes and smokeless tobacco after increases in taxes on cigarettes, and between cigarettes and cigars after increas- es in the price of cigarettes [20].
Price elasticity for tobacco prod- ucts within countries also depends on income levels of potential con- sumers. Indeed, after considering potential confounders, as aggre- gate income increases the aggre- gate demand for tobacco typically increases. The policy implication is that the price of tobacco products must increase at least as fast as in- come growth to avoid tobacco prod- ucts becoming more affordable. Younger tobacco users are expect- ed to be more sensitive to increases in prices or taxes than adults be- cause they have less disposable income and a lower level of tobacco dependence [16]. Younger users are also more influenced by peer use and are more concerned with immediate costs than with health outcomes in the long term. These expectations have been confirmed. Both smoking prevalence and inten- sity among young people diminish in response to cigarette price increas- es. The magnitude of the overall elasticity in high-income coun- tries, where most studies has been conducted, ranges between –0.50 and –1.2 [16].
Restrictions on tobacco promotion Many of the techniques now ubiq- uitous in mass marketing were first used, or first perfected, for selling manufactured cigarettes [21,22]. Advertising and other forms of pro- motion can be a powerful force for increasing product use. Legislative bans or other agreements to elimi- nate tobacco advertising are thus a critical aspect of tobacco con- trol efforts (Article 13 of the WHO FCTC) (Table 4.1.2). Advertising in- volves far more than advertisements on television, on billboards, or in
In 1 week your sense of taste and smell improves
In 5 years your risk of a stroke has dramatically decreased
In 12 months your risk of heart disease has halved
In 12 weeks your lungs regain the ability to clean themselves
In 5 days most nicotine is out of your body
In 8 hours excess carbon monoxide is out of your blood
In 1 year a pack-a-day smoker will save over $4,000
In 3 months your lung function begins to improve
STOP SMOKING START REPAIRING
Advertisement
australia.gov.au/quitnow
Authorised by the Australian Government, Capital Hill, Canberra
GO TO THE APP STORE OR ANDROID MARKET NOW TO DOWNLOAD My QuitBuddy FOR FREE
Fig. 4.1.4. Campaign poster from the Australian Department of Health and Ageing describing the gradual positive effects that quitting tobacco has on the human body.
274
newspapers; it also includes posters and the display of the products at point of sale as well as the branding and associated designs on the pack- aging. Internet advertising should also be recognized, but its control may require international coopera- tion. A comprehensive policy also requires cigarettes to be kept from view – to eliminate the promotional aspect of large, prominent cigarette displays – and removing as much branding as possible from packag- ing, ultimately by plain packaging (see “Australia’s plain packaging of tobacco products”). Restrictions on advertising need to be reasonably extensive before effects on use can be detected, at least in part due to the capacity of the industry to shift promotion to channels that are still permitted [8].
Availability of products Tobacco remains one of the most readily available consumer prod- ucts on the market. Many countries have prohibited sales to minors. A few countries, like France, limit to- bacco sales to specialist retailers. In most countries, however, to- bacco is still available in places like convenience stores, which, in turn, are often frequented by children. The density of tobacco outlets is related to the prevalence of smok- ing in the area [23]. An increased focus on restricting where tobacco products can be sold is anticipated in future policy initiatives.
Product regulation Use of filters is the only viable way of cleaning up tobacco smoke. However, use of filters is severely
limited in its efficacy because filters either remove nicotine or create conditions where nicotine cannot be carried as effectively into the lungs, thereby diminishing consumer ap- peal. While cigarettes and other forms of smoked tobacco continue to be used, it is important to pro- gressively reduce the toxicity of these products as far as possible. It is possible to remove the bulk of toxins from smokeless forms of to- bacco, so this is a viable strategy for reducing the harm that smokeless tobacco causes. Low-toxin smoke- less tobacco causes far less harm than other forms of tobacco [24], but it does contain carcinogens [25]. These products also contain nico- tine, which reinforces future use, thus prolonging the duration of the habit and increasing the cumulative use, thereby increasing the likeli- hood of damage to health.
The main forms of product regu- lation have been to set limits on the amount of tar, nicotine, and, often, carbon monoxide produced by cigarettes as determined using a standard machine. This prompted companies to market certain ciga- rettes as “light” and “mild”, imply- ing that these products were less harmful, sometimes by suggesting that they were an alternative to quit- ting. However, cigarettes that actu- ally delivered low levels of nicotine were not used in the long term by individual consumers. The industry addressed this problem by using filter ventilation: small holes in the paper surrounding the filter, thereby permitting air to mix with the smoke. This allowed cigarettes that, al- though they were determined by a testing machine to be low in tar and nicotine yield, actually delivered similar amounts of nicotine and the accompanying tar to smokers as did regular cigarettes [26]. Smokers compensated for the dilution effect; they blocked some of the holes and took larger and deeper puffs [27]. The WHO FCTC mandates pro- hibiting the misleading use of the terms “mild” and “light”, and recom- mends removing misleading yield measures from packaging where
Table 4.1.2. Key Articles of the WHO Framework Convention on Tobacco Control
Article Description
Article 5.3 Designed to restrict the influence of the tobacco industry in decision-making related to implementation of the treaty.
Article 6 Price and tax measures to reduce the demand for tobacco. Guidelines recommend trying to raise taxes so that they account for at least 70% of the retail price.
Article 8 Protection from exposure to tobacco smoke.
Article 9 Regulation of the contents of tobacco products.
Article 10 Regulation of tobacco product disclosures.
Article 11 Packaging and labelling of tobacco products. Mandates health warnings covering at least 30% of both main surfaces of packages. Guidelines suggest graphic warnings covering at least 50% of the package and consideration of plain packaging.
Article 12 Education, communication, training, and public awareness. Requires public education and training of health professionals, and participation of civil society in such activities.
Article 13 Tobacco advertising, promotion, and sponsorship. Requires prohibiting this as far as possible within constitutional limits.
Article 14 Demand reduction measures concerning tobacco dependence and cessation. Mandates developing guidelines and programmes for delivery of effective aids and services to help smokers quit.
Article 15 Illicit trade in tobacco products. Expanded with first Protocol to the treaty, requiring a range of actions to control international trade so as to eliminate illicit trade.
Article 16 Sales to and by minors. Designed to stop sales to minors, and includes suggestions for bans on vending machines and other channels where restrictions on sales may be problematic.
Articles 20–22 Research, which includes surveillance, intervention development, and evaluation to ensure that the WHO Framework Convention on Tobacco Control is evidence-based and progress can be monitored.
Chapter 4.1 • Changing behaviours – tobacco control 275
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.1
Australia’s plain packaging of tobacco products Simon Chapman
On 1 December 2012, Australia became the first country to require that all tobacco products be sold in standardized “plain” packaging. The Tobacco Plain Packaging Act 2011 mandates an identical stan- dard for all brands for all aspects of packaging [1]. Parallel legislation mandated graphic health warnings, now the world’s largest at 75% of the pack’s front surface and 90% of the back surface [2]. All brands are now sold in the same dull brown colour; the only differentiation is the brand name and variant, print- ed in a standard font. All aspects of the new design reflect detailed research with smokers about the options that are the least attrac- tive and cause the most concerns about health [3].
The Australian government in- troduced the bill following advice from its Preventative Health Task Force, which argued that Australia’s tobacco advertising ban failed to address what had long been a ma- jor loophole: the role of the pack
as advertising. As with all other in- dustries, internal tobacco industry documents and the tobacco trade press discuss candidly the central role of packs in promoting the at- tractions of brands and smoking.
Plain packaging is anticipated to have less effect on older, heavily dependent smokers, who tend to be brand-loyal and less image-con- scious. However, without branding, future generations will grow up nev- er having seen category A carcino- gens packaged in attractive packs. Today’s 20-year-olds in Australia have never seen local tobacco ad- vertising, and youth smoking rates are at an all-time low. Plain packs are expected to turbocharge this trend.
A large-scale evaluation of the impact of the packaging com- menced well before it was im- plemented. Early feedback sug- gests that many smokers believe their cigarettes now taste worse than before, an effect predictable from market research into how
packaging powerfully conditions expectations and consumers’ ex- perience of products. Reports are also common that smokers feel that the large health graphic health warnings are disturbing and embarrassing.
The new law has already sur- vived a tobacco industry challenge in the Australian High Court, with only one judge of seven not sup- porting the government. The in- dustry campaigned vigorously against the policy and is expected to engage in heavy price discount- ing to try and show the world that the policy failed and should not be adopted elsewhere. The govern- ment can counteract this by raising the tobacco tax.
Australia’s historic plain ciga- rette packaging legislation is a weapons-grade public health pol- icy now causing major concern in the international industry because of its potential to have a domino effect on other countries’ tobacco control policies. No other product is subject to such complete con- trol, sending a powerful message to the population that tobacco is an exceptionally unhealthy product, deserving exceptional controls.
References 1. Australian Government, Department
of Health and Ageing (2012). Tobacco. Plain packaging of tobacco products. Available at http://www.health.gov.au/ inter net /main /publishing.nsf/c ontent / tobacco-plain.
2. Australian Government, Department of Health and Ageing (2012). Tobacco. Health warnings. Available at http://www. health.gov.au/internet/main/publishing. nsf/content/tobacco-warn.
3. Australian Government, Department of Health and Ageing (2012). Market Research Reports on tobacco plain packaging and graphic health warn- ings. Available at http://www.health.gov. au/internet /yourhealth/publishing.nsf/ Content/mr-plainpack#.UOpDL7ZhNT4.
Fig. B4.1.1. Health warning statements on tobacco packages have evolved significantly in Australia, from text-only health warnings (left) until the introduction of graphic health warnings on 1 March 2006, to the introduction of the new plain packaging (right) on 1 December 2012.
276
they had been mandated. These bans have had a limited impact [28]. Consumers experience these ciga- rettes as slightly less harsh and in- terpret this reduced averseness as reduced harmfulness [29]. The core problem of filter ventilation has not been addressed, and remains [27].
Reducing product toxicity is one aim of Articles 9 and 10 of the WHO FCTC (Table 4.1.2). Product regu- lation also needs to consider both the addictiveness and the related attractiveness of tobacco products that continue to be harmful. Several countries, including Canada and Brazil, now ban all or nearly all addi- tives to cigarettes that are designed to affect flavour. Products contain- ing high levels of toxins still domi- nate the market in many countries.
One recent product develop- ment that has evoked interest as a harm reduction option is electronic cigarettes (e-cigarettes), also re- ferred to as electronic nicotine de- livery systems (ENDS). This prod- uct is designed to deliver nicotine to the lungs without combustion. Several groups of ex-smokers use and promote e-cigarettes as low- harm alternatives to smoking. The safety and efficacy of this product as a cessation aid is under study. Like smokeless tobacco, the po- tential role of e-cigarettes in reduc- ing smoking is unclear [30], but
evidence is emerging that e-ciga- rettes will be a much easier product to switch to and that they may be effective cessation aids [31].
Smoke-free laws Smoke-free legislation is primarily designed to protect non-smokers. Comprehensive laws that are well complied with greatly reduce invol- untary smoking. Smoke-free poli- cies also contribute to a reduction in the level of daily cigarette consump- tion by smokers, to a drop in the prevalence of smoking, and to de- ter smoking initiation among young people [32,33]. Bans on workplace smoking reduce daily cigarette con- sumption by an average of 2–4 ciga- rettes. One large United States study found that after controlling for poten- tial confounders, the percentage of a state’s population that was covered by smoke-free air laws was correlat- ed with a decreased odds of suscep- tibility to smoking, current smoking, and established smoking in young people [33]. Comprehensive smoke- free policies create strong anti- smoking norms that lead to volun- tary interventions such as smoking bans in private homes, which reduce exposure of children to second- hand smoke, with resulting effects in reducing adult smoking. Regular efforts are needed to remind the population of the health benefits of
smoke-free environments, including mass anti-tobacco campaigns and dissemination of information about the availability of cessation services or medications.
Globally, exposure to second- hand smoke varies significantly across countries, and a large part of the world’s population continues to be exposed to the health hazards of involuntary smoking both at work and at home. For example, 63% of Chinese workers report being ex- posed to smoke in their workplace (see “Tobacco and China”). The proportion of adolescents living in homes where others smoke in their presence can be high even in high- income countries.
Educational efforts The knowledge that tobacco use is harmful is central to any efforts to control its use, although sometimes the pervasiveness of this knowl- edge makes demonstrating effects from specific interventions difficult. The two main ways of informing the public are through mass media, increasingly via the Internet, and by provision of warning material on tobacco packaging. For young people, school-based education is another source of information, al- though it has only modest effects on preventing uptake unless com- plemented with other strategies [34]. Warning labels on packaging can improve knowledge and gener- ate concerns about smoking, and thereby stimulate increased quitting [35], although their net effect on cessation is probably small. Large graphic warnings are the most ef- fective. The mass media are the main source of health information for most people and, even taking the Internet into account, remain the main source of unsolicited in- formation. The mass media convey new information and remind people about knowledge they already pos- sess but have forgotten or are ig- noring (see also Chapter 4.3). The more attention is paid to tobacco in the mass media, the more quitting- related activity occurs [36].
Fig. 4.1.5. Campaign material from the Vietnamese Ministry of Health: “Tobacco has ravaging effects on your body and your baby’s. Quit today. From 01/01/2010, make sure to strictly implement the provisions of the government smoking ban in public places and workplaces.”
Chapter 4.1 • Changing behaviours – tobacco control 277
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.1
Well-constructed advertisements about the harms of smoking raise the profile of the issue and make the harms more personally relevant, thus encouraging smokers to quit and others not to start or resume smoking [37,38]. Part of the preven- tive effect seems to stem from dis- gust associated with the unattractive and disfiguring aspects of disease linked to smoking, rather than from anxiety about future harms. Anti- smoking advertisements appear to have their effects by essentially the
same mechanisms as health warn- ings on packs. Television advertise- ments can convey a much more engaging message than compa- rable pack warnings, but smokers are exposed to them far less often. Although advertising is expensive, it can reach large numbers of people quite cheaply per person influenced, so it is very cost-effective.
The mass media are effective in reaching smokers in lower socio- economic groups, especially when the advertising is both designed to
be seen as relevant and shown in programmes that these groups tend to view, and is at least as effective in influencing these groups as in influencing smokers in more socio- economically advantaged groups. Smokers in less-advantaged socio- economic groups may also require more intensive assistance if they are to quit successfully. Having the issue of smoking prominent in the mass media can also motivate health professionals by making them more aware and also more
Tobacco and China Judith Mackay
The epidemic The Chinese government owns the world’s largest tobacco company. China is the largest producer and consumer of tobacco, with about 350 million smokers, mostly male. One third of all cigarettes smoked in the world are smoked in China. In 2010 it was estimated that 53% of the male population and 2.4% of the female population in China were smokers [1]. Tobacco current- ly causes 1.2 million deaths annu- ally in China (equal to 3000 deaths every day), the highest number for any one country – and this figure is predicted to rise to more than 3 million by 2030 [2]. More than 70% of the population are regularly exposed to second-hand smoke. Only recently have small numbers of smokers attempted to quit.
Economic costs Smoking puts great pressures on economic development. Between 2000 and 2008, the attributable costs of tobacco use in China quadrupled, from US$ 7.2 billion to US$ 28.9 billion, or 0.7% of the gross domestic product [3].
Tobacco control action In 1979, the Ministries of Health, Finance, and Agriculture jointly is- sued the Circular of Publicity on the Harm of Smoking and Control on Smoking. Since the 1980s, central
and local governments have passed and expanded laws for the control of smoking in public places, pro- tection of minors, and bans on to- bacco promotion, and have issued many circulars. In 1997, Beijing hosted the 10th World Conference on Tobacco or Health, opened by President Jiang Zemin, which brought considerable attention to the issue of smoking. China ratified the WHO Framework Convention on Tobacco Control (FCTC) in 2005 and was the second country to sign the Protocol to Eliminate Illicit Trade in Tobacco Products in 2013. China is one of only four countries where tobacco control is included in both the National Development Plan and the United Nations Development Assistance Frameworks.
Obstacles and challenges The obstacles and challenges are common to most countries, and in- clude a focus on curative, not pre- ventive, medicine. The state tobac- co monopoly has jurisdiction over most tobacco control in China – a clear conflict of interest. The pub- lic still lacks accurate awareness of the harms of smoking and second- hand smoke. There is no national ban on smoking in public places and the workplace, where smok- ing is still very common. City-level smoke-free initiatives are neither
comprehensive nor well-enforced. China has weak text-only package warnings and no graphic warnings. Very cheap brands discourage smokers from attempting to quit. Despite some bans on direct tobac- co advertising, there is heavy expo- sure to tobacco promotion through event sponsorships, outdoor dis- plays, and entertainment media [2].
The future As part of comprehensive tobacco control action, the responsibility for tobacco control legislation and the WHO FCTC should be removed from the dominion of the state to- bacco monopoly, coupled with an effective taxation policy.
References 1. Centers for Disease Control and
Prevention (2010). Global Adult Tobacco Survey (GATS) Fact Sheet China: 2010. Available at http://apps.nccd.cdc. gov/gt s sdat a /A nc illar y/ Dat aRepor t s. aspx?CAID=1.
2. International Tobacco Control Policy Evaluation Project (2012). ITC China Project Report. Findings from the Wave 1 to 3 Surveys (2006 –2009). Waterloo, Ontario, Canada: University of Waterloo; Beijing, China: Office of Tobacco Control, Chinese Center for Disease Control and Prevention. Available at http://www. itcproject.org/documents/key findings/ itcchinanrenglishwebdec142012finalpdf.
3. Yang L, Sung H-Y, Mao Z et al. (2011). Tob Control, 20:266–272. http://dx.doi.org/10. 1136/tc.2010.042028 PMID:21339491
278
likely to raise the issue with their clients. The new social media, which includes the Internet and mo- bile telephony, provides both huge challenges and opportunities for tobacco control. The challenges come from the difficulty of regulat- ing any pro-smoking activity. The opportunities arise from the poten- tial of social media to provide more genuine personal engagement and the opportunity for well-informed members of the community to be- come active in communicating the messages.
Cessation aids There is now a range of effec- tive medications for smoking ces- sation when used for periods of about 6–12 weeks [39–41]. These medications, which include nico- tine itself (as nicotine replacement therapy) plus three other drugs (bu- propion, varenicline, and cytisine, which impinge on different aspects of the brain’s reward system), are all demonstrably effective. There is ev- idence that using medication for an extended period can at least delay relapse, but only for the period dur- ing which the drug is being used.
A wide range of behavioural interventions, including structured self-help manuals, tailored self- help resources, quitlines, and face- to-face individual or group pro- grammes, all help smokers to quit, and the success rates increase with increased involvement, up to sev- eral sessions [40,42]. The effective- ness of such interventions is direct- ly related to their intensity, typically measured by the amount of sup- port, at least up to about four sub- stantial sessions, but beyond that, there is little evidence that even more intensive interventions are ef- fective. Intensive advice-based pro- grammes are increasingly delivered via the telephone rather than face- to-face. Use of automated advice programmes, consisting of tailored personalized advice on the Internet and/or short, frequent advice mes- sages to mobile phones, is mark- edly increasing. Both forms of au- tomated intelligent programmes are effective. Combining advice-based programmes with medication gives the best results as the two seem to have largely independent effects [40]. There is no evidence of re- duced relapse associated with any intervention after it ceases [43].
Factors that influence smokers to try to quit are somewhat different to those that influence success in quitting [44]. Evidence is also begin- ning to emerge that determinants of short-term success may differ from those for long-term maintenance [10]. If so, different strategies are needed to help those who survive the difficult early weeks to maintain cessation in the long term, which is the ultimate challenge. Although this aspect is rarely subject to specific investigation, population- based studies indicate that a small percentage of ex-smokers continue to use nicotine replacement therapy in the long term. E-cigarettes cur- rently hold the most promise of be- ing viable substitutes for the largest number of smokers, but research is needed to document or confirm both safety and effectiveness as a cessation aid.
The outcome of interventions is not only a matter of how effective those interventions are when used but also a function of what propor- tion of smokers are prepared to use them in their quit attempts. As noted above, reference to smoking behaviour in the media drives peo- ple to seek help, as does product- specific advertising. To achieve the highest cessation rates requires a population of smokers who are motivated to quit, who are prepared to use the best possible help, and who are able to access aids that will maximize success. Uptake of help, particularly advice-based help, re- mains low, even when services are subsidized or free. This is probably due to a combination of beliefs that “I should be able to do this by my- self” and deep ambivalence about change in relation to tobacco use and other dependencies. Use of services is also influenced by the way the services are provided. For example, in the United Kingdom, there is much greater use of face- to-face services than there is of a well-organized and readily avail- able network of other services. However, in most other countries that have systems to provide help, telephone-based quitlines are gen- erally preferred to face-to-face services, at least in part due to the convenience.
Comprehensive efforts As recognized by the WHO FCTC, efforts on multiple fronts are re- quired if significant progress is to be made in tobacco control [2]. There are huge challenges. In low- and middle-income countries, govern- ments may have other priorities and sometimes lack the infrastructure to support policy. In high-income countries, smoking is increasingly concentrated among less-advan- taged socioeconomic groups, fur- ther exacerbating inequalities. Much remains to be done, but progress is being made.
Fig. 4.1.6. A WHO poster for World No Tobacco Day 2013 calls for legislative bans on tobacco advertising, promotion, and sponsorship.
Chapter 4.1 • Changing behaviours – tobacco control 279
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.1
References
1. IARC (2007). IARC Handbooks of Cancer Prevention, Vol. 11: Tobacco Control: Reversal of Risk After Quitting Smoking. Lyon: IARC.
2. WHO (2003). WHO Framework Convention on Tobacco Control. Geneva: WHO. Available at http://www.who.int/fctc/en/ index.html.
3. Eriksen M, Mackay J, Ross H (2012). The Tobacco Atlas, 4th ed. Atlanta, GA: American Cancer Society, World Lung Foundation. Available at http://www. tobaccoatlas.org.
4. Fagerström K, Eissenberg T (2012). Dependence on tobacco and nicotine products: a case for product-specific as- sessment. Nicotine Tob Res, 14:1382– 1390. http://dx.doi.org/10.1093/ntr/nts007 PMID:22459798
5. West R, Brown J (2013). Theory of Addiction, 2nd ed. Oxford: Wiley.
6. Henningfield JE, Benowitz NL (2010). Pharmacology of tobacco addiction. In: Boyle P, Gray N, Henningfield J et al., eds. Tobacco: Science, Policy, and Public Health, 2nd ed. Oxford: Oxford University Press, pp. 155–170.
7. DiFranza JR, Savageau JA, Fletcher K et al. (2002). Measuring the loss of autono- my over nicotine use in adolescents: the DANDY (Development and Assessment of Nicotine Dependence in Youths) study. Arch Pediatr Adolesc Med, 156:397–403. ht tp://dx.doi.org/10.10 01/archpedi.156. 4.397 PMID:11929376
8. Jha P, Chaloupka FJ, eds (1999). Curbing the Epidemic: Governments and the Economics of Tobacco Control. Washington, DC: World Bank.
9. Partos TR, Borland R, Yong HH et al. (2013). The quitting rollercoaster: how recent quitting history affects future cessation outcomes (data from the International Tobacco Control 4-country cohort study). Nicotine Tob Res, 15:1578– 1587. http://dx.doi.org/10.1093/ntr/ntt025 PMID:23493370
10. Borland R. (2014). Understanding Hard to Maintain Behaviour Change: A Dual Process Approach. Oxford: Wiley.
11. Global Tobacco Surveillance System Collaborating Group (2005). Global Tobacco Surveillance System (GTSS): purpose, production, and potential. J Sch Health, 75:15–24. http://dx.doi.org/ 10.1111/j.1746-1561.2005.tb00004.x PMID: 15779140
12. WHO (2012). Global Adult Tobacco Survey. Available at http://www.who.int/ tobacco/surveillance/gats/en.
13. WHO (2012). Global Youth Tobacco Survey. Available at http://www.who.int/ tobacco/surveillance/gyts/en/.
14. IARC (2008). IARC Handbooks of Cancer Prevention, Vol. 12: Tobacco Control: Methods for Evaluating Tobacco Control Policies. Lyon: IARC.
15. WHO (2008). WHO Report on the Global Tobacco Epidemic, 2008: The MPOWER Package. Geneva: WHO.
16. IARC (2011). IARC Handbooks of Cancer Prevention, Vol. 14: Tobacco Control: Effectiveness of Tax and Price Policies for Tobacco Control. Lyon: IARC.
17. WHO (2013). WHO Report on the Global Tobacco Epidemic, 2013: Enforcing Bans on Tobacco Advertising, Promotion and Sponsorship. Geneva: WHO. Available at http://apps.who.int/iris/bitstream/10665/ 85380/1/9789241505871_eng.pdf.
18. WHO (2010). WHO Technical Manual on Tobacco Tax Administration. Geneva: WHO. Available at http://www.who.int/ tobacco/publications/tax_administration/ en/.
19. Franz GA (2008). Price effects on the smoking behaviour of adult age groups. Public Health, 122:1343–1348. http:// dx.doi.org/10.1016/j.puhe. 20 08.05.019 PMID:18951594
20. Delnevo CD, Hrywna M, Foulds J, Steinberg MB (2004). Cigar use before and after a cigarette excise tax increase in New Jersey. Addict Behav, 29:1799–1807. http://dx.doi.org/10.1016/j.addbeh.2004. 04.024 PMID:15530722
21. Brandt AM (2007). The Cigarette Century: The Rise, Fall, and Deadly Persistence of the Product That Defined America. New York: Basic Books.
22. Proctor RN (2012). Golden Holocaust: Origins of the Cigarette Catastrophe and the Case for Abolition. Berkeley, CA: University of California Press.
23. Chuang Y-C, Cubbin C, Ahn D, Winkleby MA (2005). Effects of neighbourhood socioeconomic status and convenience store concentration on individual level smoking. J Epidemiol Community Health, 59:568–573. http://dx.doi.org/10.1136/jech. 2004.029041 PMID:15965140
24. Royal College of Physicians (2007). Harm Reduction in Nicotine Addiction: Helping People Who Can’t Quit. A report by the Tobacco Advisory Group of the Royal College of Physicians. London: Royal College of Physicians.
25. IARC (2012). Personal habits and indoor combustions. IARC Monogr Eval Carcinog Risks Hum, 100E:1–575. PMID:23193840
26. Kozlowski LT, Frecker RC, Khouw V, Pope MA (1980). The misuse of ‘less- hazardous’ cigarettes and its detection: hole-blocking of ventilated filters. Am J Public Health, 70:1202–1203. http:// d x . d o i . o r g /10 . 210 5 /A J P H .70 .11.12 0 2 PMID:7425194
27. Kozlowski LT, O’Connor RJ (2002). Cigarette filter ventilation is a defec- tive design because of misleading taste, bigger puffs, and blocked vents. Tob Control, 11 Suppl 1:I40–I50. http:// d x . d o i . o r g /10 .113 6 / t c .11. s u p p l _1. i 4 0 PMID:11893814
28. Yong HH, Borland R, Cummings KM et al. (2011). Impact of the removal of misleading terms on cigarette pack on smokers’ beliefs about ‘light/mild’ cigarettes: cross-country comparisons. Addiction, 106:2204–2213. http://dx.doi. org/10.1111/j.136 0 - 0 4 43. 2011.03533.x PMID:21658140
29. Shiffman S, Pillitteri JL, Burton SL et al. (2001). Smokers’ beliefs about “light” and “ultra light” cigarettes. Tob Control, 10 Suppl 1:i17–i23. PMID:11740040
30. Etter J-F, Bullen C, Flouris AD et al. (2011). Electronic nicotine delivery systems: a re- search agenda. Tob Control, 20:243–248. http://dx.doi.org/10.1136/tc.2010.042168 PMID:21415064
31. Bullen C, Howe C, Laugesen M et al. (2013). Electronic cigarettes for smok- ing cessation: a randomised controlled trial. Lancet, 382:1629–1637. http://dx.doi. o r g /10 .1016 / S 014 0 - 6 7 3 6 (13) 618 42- 5 PMID:24029165
32. IARC (2009). IARC Handbooks of Cancer Prevention, Vol. 13: Tobacco Control: Evaluating the Effectiveness of Smoke- free Policies. Lyon: IARC.
33. Farrelly MC, Loomis BR, Han B et al. (2013). A comprehensive examination of the influence of state tobacco control programs and policies on youth smoking. Am J Public Health, 103:549–555. http:// dx.doi.org/10.2105/AJPH.2012.300948 PMID:23327252
34. Pierce JP, Distefan JM, Hill D (2010). Adolescent smoking. In: Boyle P, Gray N, Henningfield J et al., eds. Tobacco: Science, Policy, and Public Health, 2nd ed. Oxford: Oxford University Press, pp. 313–322.
35. Borland R, Yong H-H, Wilson N et al. (2009). How reactions to cigarette packet health warnings influence quitting: find- ings from the ITC Four-Country survey. Addiction, 104:669–675. http://dx.doi.org/ 1 0 .1111 / j .13 6 0 - 0 4 4 3 . 2 0 0 9 . 0 2 5 0 8 . x PMID:19215595
280
36. Pierce JP, Gilpin EA (2001). News media coverage of smoking and health is asso- ciated with changes in population rates of smoking cessation but not initiation. Tob Control, 10:145–153. http://dx.doi. org/10.1136/tc.10.2.145 PMID:11387535
37. Durkin S, Brennan E, Wakefield M (2012). Mass media campaigns to promote smok- ing cessation among adults: an integra- tive review. Tob Control, 21:127–138. ht t p: //dx.doi.o rg /10.113 6 / tobac c o c on - trol-2011-050345 PMID:22345235
38. Wakefield MA, Loken B, Hornik RC (2010). Use of mass media campaigns to change health behaviour. Lancet, 376:1261–12 71. http://dx.doi.org/10.1016/S0140-6736 (10)60809-4 PMID:20933263
39. Cahill K, Stevens S, Perera R, Lancaster T (2013). Pharmacological interventions for smoking cessation: an overview and net- work meta-analysis. Cochrane Database Syst Rev, 5:CD009329. http://dx.doi. org/10.1002/14651858.CD009329.pub2 PMID:23728690
40. Treatobacco (2012). Available at www.treat obacco.net.
41. Hartmann-Boyce J, Stead LF, Cahill K, Lancaster T (2013). Efficacy of in- terventions to combat tobacco addic- tion: Cochrane update of 2012 reviews. Addiction, 108:1711–1721. http://dx.doi. org/10.1111/add.12291 PMID:23834141
42. The Cochrane Library (2012). Tobacco. Available at www.thecochranelibrary. com/view/0/index.html#http://w w w.thec ochranelibrary.com/view/0/browse.html.
43. Hajek P, Stead LF, West R et al. (2009). Relapse prevention interventions for smoking cessation. Cochrane Database Syst Rev, 8:CD003999. http://dx.doi.org/ 1 0 .1 0 0 2 / 14 6 518 5 8 . C D 0 0 3 9 9 9 . p u b 4 PMID:23963584
44. Vangeli E, Stapleton J, Smit ES et al. (2011). Predictors of attempts to stop smoking and their success in adult general population samples: a systematic review. Addiction, 106:2110–2121. http://dx.doi.org/ 10.1111/j.1360-0443.2011.03565.x PMID: 21752135
Websites Global Adult Tobacco Survey (GATS): http://www.who.int/tobacco/surveillance/ gats/en/
Global Youth Tobacco Survey (GYTS): http://www.who.int/tobacco/surveillance/ gyts/en/
Global Youth Tobacco Survey Results: www.cdc.gov/tobacco/global/gtss/tobacco_ atlas/pdfs/part3.pdf
The International Tobacco Control Policy Evaluation Project: http://www.itcproject.org
The Tobacco Atlas: http://www.tobaccoatlas.org
WHO Framework Convention on Tobacco Control: http://www.who.int/fctc/en/index.html
Chapter 4.2 • Changing behaviours – physical activity and weight control 281
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.2
Summary
• In trials, weight loss and in- creased physical activity have re- duced the risk of diabetes among participants.
• Behavioural weight manage- ment programmes typically produce weight losses of about 7–10% of initial body weight after 6–12 months of treatment, and increasingly include extended- care sessions.
• Typical goals in behavioural weight loss programmes in- clude reducing caloric intake by approximately 500–1000 kcal (2000– 4000 kJ) per day and in- creasing participation in moder- ate-intensity physical activity to 250 minutes or more per week.
• Behavioural weight management programmes focus on using self- monitoring, problem-solving, and goal setting to enact changes to the antecedents and conse- quences of behaviour.
• Benefits of adhering to weight loss, physical activity, and relat- ed guidelines are being evaluat- ed in relation to reduced cancer incidence in relevant trials.
4.2 4 PREVENTION
Changing behaviours – physical activity and weight control Rena R. Wing Kathryn R. Middleton
Christine M. Friedenreich (reviewer) Isabelle Romieu (reviewer)
Earlier chapters in this Report (see Chapter 2.6 and Chapter 3.5) present strong observational evidence that obesity and physical inactivity may be related to the development of some forms of cancer (see “Energy restriction, age, and cancer risk”) and suggest plausible biological path- ways for this association. Indeed, epidemiological studies suggest that physical activity and weight loss can lower breast cancer risk and improve survival, evidence that researchers have suggested warrants further clini- cal trials investigating the impact of behavioural weight management on preventing cancer and improving sur- vival [1]. To develop such studies, it is important to understand the basic strategies that might be used to help individuals change their weight and/ or their physical activity, and the type of results that can be achieved.
Outcomes of behavioural weight loss programmes Behavioural programmes typically produce weight losses of about 7–10% of initial body weight after 6–12 months of treatment. These weight losses have been shown to have clinically significant health benefits. The strongest evidence comes from the Diabetes Prevention Program, a multicentre randomized controlled trial, which demonstrated that behavioural weight loss led to
significantly larger reductions in dia- betes incidence than did treatment with metformin [2]. In that study, more than 3000 overweight or obese in- dividuals with impaired glucose tol- erance were randomly assigned to receive lifestyle intervention, met- formin (a drug usually used to treat diabetes), or placebo. The goals of the lifestyle intervention included weight loss of at least 7% of initial body weight (by decreasing dietary fat intake, reducing total caloric in- take, and increasing physical activ- ity) and increasing participation in moderate-intensity physical activity to 150 minutes per week. To reach these goals, a 16-session core cur- riculum was delivered to participants individually over the first 6 months of the study, with continuing contact (at least once every 2 months) during the remainder of the trial. The Diabetes Prevention Program trial was stopped early (after an average of 3.2 years of follow-up) because of the positive effects of lifestyle intervention on dia- betes incidence.
The lifestyle intervention resulted in an average weight loss of 6.5 kg at the end of the core curriculum and a weight loss of 4.5 kg at 3 years. Of participants in this arm, 49% achieved the 7% weight loss goal at 6 months and 37% at the final visit. On average, physical activity increased to 224 minutes per week
282
at the end of the core curriculum and 227 minutes at the end of the trial. The exercise goal was achieved by 74% of participants at week 24 and 67% at the final visit.
Although they are relatively mod- est, these weight losses and increas- es in physical activity produced dra- matic effects on diabetes incidence. The lifestyle intervention reduced the risk of developing diabetes by 58% relative to placebo. Weight loss was the dominant predictor of the reduced incidence of diabetes. Moreover, when participants were followed up for 10 years, the benefits of lifestyle intervention for preventing diabetes were still apparent, although there were no longer any differences in weight loss between the three arms. As diabetes has been demonstrated to be an independent predictor of cancer mortality (beyond body mass index) [3], the impact of intervention on diabetes incidence suggests that these trials may be beneficial for low- ering cancer mortality.
The findings from the Diabetes Prevention Program, indicating that modest weight losses and increases in physical activity can be achieved and can have important health ben- efits, have been confirmed recently in another clinical trial, called Look AHEAD. The Look AHEAD study
involved more than 5000 overweight or obese individuals with type 2 diabetes, who were assigned to an intensive lifestyle intervention or the control group [4]. The lifestyle inter- vention was similar to that used in the Diabetes Prevention Program, but participants were given a higher physical activity goal of 175 minutes per week and encouraged to lose 10% of their initial body weight, and meal replacement products were pro- vided to increase adherence to the dietary prescription. The intervention was delivered in group format with periodic individual sessions; partici- pants were seen weekly for the first 6 months, seen 3 times per month for months 7–12, and then seen or contacted by phone or e-mail at least twice a month. Participants in the life- style intervention arm had lost an av- erage of 8.6% of their body weight at 1 year (vs 0.7% in the control group). At 4 years, participants in the inten- sive lifestyle intervention group had maintained on average a loss of 4.7% of their initial body weight, whereas the control group had maintained a loss of 1.1%. Fitness levels also improved markedly in the lifestyle intervention group relative to the control group at both 1 year and 4 years. Improvements in glycae- mic control and in several important
cardiovascular risk factors were greater in the intensive lifestyle in- tervention group than in the control group throughout the 4 years [5]. The long-term results of the Look AHEAD trial, which indicate many health ben- efits of intensive lifestyle intervention but a lack of benefit for cardiovascular morbidity and mortality, will be de- scribed in future publications.
Indications of possible cancer- related benefits of behavioural weight loss programmes have been evi- dent for more than a decade. Thus, a randomized trial of a physical- activity-based weight management programme in breast cancer survi- vors (n = 68) demonstrated larger weight losses in the treatment group compared with the control group (5.7 kg vs 0.2 kg) and favourable changes in levels of inflammatory cytokines for intervention partici- pants but not for control participants [6]. Furthermore, within participants assigned to the treatment group, par- ticipation in physical activity was sig- nificantly associated with favourable changes in interleukin-6 levels.
Theoretical premise of behavioural weight loss programmes Behaviour-based weight loss pro- grammes, such as those used in the Diabetes Prevention Program and Look AHEAD trials, are based primarily on social learning theory [7]. These programmes focus on helping individuals make long-term changes in both their eating and exercise behaviours to produce weight loss and maintenance. This approach assumes that providing in- formation about diet and activity may be important and helpful to individu- als but is not sufficient. Rather, it is important to help people understand their current eating and exercise behaviours and the variables that are influencing these behaviours (e.g. to appreciate the learned con- nection between going to the mov- ies and eating popcorn). Behavioural treatment programmes then focus on teaching participants strategies for changing the variables that are
Fig. 4.2.1. An 8-year-old girl from the Rayerbazar area of Dhaka, Bangladesh, breaks through the ribbon during a race. Positive experiences and attitudes to physical activity may lead to long-term benefits in relation to diabetes and cancer.
Chapter 4.2 • Changing behaviours – physical activity and weight control 283
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.2
leading to inappropriate eating and sedentary behaviour as well as ways to evaluate the effect of the changes on their eating and exercise habits and ultimately their body weight.
Format of behavioural weight loss programmes Typically, behavioural weight loss programmes are offered in group set- tings, with about 10–20 participants treated together. Groups are often closed-format; participants all start at the same time and remain in the same group. Meetings (typically
60 minutes long) are held weekly for 6 months, every 2 weeks for the next 6 months, and then monthly for the following 6 months. This extended schedule of treatment contact has been shown to improve long-term outcomes [8]. Groups are typically led by a multidisciplinary team of nutritionists, exercise physiologists, and behavioural therapists. The pro- tocols for the lifestyle intervention used in the Diabetes Prevention Program and Look AHEAD trials are available online.
Energy balance Change in body weight is influ- enced by the balance of caloric intake relative to caloric expendi- ture. Thus, behavioural interven- tions typically focus on lower- ing caloric intake – typically by approximately 500–1000 kcal (2000– 4000 kJ) per day – while increasing physical activity. The dietary and physical activity com- ponents of the energy balance are reviewed below, and recommenda- tions are offered.
Energy restriction, age, and cancer risk Piet van den Brandt
Caloric restriction (CR) – reduced energy intake without malnutrition – is the most robust nutritional inter- vention known to date that increases lifespan in many species, such as yeast, nematodes, fruit flies, and mammals [1]. In rodents exposed to young-onset CR (age, 1–3 months), CR inhibits spontaneous and in- duced tumours in various organs. Effects of middle-age-onset CR on cancer risk are less clear, although beneficial effects are seen on age- related processes. In mice, a CR of 15–53% below usual ad libitum intake resulted in a proportionate linear reduction of 20–62% in tu- mour incidence. Effects of CR seem larger on spontaneous tumours: a meta-analysis showed 55% fewer spontaneous mammary tumours, with no heterogeneity according to study characteristics such as age, duration and degree of CR, and nu- trient type.
Two 20-year-long randomized trials in nonhuman primates have evaluated the effects of long-term CR on mortality and cancer risk in rhesus monkeys. In the Wisconsin National Primate Research Center (WNPRC) trial, an adult-onset CR of 30% non-significantly reduced overall mortality; incidence of cancer (gastrointestinal adenocarcinoma) was reduced by 50% compared with apes fed ad libitum. The National
Institute on Aging (NIA) trial investi- gated moderate CR in young and old monkeys and found no differences in survival. Nevertheless, cancer incidence was significantly reduced in monkeys with young-onset CR compared with controls [2]. Several differences between these two trials limit the ability to draw conclusions, including less CR, a healthier control diet, and no ad libitum controls in the NIA trial.
Indications that energy restric- tion may be important for cancer in humans stem from observational research. Overweight, as an indica- tor of a positive energy balance, is positively related to risk of various cancers. With respect to CR, Dutch and Norwegian cohort studies of in- dividuals who experienced 50–70% reduced food rationing (for < 1 year) in early life during the Second World War showed subsequently decreased risk of colorectal cancer [3] and (less consistently) of breast cancer, which may be due partly to confounding and malnutrition.
Ecological evidence on CR without malnutrition comes from Okinawans, who consume 15% few- er calories than mainland Japanese and have markedly lower cancer mortality rates. One randomized controlled trial is currently under way on effects of short-term CR (20–25% reduction for 6–12 months)
in overweight subjects. Risk factors for coronary heart disease and cancer improved, but CR adherence was low in the longer-term group. However, applying CR to subjects of normal weight to achieve a low-nor- mal body mass index of < 21 kg/m2 (as in nonhuman studies) differs from applying CR to overweight or obese individuals when studying health effects [1].
Possible mechanisms underly- ing beneficial effects of CR involve metabolic adaptations to CR, effects of CR on tumorigenic processes, for example upregulated DNA repair and downregulated insulin/IGF-1/ mTOR pathways [1], and epigenetic influences [3]. In summary, CR has beneficial effects on longevity and cancer incidence in different species; effects seem clearer with young-onset CR. Effects in humans remain promising but uncertain due to the relatively few randomized con- trolled trials that have concluded.
References 1. Omodei D, Fontana L (2011). FEBS Lett,
585:1537–1542. http://dx.doi.org/10.1016/j. febslet.2011.03.015 PMID:21402069
2. Mattison JA et al. (2012). Nature, 489:318– 321. http://dx.doi.org/10.1038/nature11432 PMID:22932268
3. Hughes LAE et al. (2009). PLoS One, 4:e7951. http://dx.doi.org/10.1371/journal. pone.0007951 PMID:19956740
284
Diet Calorie goals given to participants typically vary by baseline body weight [9]. Individuals who weigh 200 pounds (90 kg) or less at base- line are given calorie goals of approx- imately 1200 kcal (5000 kJ) per day, whereas those who weigh more than 200 pounds (90 kg) are given goals of 1500–1800 kcal (6250–7500 kJ) per day. These goals should lead to ca- loric intake deficits of approximately 500–1000 kcal (2000–4000 kJ) per day, which are associated with a weight loss of 1–2 pounds (0.5–1 kg) per week. In addition, participants are recommended to decrease their fat intake, typically by consuming less than 25% of caloric intake from fat. Furthermore, participants are encour- aged to increase intake of fruits and vegetables while decreasing intake of foods that observational studies sug- gest may be associated with obesity, such as sugar-sweetened beverages and high-fat snack foods [10].
Physical activity Participants in behavioural weight loss programmes are encouraged not only to decrease their caloric in- take but also to increase their level of participation in physical activity. This is often approached in two ways: (i) by increasing participation in moderate-intensity physical activ- ity and (ii) by increasing overall life- style physical activity. The American College of Sports Medicine [11] recommends 150–200 minutes per week of moderate-intensity activity (such as brisk walking) for individuals trying to maintain weight, and more
than 250 minutes per week for those trying to lose weight. Participants are encouraged to work up to these goals by gradually adding 10 minutes per week above baseline. In addition, participants typically track unstruc- tured, lifestyle physical activity using pedometers or accelerometers and aim to gradually increase their activ- ity to reach a goal of 10 000 steps per day.
Key components of weight management programmes In addition to providing information about diet and physical activity, other key components of weight manage- ment programmes focus on improv- ing self-regulatory skills and changing the antecedents and consequences of dietary and physical activity behav- iours. These key components are dis- cussed below.
Goal setting Self-regulation is generally viewed as an internal process that involves goal setting, self-monitoring, and evalu- ation of success or failure of goal achievement. Setting clear goals for caloric intake and physical activ- ity provides structure and direction. Individuals are encouraged to set goals that are short-term (typically a weekly weight loss goal vs a “goal weight”), measurable (e.g. eating an apple as a snack on three days dur- ing the week vs “eating more fruit”), and attainable (e.g. setting a starting activity goal to walk for 15 minutes three days per week vs running for an hour every day).
Self-monitoring Self-monitoring, a key step in evaluat- ing progress towards goal achieve- ment, may involve keeping records of body weight, caloric intake (as indi- cated by food consumed), and physi- cal activity (by using a pedometer or accelerometer). Self-monitoring al- lows individuals to assess their prog- ress towards goals and to receive feedback on the adequacy of their goal-directed behaviours. Adherence to self-monitoring has been demon- strated to be significantly associated with success in both weight loss and long-term maintenance of weight loss [12].
Problem-solving skills Problem-solving is a process by which individuals can address bar- riers to behaviour change [13]. Typically, problem-solving includes five distinct steps (Box 4.2.1). An ex- ample of a problem-solving partici- pant worksheet is shown in Fig. 4.2.2. The five-step problem-solving model is generally viewed as an iterative process, and if the chosen solution does not adequately address the barrier, individuals are encouraged to cycle back to steps 3 and 4 to try alternative solutions.
Changing behavioural antecedents Behaviour is affected both by ante- cedents – events that happen before the given behaviour – and by conse- quences, which are both positive or negative events that occur after the behaviour (Fig. 4.2.3). For example, an individual’s choice of lunch can
Step 1. Positive problem orientation. The view that problems are a nor- mal part of behaviour change ver- sus evidence of overall failure.
Step 2. Problem definition. Individuals clearly describe the bar- rier to change in objective, concrete terms.
Step 3. Generation of alternatives. Individuals brainstorm potential so- lutions, seeking quantity of ideas over quality.
Step 4. Decision-making. Participants evaluate decisions to decide which would be the best so- lution to the given problem.
Step 5. Solution implementation and verification. Participants implement the chosen solution and evaluate (through self- monitoring) whether this solution works.
Box 4.2.1. Problem-solving skills training: the five-step problem-solving model.
Chapter 4.2 • Changing behaviours – physical activity and weight control 285
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.2
be affected by their environment, and specifically the availability of nearby restaurants, and by thoughts and feelings (such as cravings, or feeling stressed or upset). These anteced- ents can often be manipulated by participants to assist with health be- haviour change, as discussed below.
Stimulus control Environmental factors have been demonstrated to affect both eating behaviours and physical activity. For example, the easy availability of high-calorie, highly palatable “junk food” can increase caloric intake, and lack of recreation facilities, safe walking areas, or sidewalks can de- crease participation in physical ac- tivity. Individuals have some control over their environments and can often enact positive environmental change at home and work.
Changing cognitions Thoughts and feelings can also rep- resent the antecedents in Fig. 4.2.3. According to the cognitive behav- ioural model of behaviour, an individ- ual’s thoughts affect their feelings, which can then affect behaviour. Thus, a person having the thought “I’ll never be able to lose weight” is likely to feel upset, frustrated, or an- gry, which may lead to the behaviour
of overeating or avoiding planned exercise. The process of cogni- tive restructuring involves identify- ing maladaptive thoughts, labelling these thoughts, and replacing them with more rational thoughts. For ex- ample, the thought “I’ll never be able to lose weight” can be replaced with the thought “I may have had a chal- lenging week, but I’ve lost 15 pounds so far and can recover from this slip.”
Changing behavioural consequences In addition to changing behavioural antecedents, the consequences of behaviour can be modified to affect future behaviour. Individuals will be more likely to engage in a behaviour
if it is tied in with a reward versus a punishment. Thus, manipulating the consequences of behaviour can lead to positive behaviour change. In general, individuals are encouraged to use non-food rewards, such as stickers, positive notes, buying new clothing, and so on. Increasing social support for behaviour change and use of programmatic incentives have also been demonstrated to lead to positive behaviour change.
Social support A group format for lifestyle weight management interventions has been used to invoke social support for in- dividual behaviour change. Group cohesion has been shown to en- hance the effectiveness of weight management treatment, even for people who indicate before interven- tion that they would prefer individual treatment. Sharing success at be- haviour change with the group can elicit positive support, acting as a social reward for the individual.
Incentives The use of incentives in weight management stems from literature in behavioural economics, which has demonstrated that people tend to discount long-term rewards in favour of short-term gains. This applies to weight management in that long- term benefits of improved dietary and activity behaviours, including weight loss and improvements in metabolic risk factors, can be less motivat- ing than short-term benefits such as the pleasure of eating or being
Fig. 4.2.2. An example of a problem-solving worksheet for programme participants.
Fig. 4.2.3. The A-B-C model of behaviour.
286
sedentary. Incentives for behaviour change and weight loss can provide additional reinforcement to the con- sequence end of the model shown in Fig. 4.2.3. Monetary incentives or deposit contracts, where partici- pants deposit money that is returned only if they meet programme goals, have been shown to improve initial treatment results, but their long-term effects are still unclear [14].
Maintenance of weight loss Long-term maintenance of weight loss has remained a substantial challenge [15,16]. After an interven- tion ends, individuals tend to regain lost weight, typically regaining one third to one half within 1 year of the end of treatment and returning to baseline weight within 3–5 years after treatment [17]. Research has focused on identifying factors that improve the long-term maintenance of weight loss; to date, one of the most successful factors has been the provision of extended care. A recent review of the literature on maintenance of weight loss sug- gested that the provision of extended care leads to the maintenance of an additional 3.2 kg of weight loss over 17.6 months after the intervention
compared with the control group [8]. Researchers are currently working to develop additional approaches to improving long-term maintenance of weight loss.
In addition, individual factors that promote long-term mainte- nance have been investigated. Using a national registry of people who have been successful in losing at least 13.6 kg and maintaining this loss for at least 1 year, research- ers have found that people who are
successful at maintaining weight loss tend to continue to consume a low-calorie, low-fat diet, eat breakfast, regularly self-monitor body weight and food intake, and engage in high levels of physical activity [18].
Dissemination and novel interventions While behavioural weight manage- ment programmes have demonstrat- ed efficacy and effectiveness, dis- semination remains a barrier to wide access to treatment. Cost and avail- ability of trained staff often limit the reach of these programmes, espe- cially the high-intensity programmes most often delivered in research settings: typically 3–6 months of weekly groups, followed by several months of extended-care groups, meeting every 2 weeks and then monthly. Thus, increasing research has focused on lower-cost deliv- ery of these programmes, includ- ing stepped-care, phone-based, and Internet-based programmes. Interventions using newer self- monitoring technologies, including activity monitors and smartphone applications that allow individuals to track their weight, food intake, and activity, hold promise as access to computers and smartphones across the population has been increasing substantially.
Fig. 4.2.4. A display of fresh produce on a market stall in the Netherlands. Increasing intake of fresh fruits and vegetables and avoiding bringing high-fat snack foods into the home can lead to health benefits.
Fig. 4.2.5. The National Health Service in the United Kingdom has a website on healthy living, called Change4Life. The “Eat well” section of the website enables people to plan healthy meals.
Chapter 4.2 • Changing behaviours – physical activity and weight control 287
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.2
EPIC as a model study vehicle Elio Riboli
The European Prospective Inves- tigation into Cancer and Nutrition (EPIC) was developed at IARC, in close collaboration with several major national research institu- tions, as a long-term multicentre prospective cohort study in west- ern Europe designed specifically to investigate the relationship be- tween lifestyle, diet, and cancer. Enrolment of EPIC participants began in 1992, with the invitation of mainly healthy subjects aged 35 –70 years from the general pop- ulation in certain geographical ar- eas. By the end of the recruitment phase, in 2000, EPIC was the larg- est prospective cohort study with a baseline biorepository aimed at investigating cancer etiology, with 521 330 participants from 23 study centres in 10 European countries.
Participants completed ques- tionnaires on diet and lifestyle factors, including physical activ- ity, and underwent measurements of anthropometric characteristics (e.g. weight, height, waist and hip circumferences), blood pressure, and pulse rate. Anonymized data are stored in a central secured da- tabase at IARC. EPIC was the first prospective study to collect and
store blood from such a large num- ber of participants: 388 467 blood samples of 30 ml each were col- lected and aliquoted into 28 plastic straws stored in liquid nitrogen, with aliquots divided between the central biorepository at IARC and each na- tional centre. Since initial baseline collection, EPIC participants have been followed up to monitor changes in major dietary and lifestyle fac- tors. Follow-up aimed at identifying cancer cases and dates and causes of death occurring among the EPIC cohort is principally based on record linkage with data from population cancer registries and regional or national death registries [1].
The design of the EPIC co- hort is key to facilitating stud- ies of cancer etiology. It was the first prospective study with a be- spoke biorepository of its size in the world. Wide geographical coverage also means that EPIC includes populations with varying dietary and lifestyle habits and dif- ferent underlying cancer incidence rates; this strategy was conceived to increase the overall statistical power to identify diet–disease re- lationships. The range of data as well as biological samples that
were collected before diagnosis from the same population enables integrated analyses of cancer risk factors. Finally, through long-term follow-up spanning almost two decades, more than 60 000 newly incident cancer cases have been reported, giving scope for well- powered studies.
EPIC has proven to be a power- ful resource to investigate the asso- ciation between complex nutritional, metabolic, and genetic characteris- tics and cancer risk. An example is the identification using EPIC of links between metabolic syn- drome and cancer risk. Metabolic syndrome is a cluster of metabolic abnormalities, including abdominal obesity, elevated blood pressure, abnormal glucose metabolism, and dyslipidaemia. It was shown that individual components of meta- bolic syndrome, such as insulin resistance (as measured by serum C-peptide levels) and central obe- sity, are associated with increased risk of several types of cancer [2]. While metabolic syndrome was associated with colorectal cancer risk, this could be accounted for by individual components of the syn- drome, namely abnormal glucose
Fig. B4.2.1. Relative risk of premature death in men and women in EPIC in relation to waist circumference, adjusted for body mass index, age, smoking status, education level, alcohol consumption, physical activity, and height.
288
metabolism and central obesity, in- dicating that complex definitions of metabolic syndrome may not be of additional clinical utility in identify- ing individuals at increased risk of colorectal cancer.
Concerning central obesity, EPIC data allowed clarification of the apparent paradox by which being very slim (body mass index [BMI], 20–23 kg/m2) and being overweight (BMI, 26–30 kg/m2) appeared to be associated with
similarly higher risk of all-cause mortality in adult life, compared with having a medium BMI (about 24–25 kg/m2). This paradox was often described as a J-shaped risk function of BMI versus mortal- ity rates. EPIC showed that waist circumference adjusted for BMI is linearly related to the risk of pre- mature death, that having a slim waist is associated with the lowest mortality, and that the apparent ex- cess is due to individuals with low
BMI but with low muscular mass and excess abdominal adiposity (Fig. B4.2.1) [3].
References 1. Bingham S, Riboli E (2004). Nat Rev
Cancer, 4:206–215. http://dx.doi.org/10. 1038/nrc1298 PMID:14993902
2. Jenab M et al. (2007). Int J Cancer, 121:368–376. http://dx.doi.org/10.1002/ijc. 22697 PMID:17372899
3. Pischon T et al. (2008). N Engl J Med, 359:2105–2120. http://dx.doi.org/10.1056/ NEJMoa0801891 PMID:19005195
Given the evidence that ex- cess body weight and low levels of physical activity are associated with cancer risk and poorer survival outcomes (see “EPIC as a model study vehicle”), there is a need to conduct clinical trials to determine whether behavioural weight loss interventions can improve surviv- al and/or help to prevent cancer [1]. Behavioural weight loss pro- grammes are effective in helping participants lose 7–10% of their ini- tial body weight; these weight losses produce numerous health benefits and should be evaluated further for cancer prevention and survival. A current randomized controlled trial, including 962 patients in 40 centres, is investigating the impact of physi- cal activity on disease outcome in colon cancer survivors [19].
Data relating to cancer out- comes affected by adherence to weight control and/or physical activity guidelines are becoming available. Consequences of adher- ence to World Cancer Research Fund recommendations were moni- tored among approximately 30 000 postmenopausal women aged 50– 76 years at baseline in 2000–2002
Fig. 4.2.6. Paradoxically, sport shown on television may play a role in determining sedentary behaviour.
with no history of breast cancer. Meeting these cancer prevention recommendations, specifically those related to alcohol, body fatness, and intake of plant foods, was as- sociated with reduced breast cancer incidence [20]. Benefits to cancer survivors have also been observed [21]. Cancer prevention through
promotion of physical activity and dietary change in cancer screen- ing settings is also being explored [22]. Future trials will investigate the role of weight loss and/or in- creased physical activity in primary prevention of various tumour types.
Chapter 4.2 • Changing behaviours – physical activity and weight control 289
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.2
References 1. Ballard-Barbash R, Hunsberger S, Alciati
MH et al. (2009). Physical activity, weight control, and breast cancer risk and survival: clinical trial rationale and design consider- ations. J Natl Cancer Inst, 101:630–643. ht t p: //dx.doi.o rg /10.10 9 3 / jnc i /djp 0 6 8 PMID:19401543
2. Knowler WC, Barrett-Connor E, Fowler SE et al.; Diabetes Prevention Program Research Group (2002). Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med, 346:393–403. http://dx.doi.org/10.1056/NE JMoa012512 PMID:11832527
3. Coughlin SS, Calle EE, Teras LR et al. (2004). Diabetes mellitus as a predictor of cancer mortality in a large cohort of US adults. Am J Epidemiol, 159:1160–1167. ht tp: //dx.doi.org /10.10 93 /aje / k wh161 PMID:15191933
4. Wadden TA, Neiberg RH, Wing RR et al.; Look AHEAD Research Group (2011). Four-year weight losses in the Look AHEAD study: factors associated with long-term success. Obesity (Silver Spring), 19:1987–1998. http://dx.doi.org/10.1038/ oby.2011.230 PMID:21779086
5. Jakicic JM, Egan CM, Fabricatore AN et al. (2013). Four-year change in cardiores- piratory fitness and influence on glycemic control in adults with type 2 diabetes in a randomized trial: the Look AHEAD trial. Diabetes Care, 36:1297–1303. http://dx.doi. org/10.2337/dc12-0712 PMID:23223405
6. Pakiz B, Flatt SW, Bardwell WA et al. (2001). Effects of a weight loss intervention on body mass, fitness, and inflammatory biomarkers in overweight or obese breast cancer survi- vors. Int J Behav Med, 18:333–341. http:// dx.doi.org/10.1007/s12529-010-9079-8 PMID:21336679
7. Bandura A (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall.
8. Ross Middleton KM, Patidar SA, Perri MG (2012). The impact of extended care on the long-term maintenance of weight loss: a systematic review and meta-analysis. Obes Rev, 13:509–517. http://dx.doi.org/10.1111/ j.1467-789X.2011.00972.x PMID:22212682
9. Wing RR (2008). Behavioral approaches to the treatment of obesity. In: Bray GA, Bouchard C, eds. Handbook of Obesity: Clinical Applications, 3rd ed. New York: Informa Healthcare, pp. 227–248.
10. Mozaffarian D, Hao T, Rimm EB et al. (2011). Changes in diet and lifestyle and long-term weight gain in women and men. N Engl J Med, 364:2392–2404. http:// dx.do i.o r g /10.10 5 6 / N E J M oa10142 9 6 PMID:21696306
11. Donnelly JE, Blair SN, Jakicic JM et al.; American College of Sports Medicine (2009). American College of Sports Medicine Position Stand. Appropriate physical activity intervention strategies for weight loss and prevention of weight regain for adults. Med Sci Sports Exerc, 41:459–471. http://dx.doi.org/10.1249/MSS. 0b013e3181949333 PMID:19127177
12. Burke LE, Wang J, Sevick MA (2011). Self- monitoring in weight loss: a systematic review of the literature. J Am Diet Assoc, 111:92–102. http://dx.doi.org/10.1016/j.jada. 2010.10.008 PMID:21185970
13. D’Zurilla T, Nezu AM (2006). Problem- Solving Therapy: A Positive Approach to Clinical Intervention, 3rd ed. New York: Springer.
14. Burns RJ, Donovan AS, Ackermann RT et al. (2012). A theoretically grounded systematic review of material incentives for weight loss: implications for interven- tions. Ann Behav Med, 44:375–388. http:// dx.doi.org/10.1007/s12160 - 012-9403- 4 PMID:22907712
15. Jeffery RW, Drewnowski A, Epstein LH et al. (2000). Long-term maintenance of weight loss: current status. Health Psychol, 19 Suppl: 5–16. http://dx.doi.org/10.1037/0278-6133. 19.Suppl1.5 PMID:10709944
16. Perri MG, Foreyt JP, Anton SD (2008). Preventing weight regain after loss. In: Bray GA, Bouchard C, eds. Handbook of Obesity: Clinical Applications, 3rd ed. New York: Informa Healthcare, pp. 249–268.
17. Thomas PR, Stern JS (1995). Weighing the Options: Criteria for Evaluating Weight- Management Programs. Washington, DC: National Academy Press.
18. Wing RR, Hill JO (2001). Successful weight loss maintenance. Annu Rev Nutr, 21:323– 341. http://dx.doi.org/10.1146/annurev.nutr. 21.1.323 PMID:11375440
19. Courneya KS, Booth CM, Gill S et al. (2008). The Colon Health and Life-Long Exercise Change trial: a randomized trial of the National Cancer Institute of Canada Clinical Trials Group. Curr Oncol, 15:279– 285. http://dx.doi.org/10.3747/co.v15i6.378 PMID:19079628
20. Hastert TA, Beresford SAA, Patterson RE et al. (2013). Adherence to WCRF/AICR cancer prevention recommendations and risk of postmenopausal breast can- cer. Cancer Epidemiol Biomarkers Prev, 22:1498–1508. http://dx.doi.org/10.1158/ 1055-9965.EPI-13-0210 PMID:23780838
21. Inoue-Choi M, Lazovich D, Prizment AE, Robien K (2013). Adherence to the World Cancer Research Fund/American Institute for Cancer Research recommendations for cancer prevention is associated with better health-related quality of life among elderly female cancer survivors. J Clin Oncol, 31:1758–1766. http://dx.doi.org/10.1200/ JCO.2012.45.4462 PMID:23569318
22. Anderson AS, Mackison D, Boath C, Steele R (2013). Promoting changes in diet and physical activity in breast and colorectal cancer screening settings: an unexplored opportunity for endorsing healthy behaviors. Cancer Prev Res (Phila), 6:165–172. http:// dx.doi.org/10.1158/1940-6207.CAPR-12- 0385 PMID:23324132
Websites Diabetes Prevention Program Study Docu- ments Web Site: http://www.bsc.gwu.edu/DPP/index.htmlvdoc
Look AHEAD: Action for Health in Diabetes: https://www.lookaheadtrial.org/
290
Summary • Population-wide campaigns can
be an effective and efficient way to modify cancer risk in popu- lations where more immediate channels of personal commu- nication and influence are not feasible.
• In the design and evaluation of campaigns, a balance should be struck between influencing the environment and context in which individual cancer-related behaviours take place, and the intrapersonal determinants of those behaviours.
• Hypothesized causal pathways for change in individuals should be specified in advance, as this will help focus the interventions that make up the campaign as well as defining measures of outcome that will demonstrate and explain effects.
• Campaign messages need to take into account psychological factors that oppose, and facili- tate, change in cancer-related behaviour.
• All campaigns should be subject to evaluation using specified cri- teria, particularly with reference to novel elements.
4.3 4 PREVENTION
Designing and evaluating population-wide campaigns David Hill Melanie Wakefield
Joakim Dillner (reviewer) Surendra S. Shastri (reviewer)
• While campaigns should be sensitive to local cultural fac- tors, there is evidence that many campaigns can be successfully re-used or adapted for use in dif- ferent countries.
When is a campaign the right approach? Population-wide campaigns for pri- mary or secondary prevention of cancer should be considered when the risk or risk factors for a prevalent cancer are spread widely through a population. Such campaigns invari- ably make use of mass media, of- ten through carefully planned paid advertising, as well as other simul- taneous communication and policy interventions. Campaigns would not normally be considered for rare tumours or those concentrated in subpopulations, such as in certain occupations or locations, or for in- fluencing risk behaviours in small demographic subgroups. There are almost certainly other more effec- tive and efficient alternative ways to directly communicate and influence risk in such cases, for instance by communicating at antenatal clinics where pregnant smokers are con- centrated, or by using broadcast or print media in the preferred language of a targeted population subgroup.
Campaigns in cancer preven- tion are often conceived and fund- ed by public health authorities as time-limited operations, whereas in reality their objectives can rarely be achieved or maintained without long-term investment of time, effort, and money. A cancer prevention campaign is better thought of as a health service for which the need is continuous (such as palliative care or ambulances) than as a “project” that has a defined end.
This chapter considers cam- paigns where the objective is to change a defined cancer-related behaviour in members of a defined population. It is not concerned with cancer “awareness-raising” cam- paigns. Indeed, such campaigns are of questionable public health value. Sometimes awareness-raising cam- paigns can do harm, as when pros- tate cancer awareness leads to inap- propriate screening for the disease and to harms that at least arguably outweigh any benefits [1,2]. Neither should campaigns be undertaken as a substitute for potentially effective public health policy and regulation. Rather, they should build on good policy and generate public accept- ance of the need for regulations that facilitate change in behaviours known to increase cancer risk.
Campaigns can be implemented for primary or secondary prevention,
Chapter 4.3 • Designing and evaluating population-wide campaigns 291
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.3
but their use in secondary preven- tion is more limited. Since second- ary prevention depends principally on there being a means of detecting cancer early enough for treatment to be more effective than otherwise, health services (for screening, diag- nosis, and treatment) need to be in place before the appropriate target group is invited or persuaded to at- tend. Contacting individuals directly with invitation letters is far more ef- fective than public advertising of a service alone [3], so a campaign is not likely to be a major element in secondary prevention. In addition, unless the relevant health services are equally available to all members of a target population, population- wide campaigns would generate de- mand that could not be met. Thus, when screening services are intro- duced into a population in stages, it makes little sense to use mass me- dia to communicate. However, as- suming that all required services are ready, there may be a limited role in the early stages for mass com- munication to engender a level of awareness and public understand- ing to prime acceptance of personal invitations when they arrive. Later in the programme’s life, there may also be a role for mass media in refresh- ing public interest should participa- tion rates start to fall [4]. Overall, mass media campaigns alone are unlikely to achieve desired participa- tion rates for secondary prevention programmes.
Habitual and non-habitual behaviours There is a fundamental distinction to be made between strategies de- signed to change risky habitual be- haviours such as tobacco use, over- nutrition, under-exercising, and al- cohol consumption, and strategies aimed to prompt an individual to act only once or twice – such as human papillomavirus (HPV) or hepatitis B virus (HBV) vaccination – or inter- mittently, as is required for effective screening. Campaigns to change habitual behaviours are far more difficult to implement successfully.
The norm is for incremental change, at best, in population rates of a ha- bitual target behaviour, whereas participation rates approaching 50% may rapidly follow the introduc- tion of a screening or vaccination service that was supported by direct invitations and media publicity [5,6].
Habits are the result of a per- son’s lifelong operant learning history, and are likely to be highly resistant to change. In the case of tobacco at least, the difficulties of reversing a habit are further com- plicated by the addictive qualities of nicotine. Nevertheless, campaigns have been shown to be effective in changing population risk behav- iours, including tobacco use, albeit at incremental levels. It is notewor- thy that small percentage changes in risk factor behaviours equate to large numbers of people when a risk factor is common, and that where relative risks are large, the potential for reducing early death is substan- tial. Furthermore, small percentage changes that are sustained over several years add up to big effects. As well as initiating trends, cam- paigns play a vital role in sustaining them [7,8].
Types of campaign message Formulating the campaign message – what the communication actually says to the target individual and how
it is said – is at the core of campaign planning. Judicious application of psychological knowledge is helpful, if not essential. It is important that campaign planners are explicit about the way in which their message is expected to influence the receiver, because this sharpens the focus of the intervention and assists in formu- lating measures to evaluate whether the campaign is “working”. Many lab- oratory studies have investigated the underlying nature of effective health messages – for instance, the rela- tive merits of gain- or loss-framed wording [9] – but the applicability of results obtained to population- wide communications is uncertain. A smaller number of studies have explored message characteristics as they apply to broadcast cancer prevention campaigns [10].
A common issue is when, and when not, to use confronting, shock- ing, or “scary” message content [11]. One would hesitate to use such mes- sages to promote participation in cancer screening. This is because people contemplating a screening offer are likely in a psychologically conflicted “approach–avoidance” state of mind, in which they are at- tracted by the possibility that they will get a reassuring all-clear report but at the same time are fearful of (and wishing to avoid) the possibility of a cancer diagnosis. Hence, a message
Fig. 4.3.1. Series of images from a video used by Australia’s National Tobacco Campaign in 1997–2001. This particular advertisement highlighted how smoking causes arteries to become blocked, graphically emphasizing that “every cigarette is doing you damage”.
292
emphasizing how bad cancer can be might simply amplify avoidance without influencing approach ten- dencies. Positive portrayals of the reassurance offered by a favourable test result might be more effective. In contrast, scary messages are ap- propriate for a person contemplating quitting smoking, who has an “avoid- ance–avoidance” conflict between wanting to avoid lung cancer and wanting to avoid withdrawal symp- toms. Messages that increase the desire to avoid lung cancer – by us- ing vivid imagery, or eliciting pow- erful negative emotions through the use of personal testimonials or by dramatizing someone confronting a situation in which members of the target audience could envision themselves – might well tip the bal- ance in favour of behaviour change in the smoker, who may then be more willing to tolerate withdrawal symptoms to avoid the feared health consequences.
In recent years, there has been a body of work on the efficacy of spe- cific cancer prevention messages in different populations around the world, and certain principles stand out. At least in tobacco control, mes- sages that communicate the serious health harms of smoking by elicit- ing emotional discomfort in viewers elicit similar responses in smokers in many countries of the world, sug- gesting great potential for sharing and recycling of advertisements across jurisdictions. These kinds of messages have high memorability, elicit early responses predictive of smoking behaviour change across many population subgroups within and between countries, and are more likely to generate cessation attempts in response to campaign exposure. This suggests that they are likely to be effective messages for wide dis- semination [12–16]. Overall, this re- search supports the assumption on which many campaigns have been based: that an underlying dynamic of health behaviour change arises from a desire to reduce psychologi- cal discomfort felt when continuing to engage in the risk behaviour.
Campaign formulation Prevention campaigns are usually created and implemented by teams of professionals from varying back- grounds, who all contribute according
to their skills. The teams may include behavioural scientists, medical ex- perts, health promotion specialists, copywriters, graphic designers, film- makers, media specialists, journal- ists, and so on. Consistent messag- ing therefore needs to be built on a shared understanding of the way in which the communications are ex- pected to influence the behaviour of people who are the targets of the campaign. Ideally, a small group, which includes a behavioural scien- tist experienced in such campaigns, would develop a behavioural model tailored to this campaign. Useful models can be found in the literature (Box 4.3.1) [17,18].
Models are based on principles derived from basic research in the behavioural sciences, for instance, in the fields of cognition, social learning, operant conditioning (pos- itive and negative reinforcement), memory, modelling, motivation, and self-efficacy. By their nature, mod- els in the theoretical literature are comprehensive and it is not always useful to try to apply them entirely in formulating a campaign. Rather, the planning group should take in- spiration from them and then cre- ate and articulate a model suited to the intended campaign. The model becomes the reference point for all those who apply their creative talents to developing the campaign [19].
Fig. 4.3.2. Poster from the Spanish Association against Cancer (AECC) cam- paign for skin cancer prevention: “Stand up to the sun. Protect yourself against the sun. Whenever outside, put sunscreen on. At the beach or swimming pool, wear a shirt, a hat, and sunglasses. Don’t let them go into the sun between 12 and 4 pm. Even in the shade, you have to pro- tect yourself.”
Hill and Dixon [18] proposed a set of principles they call the “Big Five”, which encapsulate a large body of psychological and behavioural re- search on determinants of human behaviour.
Few, if any, cancer-related be- haviours are the result of a single influence, so it follows that the more determinants a campaign can activate simulta- neously, the more likely it is that change will occur.
Change in behaviour is likely to occur to the extent that a person:
1. Wants to do it (motivation)
2. Sees others doing it (modelling)
3. Has the capacity to do it (re- sources and self-efficacy beliefs)
4. Remembers to do it (memory)
5. Is reinforced for doing it, or suf- fers for not doing it (positive and negative reinforcement).
The “Big Five” principles can be used in campaign planning as a checklist to maximize forces for population behaviour change.
Box 4.3.1. “Big Five” principles of behaviour change
Chapter 4.3 • Designing and evaluating population-wide campaigns 293
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.3
As well as making explicit the in- tended pathways of influence of cam- paign communications, the elements of a behavioural model enrich the campaign by providing a checklist of opportunities to intervene. Most be- haviours, and certainly those behav- iours involved in cancer prevention, have multiple “causes”. It therefore follows that the more of these causes can be influenced by the campaign, the more likely it is that behaviour change will occur. For instance, a campaign that not only modelled the desired behaviour in a respected per- son but also strengthened motivation by communicating a fresh reason to change and promised positive rein- forcement would more likely be ef- fective than one that relied solely on modelling.
Market segmentation Commercial marketing assumes a product or products that can be ren- dered relatively more attractive to a consumer. Different products can be created, varied, and packaged for population segments in ways that recommended health behaviours cannot. But even assuming that a behaviour can be equated to a prod- uct, one cannot create variants of the behaviour simply to suit the tastes of population segments.
Experience with primary preven- tion campaigns generally suggests that at any given time only a certain
proportion of those at risk are ready to consider behaviour change. This is the segment that mass communi- cations are most likely to activate. This is also a reason why campaigns need to extend over long periods: so that they will influence new seg- ments or “waves” of people as they become ready to change.
That good health and avoiding disease is a universal (i.e. “unseg- mented”) human aspiration probably explains the fortunate finding that anti-smoking television advertise- ments made in high-income coun- tries can readily be adapted for ef- fective use in campaigns in low- and middle-income countries [12,16,20].
Channels of communication The choice of channels to reach a target audience should be based on analysis of that population group’s exposure to various communication channels. Such data are usually available commercially in low- and middle-income countries as well as high-income countries. This analysis will underpin media planning deci- sions about relative balance of effort and expenditure between television, radio, print, outdoor, transit, direct mail, and so on.
Although social media such as Facebook and Twitter are likely to be a means to achieve advocacy goals and strengthen the impact of broad- cast campaigns to change cancer- related behaviour, the evidence base remains small [21]. However, prom- ising results have been reported for pioneering work in low- and middle- income countries, including Viet Nam and Egypt [22,23].
Evaluation It is helpful to think of a campaign as a hypothesis (or set of hypotheses) to be tested. Whether the campaign will “work” can never be known in advance, no matter how well ground- ed it is in prior research. Every new campaign will have novel elements, if only because it should be formu- lated to resonate with the specific culture in which it is implemented. Those implementing campaigns
should pay as much attention to evaluation as to implementation, since that is the path to continuous improvement.
Careful pre-testing of campaign elements can lead to final improve- ments in communications before the campaign is launched. It may give useful insights into how the campaign will be received and how to manage misunderstandings or other difficul- ties that may arise [23]. Most cam- paigns rely heavily on self-reports from individuals, so it is important to use forms of questioning that minimize response bias [24] and that preferably have been validated objectively (e.g. biochemical valida- tion of smoking status, or unobtru- sive observation).
Fig. 4.3.3. The “Move against Cancer” campaign of the Lalla Salma Association Against Cancer (Morocco): “30 minutes of daily exercise is enough to reduce the risks of cancer. Do not wait to be in danger. Move against cancer.”
Fig. 4.3.4. Front page of a leaflet produced by the Hong Kong Cancer Fund as part of a campaign for cervical cancer preven- tion: “Dual protection: regular Pap test and vaccination. The best protection against cervical cancer.”
294
It is when measures of campaign effectiveness are being considered that the investment in clear articu- lation of hypothesized pathways of influence, as mentioned above, becomes apparent. The campaign evaluation should aim to quantify change at as many points as pos- sible in the pathway, not merely the targeted risk behaviour itself. For instance, a campaign aiming to get smokers to make a cessation attempt by broadcasting new mes- sages about macular degeneration should use representative sam- ple surveys to assess population changes in knowledge about macu- lar degeneration and to test whether such knowledge was associated with intentions to quit, as well as to as- sess changes over time in quitting behaviour itself.
As well as being effective, cam- paigns should be cost-effective – meaning they should achieve pre- vention outcomes at a lower per capita cost than alternative ap- proaches. For example, there is good evidence that anti-tobacco
campaigns and skin cancer preven- tion campaigns are cost-effective [25,26]. Unsurprisingly, tobacco control campaigns (together with rigorous evaluations) have been the most common, leading to a substan- tial knowledge base, some of which has relevance for campaigns on other cancer risk behaviours.
Examples of successful campaigns Anti-tobacco An early, landmark campaign was the New South Wales “Quit. For Life” campaign (starting in 1983), which used a powerful image of tar being squeezed from a sponge to convey the build-up of carcinogenic partic- ulate matter in lungs from tobacco smoke. The campaign resulted in an estimated short-term reduction in smoking prevalence of 2.8% [27]. This effect was replicated with the onset of the campaign in the adjoin- ing state of Victoria the following year [28].
California conducted a well- funded and long-term campaign and showed reductions in smoking prevalence, resulting in a 4% differ- ence between California and the rest of the USA in 1995, although momentum later waned, probably as a result of decreased government funding and increased pro-tobacco advertising [29].
New York State used strong graphic and emotional content in a televised mass media campaign supplemented by advertisements to build confidence in the ability to quit smoking, and between 2003 and 2009 reported a fall in prevalence of 18%, compared with a fall of only 5% nationally [30].
In Massachusetts, after the in- troduction of a comprehensive cam- paign that included significant price (tax) increases in 1993, the preva- lence of smoking fell at an average annual rate of 4% until 1999, in con- trast to little or no change in the rest of the USA, excluding California [31].
A national television and radio campaign to warn of the hazards of smokeless tobacco use in India
was shown to produce effects on precursors of behaviour change, in- cluding intentions to modify smoke- less tobacco use, in the target pop- ulation [32].
Sun protection The commencement of the SunSmart campaign in Victoria, Australia, in 1989 was followed by a significant improvement in sun-related beliefs and behaviours and a reduction in prevalence of sunburn (assumed to be an intermediate marker of car- cinogenic risk), after adjusting for ambient levels of ultraviolet radia- tion at the time observations were made [33], and long-term follow-up showed a continuing effect [34].
Diet/nutrition The “1% or Less” campaign was an intensive 6-week mass media cam- paign to promote the use of low-fat milk in Wheeling, West Virginia, in the USA. After the campaign, 34% of milk drinkers indicated that they had changed to low-fat milk, compared with 3.6% in the control community [35]. Also, sales of low-fat milk in- creased from 29% to 46% in the in- tervention community.
Physical activity The VERB campaign (Fig. 4.3.6), which was implemented nationally in the USA from 2002 to 2006, used commercial advertising and market- ing techniques to increase physical activity levels in children. Changes that suggested an effect of the cam- paign were reported in attitudes and behaviour [36].
Campaign principles applied to subpopulations Using Vietnamese-language media, a campaign to promote colorectal cancer screening among Vietnamese Americans in Alameda and Santa Clara counties in California in 2004– 2006 resulted in greater screening participation than in comparison communities that did not receive the campaign [37].
Also using Vietnamese-language media, a campaign in Houston,
Fig. 4.3.5. A campaign for breast cancer prevention in Singapore: “Are you ob- sessed with the right things? The differ- ence between a pimple and breast cancer is that of life and death. Regular breast checks are the best way to fight cancer. Show support for the women in your life by purchasing a Pink Ribbon.”
Chapter 4.3 • Designing and evaluating population-wide campaigns 295
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.3
Texas, to promote the receipt of three hepatitis B virus vaccinations for Vietnamese American children aged 3–18 years resulted in in- creased knowledge among parents about such vaccinations and a sig- nificantly higher increase in the rate of vaccinations compared with a control community [38].
The cancer prevention environment All behaviour is a product both of intra-individual factors and of the en- vironments in which behaviours take place. Campaigns take place in, and can be affected both positively and negatively by, the prevailing social, policy, and regulatory environments. So the best campaigns not only rely on a conducive environment; suc- cessful prevention campaigns help to create such environments.
Ideally, to make a public health impact, the social and regulatory en- vironments will need to move in con- cert with successive phases of public communication campaigns.
Fig. 4.3.6. One of many posters used for the VERB campaign in the USA. The 2002– 2006 campaign was aimed at increasing and maintaining physical activity among tweens (ages 9–13).
296
References
1. Chapman S, Barratt A, Stockler M (2010). Let Sleeping Dogs Lie? What Men Should Know before Getting Tested for Prostate Cancer. Sydney, Australia: Sydney University Press.
2. Schröder FH, Hugosson J, Roobol MJ et al.; ERSPC Investigators (2009). Screening and prostate-cancer mortality in a ran- domized European study. N Engl J Med, 360:1320–1328. http://dx.doi.org/10.1056/ NEJMoa0810084 PMID:19297566
3. Ferroni E, Camilloni L, Jimenez B et al.; Methods to Increase Participation Working Group (2012). How to increase uptake in oncologic screening: a systematic review of studies comparing population-based screening programs and spontaneous access. Prev Med, 55:587–596. http:// dx.doi.org/10.1016/j.ypmed.2012.10.007 PMID:23064024
4. Mullins R, Wakefield M, Broun K (2008). Encouraging the right women to attend for cervical cancer screening: results from a targeted television campaign in Victoria, Australia. Health Educ Res, 23:477– 486. ht tp://dx.doi.org /10.10 93/ her/cym 021 PMID:17615181
5. Ladner J, Besson MH, Hampshire R et al. (2012). Assessment of eight HPV vaccina- tion programs implemented in lowest in- come countries. BMC Public Health, 12:370. ht t p: //dx.doi.o r g /10.118 6 /1471-24 5 8 - 12-370 PMID:22621342
6. Brotherton J, Gertig D, Chappell G et al. (2011). Catching up with the catch- up: HPV vaccination coverage data for Australian women aged 18–26 years from the National HPV Vaccination Program Register. Commun Dis Intell Q Rep, 35:197–201. PMID:22010515
7. Wakefield MA, Bowe SJ, Durkin SJ et al. (2013). Does tobacco-control mass me- dia campaign exposure prevent relapse among recent quitters? Nicotine Tob Res, 15:385–392. http://dx.doi.org/10.1093/ntr/ nts134 PMID:22949574
8. Niederdeppe J, Farrelly MC, Hersey JC, Davis KC (2008). Consequences of dra- matic reductions in state tobacco control funds: Florida, 1998–2000. Tob Control, 17:205–210. http://dx.doi.org/10.1136/tc. 2007.024331 PMID:18390911
9. Gallagher KM, Updegraff JA (2012). Health message framing effects on attitudes, in- tentions, and behavior: a meta-analytic re- view. Ann Behav Med, 43:101–116. http:// dx.doi.org/10.1007/s12160 - 011-9308 -7 PMID:21993844
10. Durkin S, Brennan E, Wakefield M (2012). Mass media campaigns to promote smok- ing cessation among adults: an integra- tive review. Tob Control, 21:127–138. http://dx.doi.org/10.1136/tobaccocontrol- 2011-050345 PMID:22345235
11. Hill D, Chapman S, Donovan R (1998). The return of scare tactics. Tob Control, 7:5 –8. http://dx.doi.org/10.1136/tc.7.1.5 PMID:9706747
12. Wakefield M, Durrant R, Terry-McElrath Y et al. (2003). Appraisal of anti-smoking advertising by youth at risk for regu- lar smoking: a comparative study in the United States, Australia, and Britain. Tob Control, 12 Suppl 2:ii82–ii86. http:// dx.doi.org /10.113 6/tc.12.suppl _ 2.ii82 PMID:12878778
13. Durkin SJ, Biener L, Wakefield MA (2009). Effects of different types of antismoking ads on reducing disparities in smoking ces- sation among socioeconomic subgroups. Am J Public Health, 99:2217–2223. http:// dx.doi.org/10.2105/AJPH.2009.161638 PMID:19833980
14. Farrelly MC, Duke JC, Davis KC et al. (2012). Promotion of smoking cessation with emo- tional and/or graphic antismoking adver- tising. Am J Prev Med, 43:475–482. http:// dx.doi.org/10.1016/j.amepre.2012.07.023 PMID:23079169
15. Davis KC, Nonnemaker J, Duke J et al. (2013). Perceived effectiveness of ces- sation advertisements: the importance of audience reactions and practical im- plications for media campaign planning. Health Commun, 28:461–72. http://dx.doi. o r g /10.10 8 0 /10 410 2 3 6 . 2 012 . 6 9 6 5 3 5 PMID:22812702
16. Wakefield M, Bayly M, Durkin S et al.; International Anti-Tobacco Advertisement Rating Study Team (2013). Smokers’ re- sponses to television advertisements about the serious harms of tobacco use: pre-testing results from 10 low- to middle- income countries. Tob Control, 22:24–31. http://dx.doi.org/10.1136/tobaccocontrol- 2011-050171 PMID:21994276
17. Hornik R, Yanovitzky I (2003). Using theory to design evaluations of communication campaigns: the case of the national youth anti-drug media campaign. Commun Theory, 13:204–224. http://dx.doi.org/10. 1111/j.1468-2885.2003.tb00289.x
18. Hill D, Dixon H (2010). Achieving behav- ioural changes in individuals and popula- tions. In: Elwood JM, Sutcliffe SB, eds. Cancer Control. Oxford: Oxford University Press, pp. 43–61.
19. Hill D, Carroll T (2003). Australia’s National Tobacco Campaign. Tob Control, 12 Suppl 2:ii9–ii14. http://dx.doi.org/10.1136/tc.12. suppl_2.ii9 PMID:12878768
20. Cotter T, Perez D, Dunlop S et al. (2010). The case for recycling and adapting anti-to- bacco mass media campaigns. Tob Control, 19:514–517. http://dx.doi.org/10.1136/tc.20 09.035022 PMID:20852321
21. Chou WY, Prestin A, Lyons C, Wen KY (2013). Web 2.0 for health promo- tion: reviewing the current evidence. Am J Public Health, 103:e9–e18. http:// dx.doi.org/10.2105/AJPH.2012.301071 PMID:23153164
22. Hefler M, Freeman B, Chapman S (2012). Tobacco control advocacy in the age of social media: using Facebook, Twitter and Change. Tob Control, 22:210–214. http://dx.doi.org/10.1136/tobaccocontrol- 2012-050721 PMID:23047890
23. World Lung Foundation Tobacco Control Mass Media Resource. 360 Mass Media Process. Available at http://67.199.72.89/ mmr/english/360formativeResearch.html.
24. Dillman DA (1978). Mail and Telephone Surveys: The Total Design Method. New York: Wiley.
25. Hurley SF, Matthews JP (2008). Cost- effectiveness of the Australian National Tobacco Campaign. Tob Control, 17:379– 384. http://dx.doi.org/10.1136/tc.2008.025 213 PMID:18719075
26. Carter R, Marks R, Hill D (1999). Could a national skin cancer primary prevention campaign in Australia be worthwhile? An economic perspective. Health Promot Int, 14:73–82. http://dx.doi.org/10.1093/hea pro/14.1.73
27. Dwyer T, Pierce JP, Hannam CD, Burke N (1986). Evaluation of the Sydney “Quit. For Life” anti-smoking campaign. Part 2. Changes in smoking prevalence. Med J Aust, 144:344–347. PMID:3485760
28. Pierce JP, Macaskill P, Hill D (1990). Long-term effectiveness of mass media led antismoking campaigns in Australia. Am J Public Health, 80:565–569. http:// d x . d o i . o r g / 10 . 2 10 5 / A J P H . 8 0 . 5 . 5 6 5 PMID:2327533
29. Pierce JP, Gilpin EA, Emery SL et al. (1998). Has the California tobacco control program reduced smoking? JAMA, 280:893–899. http://dx.doi.org/10.1001/jama.280.10.893 PMID:9739973
30. Davis KC, Farrelly MC, Duke J et al. (2012). Antismoking media campaign and smok- ing cessation outcomes, New York State, 2003–2009. Prev Chronic Dis, 9:E40. http://dx.doi.org/10.5888/pcd9.110102 PMID:22261250
31. Biener L, Harris JE, Hamilton W (2000). Impact of the Massachusetts tobacco control programme: population based trend analysis. BMJ, 321:351–354. http:// dx.doi.org /10.113 6/ bmj.321.7257.3 51 PMID:10926595
Chapter 4.3 • Designing and evaluating population-wide campaigns 297
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.3
32. Murukutla N, Turk T, Prasad CV et al. (2012). Results of a national mass me- dia campaign in India to warn against the dangers of smokeless tobacco con- sumption. Tob Control, 21:12–17. http:// d x . d o i . o r g / 10 .113 6 / t c . 2 010 . 0 3 9 4 3 8 PMID:21508418
33. Hill D, White V, Marks R, Borland R (1993). Changes in sun-related attitudes and be- haviours, and reduced sunburn prevalence in a population at high risk of melanoma. Eur J Cancer Prev, 2:447–456. http://dx.doi. org/10.1097/00008469-199311000-00003 PMID:8287008
34. Dobbinson SJ, Wakefield MA, Jamsen KM et al. (2008). Weekend sun protection and sunburn in Australia: trends (1987–2002) and association with SunSmart television advertising. Am J Prev Med, 34:94–101. ht tp://dx.doi.org/10.1016/j.amepre. 20 07.09.024 PMID:18201638
35. Reger B, Wootan MG, Booth-Butterfield S (1999). Using mass media to promote healthy eating: a community-based dem- onstration project. Prev Med, 29:414– 421. http://dx.doi.org/10.1006/pmed.1998.0570 PMID:10564633
36. Huhman ME, Potter LD, Duke JC et al. (2007). Evaluation of a national physical ac- tivity intervention for children: VERB cam- paign, 2002–2004. Am J Prev Med, 32:38– 43. http://dx.doi.org/10.1016/j.amepre.20 06.08.030 PMID:17218189
37. Nguyen BH, McPhee SJ, Stewart SL, Doan HT (2010). Effectiveness of a con- trolled trial to promote colorectal cancer screening in Vietnamese Americans. Am J Public Health, 100:870–876. http:// dx.doi.org/10.2105/AJPH.2009.166231 PMID:20299659
38. McPhee SJ, Nguyen T, Euler GL et al. (2003). Successful promotion of hepati- tis B vaccinations among Vietnamese- American children ages 3 to 18: results of a controlled trial. Pediatrics, 111:1278–1288. http://dx.doi.org/10.1542/peds.111.6.1278 PMID:12777542
298
Summary • Extensive research during the
past few decades has provided strong evidence for preven- tion of cancer and other major noncommunicable diseases. Although preventive actions among people at high risk are important and there are multiple tumour-specific measures, the greatest potential for cancer pre- vention in the general population is through integrated health pro- motion and policies that target certain lifestyle-related risk fac- tors to prevent noncommunica- ble diseases.
• The WHO Global Strategy for the Prevention and Control of Noncommunicable Diseases targets four behavioural risk factors – tobacco, unhealthy diet, physical inactivity, and harmful use of alcohol – with specific strategies.
• The WHO Framework Convention on Tobacco Control describes the most effective measures, espe- cially for demand reduction, and is binding for the countries that have ratified it.
• The WHO Global Strategy on Diet, Physical Activity, and Health includes a range of
4.4 4 PREVENTION
Prevention strategies common to noncommunicable diseases Pekka Puska Thiravud Khuhaprema (reviewer)
Richard Muwonge (reviewer)
evidence-based interventions to influence diet and physical activity in the general popula- tion. Emphasis is placed on measures to promote produc- tion, availability, and marketing of healthier food stuffs.
• The WHO Global Strategy to Reduce the Harmful Use of Alcohol describes initiatives to reduce the harmful use of alco- hol, with emphasis on alcohol policy.
• The major challenge to fram- ing preventive measures for noncommunicable diseases is the gap between the scientific knowledge of risk and its ex- trapolation to achieve reduced incidence.
During the past two decades, the globalization of public health priori- ties has proceeded rapidly. Studies of the burden of disease worldwide have established that noncommu- nicable diseases (NCDs) are now the main causes of death, not only in developed countries but also in developing countries [1]. A landmark publication was The World Health Report 2002, which documented both the change in the burden of disease and the impact of the main risk factors [2].
As a consequence of the emerging NCD epidemic, the WHO Global Strategy for the Prevention and Control of Noncommunicable Diseases was approved by the World Health Assembly in 2000 [3]. It was based on the global increase of NCDs, the causal role of certain behaviour-linked risk factors, the ac- cumulated scientific evidence indi- cating possibilities for NCD preven- tion, and the experience in several countries.
The potential of NCD prevention The WHO Global Strategy for the Prevention and Control of Noncommunicable Diseases [3] acknowledged for the first time that NCDs are a priority area for WHO. The strategy focused on the four main groups of NCDs: cardiovascu- lar disease, cancer, diabetes, and chronic pulmonary disease. It in- cluded comprehensive NCD control activities but emphasized prevention as the key public health approach.
Although the causal role of the main risk factors for NCDs was es- tablished beyond any reasonable doubt before 2000, the understand- ing of the great potential of NCD pre- vention has grown most rapidly since then. In Finland, large reductions in the age-specific rates of cardio- vascular disease and cancer have
Chapter 4.4 • Prevention strategies common to noncommunicable diseases 299
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.4
occurred as a result of the preventive work begun in the 1970s [4].
In the early years of NCD preven- tion, emphasis was placed on early detection, individual treatment, and education about causative agents. This is commonly known as the “high-risk” approach. However, both epidemiological and behavioural/ social considerations emphasize “population-based” prevention as the most effective public health ap- proach [5,6].
The WHO Global Strategy for the Prevention and Control of Noncommunicable Diseases in- troduced the concept of integrated prevention. It was recognized that although specific activities related to different diseases are needed, the most effective public health ap- proach requires intervention to re- duce the impact on the entire popu- lation of risk factors that are common to several major NCDs. The WHO strategy specifically addresses to- bacco use, unhealthy diet, physical inactivity, and harmful use of alcohol.
This chapter discusses cancer prevention in the general popula- tion through this integrated NCD prevention perspective. This is the most cost-effective and sustainable public health approach to achiev- ing a major reduction of the cancer burden in general populations. However, it should be recognized that several other measures, dis- cussed elsewhere in this Report, can substantially contribute to cancer prevention. These include various screening programmes, measures to control certain infectious diseases, and action to prevent skin cancers by restricting exposure to ultraviolet radiation and to prevent lung cancer by addressing outdoor and indoor air pollution, specifically including clean stove programmes in relevant countries.
Population-based integrated NCD prevention targets risk-related behav- iours or lifestyles in the community as a whole and requires emphasis on the built and social environments as determining lifestyle, and on broad health promotion and policy interventions. This has also led to
consideration of the social determi- nants of particular lifestyles, and pos- sibilities to influence them [7]. The early approaches to health educa- tion have, more recently, been sup- plemented by attempts to influence environments – physical and social. Such an “ecological” approach calls for comprehensive activities by differ- ent authorities, which may be equat- ed to multisectoral actions, among which are policy measures by differ- ent sectors of the government, such as the Finnish initiative Health in All Policies [8].
Since the 2000 WHO strategy on NCDs, WHO has developed several major methodologies and tools for national interventions directed to- wards four behavioural risk factors: tobacco use, unhealthy diet, physical inactivity, and harmful use of alcohol.
WHO Framework Convention on Tobacco Control The WHO Framework Convention on Tobacco Control (FCTC) was adopt- ed in 2003 [9]. As of September 2013, 177 countries have ratified it. This is first time that international law has been used in the field of public health. The WHO FCTC specifies ev- idence-based elements of success- ful tobacco control; most initiatives are directed towards reducing de- mand for tobacco products. Demand may be reduced through education and communication (Fig. 4.4.1). Comprehensive educational, pub- lic awareness, and training pro- grammes on tobacco control are di- rected towards health workers, other professionals, community groups, and decision-makers. Reducing tobacco use is difficult because of the strong addictive effects of nico- tine and also because of the social dependence. Both pharmacological and non-pharmacological methods, the latter involving psychological and educational approaches, have been developed to effectively help tobac- co users to quit. Tobacco control policies should include measures to provide cessation services [10].
The global tobacco epidemic is an immediate outcome of the
powerful leverage of the multina- tional tobacco industry, exercised through advertising, promotion, sponsorship, and lobbying. Thus, an important component of tobacco control is a comprehensive ban on advertising, promotion, and sponsor- ship. In relation to this challenge, an international agreement is especially important because of the cross-bor- der spread of advertising material.
Price and tax measures are ef- fective and important means of re- ducing tobacco consumption, par- ticularly among young people (see Chapter 4.1). Although there are no safe tobacco products, national or other relevant authorities can in- troduce regulations on testing and measuring tobacco products, and on levels of particular components in, and emissions from, these products.
Exposure to second-hand to- bacco smoke, especially indoors, is recognized as causing multiple NCDs [11] and adversely affects vul- nerable groups within the population, such as children. Smoke-free envi- ronments discourage people from initiating smoking and continuing to smoke. Thus, any tobacco control policy should include prohibition of smoking in indoor workplaces, on public transportation, in other indoor settings, and in public places such as stadiums.
Packaging for tobacco products should not promote the product by presenting misleading messages. Such misleading messages may in- clude terms like “low tar”, “light”, “ul- tralight”, or “mild”. Tobacco packages
Fig. 4.4.1. An anti-smoking message painted on a pedestrian crossing in the Orchard Road area of Singapore.
300
should carry large and clear health warnings in text or in the form of pic- tures, and/or be generic (Fig. 4.4.2). Australia was the first country to adopt the plain packaging of ciga- rettes [12].
The WHO FCTC specifies a range of measures to reduce sup- ply. The sale of tobacco to minors should be prohibited, and this leg- islation should be comprehensively enforced. Vending machines should be placed so that children cannot obtain access. Surprisingly, a non- trivial percentage of the tobacco used worldwide is smuggled, man- ufactured illicitly, or counterfeited. Therefore, elimination of illicit trade in tobacco products is an issue
requiring international collaboration by relevant authorities.
Although the scientific basis for tobacco control is very strong, fur- ther research is needed in several areas, including, for example, the vulnerability of particular popula- tions. It is important for every coun- try to identify its own research needs in the light of local circumstances. Monitoring trends in tobacco use in the general population and in sub- groups is crucial. It is also important to monitor determinants and patterns of tobacco use, as well as the im- pact of activities related to tobacco control.
In 2009, the WHO FCTC Secretariat published a report on the
global implementation of the con- vention [13]. The report concluded that the implementation varies sub- stantially. Overall, most countries report high implementation rates for measures on packaging and label- ling, sales to minors, and education and training. In general, implemen- tation rates for programmes direct- ed towards cessation remain low.
Promoting healthy diet and physical activity Diet and physical activity are differ- ent from tobacco smoking in many fundamental respects [14]. Tobacco smoking is a very harmful practice and, in principle, not needed at all. Diet and physical activity are part of everyday life. Physical ac- tivity and eating a healthy diet are positive behaviours to be promoted [15]. Also, the health implications of diet are multiple and complicated. Furthermore, although there are general recommendations, the re- lationships of different nutrients and foods to different NCDs vary.
In 2004, WHO adopted a Global Strategy on Diet, Physical Activity, and Health [16]. This strategy pro- vides countries with a comprehensive range of options to influence dietary practices and physical activity in their populations (Fig. 4.4.3). As back- ground for this strategy, WHO, in col- laboration with FAO, published the report Diet, Nutrition and Prevention of Chronic Diseases [17]. Based on expert advice, the report reviewed the evidence relating to nutrition as a means of preventing cardiovascular disease, cancer, diabetes, obesity, osteoporosis, and dental caries.
The WHO strategy’s recommen- dations concerning diet for popu- lations and individuals were the following: • Achieve energy balance and a
healthy weight. • Limit energy intake from total fats,
and shift fat consumption away from saturated fats to unsaturated fats; eliminate the intake of trans- fatty acids (also called trans fats).
Fig. 4.4.2. New graphics and warnings on cigarette packaging are being implemented in several countries. This example from Brazil includes “Horror, Danger, Gangrene, Impotence, Pain, Death”. Although community awareness of smoking as a cause of lung cancer predominates, morbidity and mortality due to smoking include the consequences of respiratory, cardiovascular, and other diseases.
Chapter 4.4 • Prevention strategies common to noncommunicable diseases 301
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.4
• Increase consumption of fruits and vegetables, legumes, whole grains, and nuts.
• Limit the intake of free sugars. • Limit salt (sodium) consumption.
Individuals were encouraged to engage in optimal levels of physi- cal activity throughout their lives. Different types and amounts of physical activity are required to achieve different health outcomes; at least 30 minutes of regular, moderate-intensity physical activ- ity on most days reduces the risk of cardiovascular disease, diabetes (Fig. 4.4.4), colon cancer, and prob- ably breast cancer.
The strategies – both policies and necessary actions to achieve them – must be adopted compre- hensively, taking a long-term per- spective and engaging all sectors of society. The complex interactions between personal choices, social norms, and economic and envi- ronmental factors must be recog- nized. A life-course perspective is essential, and the strategies should be part of broad public health programmes. Priority should be
accorded to activities targeting the lowest-income population groups and communities.
Governments Governments have a primary steer- ing role and are encouraged to act within the context of comprehensive NCD prevention and health promo- tion. National strategies should identify the required measures and should include specific goals, ob- jectives, and actions. Governments should take action in the following areas: • Education, communication, and
public awareness. • Marketing, advertising, sponsor-
ship, and promotion. • Labelling of food products. • Health claims.
School policies and programmes should support the adoption of healthy diets and physical activity of children, and health services should target the habits of patients and the general population. Governments should also invest in research and evaluation of nutrition and physical activity in the general population. National institutions for public health, nutrition, and physical activity have an important role to play in the im- plementation of programmes and in monitoring and evaluation.
Private sector and civil society The private sector can also influ- ence dietary practices and physi- cal activity [18]. The food industry, retailers, catering organizations (including those responsible for meals at schools or workplaces), sporting-goods manufacturers, ad- vertising and recreation businesses, and the media may all contribute to the adoption of good health prac- tices. Moreover, commentaries note that food systems are rarely driven to deliver optimal human diets but pre- dominantly to maximize profits, with increased obesity as the immediate outcome [19].
Recommendations to the food processing industry may include the following:
• Limit the levels of saturated fats, trans fats, sugars, and salt in products.
• Continue to develop and provide affordable, healthy, and nutritious choices to consumers.
• Introduce new products with better nutritional value than was previ- ously available.
• Provide consumers with adequate and understandable product and nutritional information.
• Practise responsible marketing. • Provide simple, clear, and con-
sistent food labels and evidence- based health claims.
• Assist in the development and im- plementation of physical activity programmes.
Workplaces provide settings for health promotion and disease pre- vention, and should make provision for healthy food choices and encour- agement of physical activity.
Civil society and nongovernmen- tal organizations can advocate for healthy lifestyles and influence the food industry to provide healthy prod- ucts. Such organizations can mobilize community sentiment and influence the public agenda. Nongovernmental and civic groups can support the dis- semination of information on healthy diets and physical activity.
Alcohol control Alcohol consumption is among the top 10 causes of mortality worldwide,
Fig. 4.4.3. Poster from the Strong4Life campaign to prevent childhood obesity, in Atlanta, USA. Obesity increases the risk of diabetes, cardiovascular disease, and certain cancers – the major noncommuni- cable diseases – worldwide.
Fig. 4.4.4. Testing of blood glucose with a blood glucose meter, used to help manage diabetes. Reduced incidence of diabetes, cardiovascular disease, and certain tu- mour types may be identified as benefits from reduced obesity, an outcome sought through adoption of better dietary practices and increased physical activity.
302
The global economics of chronic and noncommunicable diseases Felicia Marie Knaul and Andrew Marx
Chronic and noncommunicable dis- eases (CNCDs) have massive rami- fications for global poverty, financial security, and equity [1,2]. The World Economic Forum’s Global Risks 2010 report ranked chronic disease among the three most likely and severe risks facing the planet, with potential economic losses of more than US$ 1 trillion [1].
Between 2005 and 2015, income losses of US$ 558 billion in China and US$ 237 billion in India are pre- dicted due to stroke, heart disease, and diabetes alone [3]. And the eco- nomic impact of CNCDs on low- and middle-income countries (LMICs) will become more severe over time, particularly affecting working-age populations. Although the predict- ed proportion of deaths from NCDs among those aged 15–59 years will fall globally by 2030, it will increase in LMICs [2].
Tobacco is a huge economic risk for LMICs. Tobacco’s estimat- ed annual burden of US$ 500 bil- lion exceeds the total yearly health expenditure of all LMICs combined. Tobacco consumption is estimated to reduce global economic gross do- mestic product by 3.6% per year. If consumption continues to rise, the global annual economic cost of to- bacco may reach US$ 1 trillion be- tween 2020 and 2030 [4].
The estimated annual cost of scaling up interventions to reduce risk factors such as tobacco use and harmful alcohol consumption is only US$ 2 billion for all LMICs – less than US$ 0.40 per person. After including additional interventions
such as vaccinations against hepa- titis B virus to prevent liver cancer and against human papillomavirus to prevent cervical cancer, the an- nual cost increases to US$ 9.4 bil- lion: an annual per capita investment of less than US$ 1 for low-income countries, US$ 1.50 for lower-mid- dle-income countries, and US$ 3 for upper-middle-income countries [3].
A 50% rise in chronic disease incidence and mortality, such as that projected for Latin America from 2002 to 2030, could slow an- nual economic growth by more than 2%, widening disparities between high-income countries and LMICs [5]. This projected economic burden exceeds any to date, including those from malaria and HIV/AIDS [3].
CNCDs also have a tremendous negative impact on the economic well-being of families. CNCDs, es- pecially cancer, increase the risk of catastrophic health expenditure, impairing the ability of families to invest in education and nutrition. In South Asia, for example, the chanc- es of catastrophic expenditures from hospitalizations are 160% higher for cancer than for infectious disease [6]. As typified by data from Egypt, people with NCDs have a 25% lower probability of being employed [7]. Furthermore, the burden of caregiv- ing generally falls heavily on women and girls, exacerbating gender ineq- uities by reducing their labour force participation and access to educa- tional opportunities [2].
Inadequately investing in pre- vention and treatment of CNCDs is recognized as short-sighted
and misguided [1]. In the face of resource constraints, a short-term view would encourage LMICs to focus only on achieving the United Nations Millennium Development Goals. Yet, ignoring CNCDs plac- es many countries at risk of not meeting many of the Millennium Development Goals because of es- calating health costs and the health risks to mothers, infants, and young children [3]. Failure to protect pop- ulations from preventable health risks will inevitably and severely detract from both economic devel- opment and social well-being [3].
References 1. Global Risk Network of the World Economic
Forum (2010). Global Risks 2010: A Global Risk Network Report. Geneva: World Economic Forum.
2. Nikolic IA et al. (2011). Chronic Emergency: Why NCDs Matter. Health, Nutrition and Population (HNP) Discussion Paper. Washington, DC: World Bank.
3. WHO (2011). Global Status Report on Noncommunicable Diseases 2010. Geneva: WHO.
4. Shafey O, Eriksen M, Ross H, Mackay J (2009). The Tobacco Atlas, 3rd ed. Atlanta, GA: American Cancer Society.
5. Stuckler D (2008). Milbank Q, 86:273– 326. http://dx.doi.org/10.1111/j.1468-0009. 2008.00522.x PMID:18522614
6. Engelgau MM et al. (2011). Capitalizing on the Demographic Transition: Tackling Noncommunicable Diseases in South Asia. Washington, DC: World Bank.
7. Rocco L et al. (2011). Chronic Diseases and Labor Market Outcomes in Egypt. Policy Research Working Paper 5575. Washington, DC: World Bank.
responsible for 2–3 million deaths per year [20]. NCDs are among the adverse health outcomes of alcohol consumption, which extend from the acute effects of intoxication, which contributes to accidents and anti-social behaviour, to chronic in- jury, including liver disease, certain
cancers (Fig. 4.4.5), and cardiovas- cular disease.
A special aspect of alcohol con- sumption is its addictive nature. Thus, alcohol is “no ordinary commodity” [21]. While certain adverse health consequences of drinking, including road accidents and criminal assault,
are usually a result of “harmful use”, the prevalence of harmful use is close- ly related to the level of alcohol con- sumption in the general population. Accordingly, interventions should not be confined to “high-risk interventions” among problem users, but should address general alcohol policy and
Chapter 4.4 • Prevention strategies common to noncommunicable diseases 303
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.4
be population-based interventions. There is consensus about effective interventions to reduce alcohol con- sumption [21]. Price and availability are the most effective matters for interventions. Among other interven- tions are measures to limit drinking and driving, and mini-interventions in health services. Information cam- paigns alone seem to have no direct impact.
In 2010, WHO adopted the Global Strategy to Reduce the Harmful Use of Alcohol [22]. This strategy includes evidence-based
policies and measures concerning pricing, availability, and market- ing. Pricing policies are important because consumers are sensitive to changes in the price of alcohol. Policy options include: taxation, es- pecially taking account of the alcohol content of particular beverages; re- stricting the use of price reductions for promoting sales; establishing minimum prices; and price conces- sions for non-alcoholic beverages. Pricing policies should take account of inflation and variation in incomes over time. Strategies that regulate
the availability of alcohol through regulation, policies and programmes include: appropriate systems to reg- ulate production, wholesaling, and serving (regulations may include government monopolies, regulating hours or days of sale, or sales in cer- tain places or during certain events); minimum age for youth being permit- ted to purchase these products; and policies to reduce illicit production or smuggling. Marketing policies should be aimed at reducing the im- pact of advertising and other mar- keting, particularly as these matters affect young people. Frameworks should be established to regulate the nature of, and amount of expenditure on, marketing and sponsorship.
Health services are central to mitigating harm at the individual level among those with conditions caused by alcohol. Health services should address prevention through initiatives to reduce consumption and through treatment interven- tions for individuals and their fami- lies. Mini-interventions, in the form of identification of persons at risk and brief interventions, have proven to be cost-effective.
Conclusion A range of evidence-based inter- ventions have been identified to reduce the deleterious effects of major lifestyle-related risk fac- tors common to cancer and other NCDs. This approach was recog- nized in recent global political de- velopments that culminated in the United Nations High-Level Meeting on the Prevention and Control of Noncommunicable Diseases that took place in 2011 in New York [23]. Influencing the behavioural risk fac- tors common to several major NCDs in the general population is a cost-ef- fective and sustainable public health approach immediately relevant to cancer prevention. Although effec- tive treatment of cancer patients is a necessity, it is generally recognized that more should be done to prevent the otherwise increasing burden of malignant disease.
Fig. 4.4.5. Poster from the 2012 Government of Western Australia campaign on alcohol and cancer risks. Intoxication and addiction are widely recognized hazards of drinking; chronic disease, including cancer, may be less well known.
304
References 1. WHO (2008). The Global Burden of
Disease: 2004 Update. Geneva: WHO.
2. WHO (2002). The World Health Report 2002: Reducing Risks, Promoting Healthy Life. Geneva: WHO.
3. WHO (2000). Global Strategy for the Prevention and Control of Noncommunicable Diseases (WHAA53/14). Geneva: WHO.
4. Puska P, Vartiainen E, Laatikainen T et al., eds (2009). The North Karelia Project: From North Karelia to National Action. Helsinki: National Institute for Health and Welfare.
5. Rose G (1992). The Strategy of Preventive Medicine. Oxford: Oxford University Press.
6. Puska P (2005). Community change and the role of public health. In: Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology: From Aetiology to Public Health. New York: Oxford University Press, pp. 893–907.
7. WHO (2008). Closing the Gap in a Generation: Health Equity through Action on the Social Determinants of Health. Final Report of the Commission on Social Determinants of Health. Geneva: WHO.
8. Puska P, Ståhl T (2010). Health in All Policies – the Finnish initiative: back- ground, principles, and current issues. Annu Rev Public Health, 31:315–328. http:// dx.doi.org/10.1146/annurev.publhealth. 012809.103658 PMID:20070201
9. WHO (2003). WHO Framework Convention on Tobacco Control. Geneva: WHO.
10. Pirie K, Peto R, Reeves GK et al.; Million Women Study Collaborators (2013). The 21st century hazards of smoking and benefits of stopping: a prospective study of one million women in the UK. Lancet, 381:133–141. http://dx.doi.org/10.1016/S01 40-6736(12)61720-6 PMID:23107252
11. Oberg M, Jaakkola MS, Woodward A et al. (2010). Worldwide burden of disease from exposure to second-hand smoke: a retro- spective analysis of data from 192 coun- tries. Lancet, 377:139–146. http://dx.doi. org /10.1016/S0140 - 673 6(10) 613 8 8 - 8 PMID:21112082
12. Wakefield MA, Hayes L, Durkin S, Borland R (2013). Introduction effects of the Australian plain packaging policy on adult smokers: a cross-sectional study. BMJ Open, 3:e003175. http://dx.doi. o r g / 10 .113 6 / b m j o p e n - 2 0 13 - 0 0 317 5 PMID:23878174
13. WHO (2009). 2009 Summary Report on Global Progress in Implementation of the WHO Framework Convention on Tobacco Control. Geneva: WHO.
14. Friel S, Labonte R, Sanders D (2013). Measuring progress on diet-related NCDs: the need to address the causes of the causes. Lancet, 381:903–904. http://dx.doi. org/10.1016/S0140 - 673 6(13) 6 0 6 69 - 8 PMID:23499037
15. Lee I-M, Shiroma EJ, Lobelo F et al.; Lancet Physical Activity Series Working Group (2012). Effect of physical inactivity on ma- jor non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet, 380:219–229. http:// dx.doi.org/10.1016/S0140-6736(12)61031-9 PMID:22818936
16. WHO (2004). Global Strategy on Diet, Physical Activity and Health. Geneva: WHO.
17. Joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases (2003). Diet, Nutrition and the Prevention of Chronic Diseases: Report of a Joint FAO/WHO Expert Consultation. Geneva: WHO (WHO Technical Report No. 916).
18. Hancock C, Kingo L, Raynaud O (2011). The private sector, international devel- opment and NCDs. Global Health, 7:23. http://dx.doi.org/10.1186/1744-8603-7-23 PMID:21798001
19. Stuckler D, Nestle M (2012). Big food, food systems, and global health. PLoS Med, 9:e1001242. http://dx.doi.org/10.1371/ journal.pmed.1001242 PMID:22723746
20. WHO (2009). Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risk Factors. Geneva: WHO.
21. Babor T, Caetano R, Casswell S et al. (2010). Alcohol: No Ordinary Commodity. Research and Public Policy, 2nd ed. Oxford: Oxford University Press.
22. WHO (2010). Global Strategy to Reduce the Harmful Use of Alcohol. Geneva: WHO.
23. United Nations (2011). Political Declaration of the High-Level Meeting of the General Assembly on the Prevention and Control of Non-communicable Diseases. New York: United Nations. Available at www. who.int/nmh/events/un_ncd_summit2011/ political_declaration_en.pdf.
Chapter 4.5 • Legislative and regulatory initiatives 305
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.5
Summary • A range of legislative measures,
and corresponding regulations, affect exposure to carcinogens, although very few are specifi- cally framed in that context.
• Certain behaviours determine personal exposure to carcino- gens, primarily involving tobac- co smoking, alcohol consump- tion, or deliberate sun exposure. Such exposures may be af- fected by legislative measures, which reflect societal attitudes and concerns and are thus not necessarily framed with specific reference to a carcinogenic risk.
• Beyond national or subnation- al initiatives, the international WHO Framework Convention on Tobacco Control establish- es treaty obligations on Parties to adopt relevant legislation, among other measures.
• Prevention of occupational can- cer underpins the banning of a limited number of agents, includ- ing asbestos, and the specifica- tion of workplace exposure lim- its for many other carcinogenic substances. However, work- related carcinogenic risks that are not identified with specific
4.5 4 PREVENTION
Legislative and regulatory initiatives Bernard W. Stewart Robert A. Baan
Vincent Cogliano (reviewer) Jonathan Liberman (reviewer) Christopher J. Portier (reviewer)
agents are not amenable to such action.
• The incidence of cancer attribut- able to atmospheric, waterborne, or foodborne contaminants may be reduced by regulatory inter- vention that results in decreased exposure to pollutants. In cer- tain instances, health benefits have been established.
• Community awareness of regu- lations to prevent or limit carcin- ogen exposure is often focused on particular circumstances of involuntary exposure, rather than on measures that reduce such exposure by affecting behaviour.
Scope A range of legal processes, includ- ing international treaties, national and subnational legislation, and as- sociated regulatory measures, are directed at, or relevant to, cancer control. Such measures rely on the outcomes of medical and scientific research, including behavioural re- search, and are discussed in that context in this chapter. This approach represents a very limited perspective. A comprehensive consideration of le- gal processes relevant to cancer con- trol would involve describing not only relevant measures but also their
implementation as reflected in moni- toring and inspection, their interpre- tation as indicated by litigation and consequential judicial determinations (see Chapter 6.6), and their impact as indicated, for example, by research to establish the efficacy of a particular measure in reducing exposure, and ultimately in reducing any related bur- den of disease.
Regulatory measures adopted under legislation to address the risk presented by exposure to carcino- gens are almost invariably specific to particular classes of agents or cir- cumstances of exposure. There are controls on occupational, environ- mental, pesticide, pharmaceutical, and foodborne exposures. One of the most developed regulatory frame- works can be seen in the USA. In that country, relevant legislation includes the Clean Air Act, the Clean Water Act, the Toxic Substances Control Act, the Food Quality Protection Act, the Comprehensive Environmental Response, Compensation, and Lia- bility Act (commonly known as the Superfund law), the Occupational Safety and Health Act, and the Food, Drug and Cosmetic Act. In common with similar legislation in other coun- tries, regulations adopted under such measures serve to limit exposure to carcinogens. Rarely, if ever, is legis- lation oriented towards cancer pre- vention; the Delaney Clause (1958),
306
an amendment (now repealed) to the United States Food, Drug and Cosmetic Act, is an exception [1]. However, as a means of limiting or preventing exposure to carcino- gens, regulatory processes and the corresponding enabling legislation play a crucial role. The range of mat- ters relevant to cancer control and appropriately subject to legislation and/or regulation expands in parallel with the understanding of the term “environment” (see “Environmental pollution: old and new”).
Apart from limiting exposure to carcinogens, cancer control may be furthered by legislation cover- ing a broad scope. The health-care budget is a major aspect of many national economies and is likely to specify expenditure for cancer con- trol, particularly involving diagno- sis of, treatment of, recovery from, and palliation of malignant disease. Lesser funding, but possibly with wider ramifications, may involve ser- vices relating to population-based cancer screening or the instiga- tion of vaccination programmes. Statutory action underpins the op- eration of population-based cancer registries. The entire health-care workforce is dependent on educa- tion and training, which is supported by legislation. Against such a broad background, this chapter is limited to measures to prevent or reduce the
likelihood of exposure to carcino- gens in various contexts.
In discussions of conditions un- der which exposure to carcinogens may occur, a distinction is often made between behaviours and un- avoidable situations or involuntary circumstances [2]. Rarely are rec- ognized or suspected carcinogens summarily subjected to regulatory control by the imposition of a ban. To address behavioural exposures to carcinogens, banning particu- lar products or agents is often not practicable, even if that option is mentioned to emphasize a particu- lar cause-and-effect relationship. In relation to behavioural exposures, legislative measures short of bans may have a marked and demonstra- ble effect on the likelihood of expo- sure to carcinogens.
Minimizing behaviour-related exposures The circumstances of behaviour- related carcinogenic risk were first identified largely in high-income countries but are now recognized in many instances as occurring worldwide. Such behaviour-related exposures include tobacco smoking, alcohol consumption, and excessive
food intake, as well as unnecessary sun exposure, which is particular to fair-skinned populations. For these factors, although cancer is just one of the adverse health outcomes, it is often the focus of community awareness because it is generally acknowledged as the most feared disease. Accordingly, advocacy di- rected towards the adoption of rel- evant measures may be framed pri- marily with reference to cancer. The benefits of adopting measures to re- duce exposure to behaviour-related carcinogens may extend beyond re- ducing the incidence of cancer.
Tobacco smoking The WHO Framework Convention on Tobacco Control (FCTC), the first- ever international treaty on health under the auspices of WHO, marks a peak in statutory measures calcu- lated to reduce the burden of disease [3]. The WHO FCTC is one of the fruits of medical and biological evi- dence that tobacco smoking causes a broad spectrum of diseases, spe- cifically including cancer. Certain legislative measures that may be adopted to reduce tobacco consump- tion have also been established by research and hence can be adopted
Fig. 4.5.1. The Legislative Palace in Mon- tevideo, Uruguay, is the site where the Uruguayan parliament meets. Governments worldwide have adopted legislation estab- lishing a range of statutory authorities to regulate, for example, exposure to carcino- gens through the use of certain categories of consumer and other products.
Fig. 4.5.2. A poster illustrating the WHO Framework Convention on Tobacco Control as the theme of World No Tobacco Day 2011.
Chapter 4.5 • Legislative and regulatory initiatives 307
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.5
Environmental pollution: old and new Rodolfo Saracci
“Environmental pollution” is a term collectively denoting all agents nox- ious to health that contaminate the human environment. In the broad sense, “the environment” includes everything outside the genetic en- dowment of an individual. Hence, blood represents the body’s internal environment, and food, cosmetics, and tobacco smoke are elements of the personal environment, while air, soil, and water as they occur in the home, in buildings, at workplaces, and outdoors constitute the environ- ment in the stricter and more usual sense of the word.
Noxious pollutants of the envi- ronment (in the stricter sense) have been a scourge since time imme- morial. For centuries, inadequate sanitation of water polluted with sewage caused a heavy burden of infectious disease and death. Early industrial development added ur- ban overcrowding, facilitating the transmission of pathogenic micro- organisms, and a massive increase
in the volume of waste and in toxic chemicals from industrial sources. The same still occurs today, particu- larly in low-income areas of contem- porary megacities, exposing large populations to the combined risks of “old” infectious diseases, most often acute, and of “new”, pollutant- induced ailments like chronic re- spiratory diseases and a variety of cancers, which develop slowly over years or decades after the beginning of the exposures.
The occurrence of cancers in- duced by environmental contami- nants depends on several circum- stances. Pollutants may be present at low concentrations, producing weak effects, in places where effec- tive hygiene facilities and regulations are available, as in high-income countries. Instead, an increase in cancer risk may derive from the fact that several carcinogenic pol- lutants are often jointly present and active within the same medium, for example polycyclic aromatic
hydrocarbons, benzene, arsenic, and other inorganic chemicals in urban air or chlorination by-products in drinking-water. Time of exposure plays a role; for instance, exposure during the perinatal period or in early life might predispose to onset of can- cers both in childhood and in adult life. Whereas pollutants are often present at relatively low concentra- tions in the general environment, high concentrations can occur lo- cally around point sources, typically industrial discharges of fumes in air or disposal of waste in waters and soils. Given the variety of exposure circumstances, the proportion of all cancers attributable to carcino- genic pollutants varies from place to place and with time. However, “overall” estimates applicable to whole countries or regions have been attempted, with attributable fractions ranging from 1% (or even less) to 5–10%.
as specific measures recognized in the WHO FCTC. Parties to the WHO FCTC undertake to increase tobacco taxes; protect citizens from exposure to tobacco smoke at work and in oth- er places; ban tobacco advertising, promotion, and sponsorship, and ap- ply restrictions to advertising where constitutional constraints preclude a complete ban; prohibit the use of deceptive terms such as “light” and “mild”; regulate the packaging and labelling of tobacco products, in- cluding adoption of health warnings; regulate the testing of emissions from tobacco products; promote public awareness of tobacco con- trol issues; implement programmes to provide for cessation of tobacco use; adopt measures to combat il- licit trade in tobacco; prohibit the sale of tobacco products to minors; and implement policies to support alternative sources of income for tobacco workers, growers, and in- dividual sellers.
Some of these provisions, such as those related to cessation pro- grammes, are not immediately ad- dressed by and through legislation. Otherwise, many of the policies may be wholly or largely achieved through legislation to reduce the use of tobac- co products. Banning the production and sale of tobacco products is not envisaged under the WHO FCTC and is not considered to be an option in most countries. However, legislative measures to control or discourage practices that contribute to ill health have the potential to contribute mark- edly to cancer control as well.
The measures adopted by par- ticular countries, or by countries in partnership, to implement tobacco control legislation vary both in the terms of their design and in the time frame for their implementation. In some instances, the prerogative may lie with state or provincial govern- ments to adopt particular legislation. Both the attributable risk of smoking
Fig. 4.5.3. The Tobacco Legislation Amend- ment Bill 2012, passed by the Legislative Assembly of New South Wales, Australia, provides a basis for regulations to limit smoking in public places.
308
for lung cancer and the relative risk are so large that the effect of particu- lar measures to discourage smoking may be readily evident in terms of case numbers – a scenario that does not apply to many cancer prevention initiatives. Thus, lung cancer rates in California during 1988–1997 could be distinguished from those in the rest of the USA, an outcome correlated with legislative measures against tobacco use adopted in California [4].
Legislation to control tobacco use is particularly relevant to developing countries, in many of which tobacco consumption has risen dramatically. This challenge is being addressed. In India, for example, beginning with the Cigarettes Act of 1975, several leg- islative strategies and programmes to curb tobacco use have been im- plemented, with limited success; currently, the Cigarettes and Other Tobacco Products Act of 2003 is de- signed to curb the use of tobacco, to protect and promote public health [5]. Tobacco control initiatives are also being undertaken in China [6].
Alcohol consumption In countries where consumption of alcoholic beverages is legal, the harm attributable to alcohol consump- tion may be limited by encouraging personal responsibility for drinking behaviour and by adopting a range
Fig. 4.5.4. Young men buy alcohol from a vendor at a local market in India. In countries where the sale of alcoholic beverages is permitted, consumption is primarily a matter of personal choice. However, the community may regulate availability by specifying that, for example, people must be of a certain minimum age to legally purchase alcohol.
Table 4.5.1. Text of alcohol warning labels from eight countries
Country Text of label
Brazil “Avoid the risks of excessive alcohol consumption”
Ecuador “Warning. The excessive consumption of alcohol restricts your capacity to drive and operate machinery, may cause damage to your health, and adversely affects your family”
Mexico “Excessive consumption of this product is hazardous to health”
Portugal “Drink alcohol in moderation”
Republic of Korea
One of three messages, including: “Warning: Excessive consumption of alcohol may cause liver cirrhosis or liver cancer and, especially, women who drink while they are pregnant increase the risk of congenital anomalies”
Thailand “Warning: Drinking liquor reduces driving ability” and “Forbidden to be sold to children under 18 years old”
USA “GOVERNMENT WARNING:” “According to the Surgeon General, women should not drink alcoholic beverages during pregnancy because of the risk of birth defects” or “Consumption of alcohol impairs your ability to drive a car or operate machinery, and may cause health problems”
Zimbabwe “Alcohol may be hazardous to health if consumed to excess” or “Operation of machinery or driving after the consumption of alcohol is not advisable”
of regulatory options. In contrast to tobacco use, the immediate focus on harm induced by irresponsible al- cohol consumption involves, for ex- ample, traffic accidents and violent assaults. A longer-term view is that alcohol consumption contributes to chronic disease, including cancer. WHO has adopted a Global Strategy
to Reduce the Harmful Use of Alcohol, which addresses the affordability, availability, and promotion of alcohol – all matters that are subject to statu- tory regulation. Commentary on this WHO initiative refers specifically to the prospect of expansion of the alco- hol market in low- and middle-income countries. For example, in Thailand, alcohol consumption was once low, but the adult per capita consumption rose from 0.26 l in 1961 to 8.47 l in 2001 [7].
During the past decade, leg- islative progress in the European Commission and in Australia and New Zealand to reduce alcohol-relat- ed cancers, among other harms, has been modest at best, and there are differences between countries in the involvement of the alcohol industry in developing relevant measures [8]. In Canada, as in many similar coun- tries, the minimum legal drinking age, service to minors, and drinking and driving are subject to regulation, and there are measures addressing the price of alcohol, density of outlets, server intervention, control systems, and community interventions [9].
Chapter 4.5 • Legislative and regulatory initiatives 309
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.5
At least 15 countries have adopted warning labels for alcoholic beverag- es, and the health effects mentioned include effects on the unborn in some countries; some of these warnings are shown in Table 4.5.1. Cancer is rarely addressed in this context, although reduced consumption of alcohol will contribute to lower inci- dences of cancers of the oral cavity, oesophagus, liver, female breast, and some other sites.
Exposure to ultraviolet radiation Unlike virtually all other carcinogens, ultraviolet radiation, primarily from the sun, presents a risk to a particu- lar racial group, namely fair-skinned people. Among such people, the extent of avoidable sun exposure is determined by behaviour and by circumstances. For children, there is a recognized need to address sun protection policies under broad strictures concerning safety [10]. Regulatory measures that specify requirements for childcare centres have been adopted in particular jurisdictions, as exemplified by re- quirements in Germany [11].
The initial focus of calls to limit provision of commercial indoor tan- ning operations was the protection of adolescents and young adults. Unequivocal evidence of increased melanoma risk caused by use of ultraviolet-emitting tanning devices, as specified in IARC Monographs Volume 100D, has resulted in access to indoor tanning becoming increas- ingly restricted around the world [12]. Commercial tanning services are set to be banned in most Australian
states in the next few years, and other countries may follow.
For people who work outdoors, appropriate sun protection may involve adoption of particular be- haviours, complemented by provi- sion of protective clothing, which may be subject to regulation. It is arguably the task of health educa- tion to improve the recognition by adults of the hazard presented by recreational sun exposure. However, the limited progress that has been made in encouraging young people to be “sunsmart” – a term used in Australian campaigns – is prompting consideration of other options, pos- sibly including regulation [13].
Diet Regulatory measures to reduce obe- sity are relevant to cancer control but are adopted in the broad context of controlling diabetes and cardiovas- cular disease. Measures to encour- age good nutrition are available and are being further developed (See “Taxing sugar-sweetened beverag- es: the Brazilian case”), but these are not aimed at reducing exposure to agents generally recognized as carcinogens. In terms of involuntary exposure to carcinogens via food, most attention has been focused on aflatoxin contamination [14]. A sin- gular focus on cancer resulting from food contamination underpinned historic United States legislation – the Delaney Clause mentioned
Fig. 4.5.5. A polyvinyl chloride production plant in Turkey. Occupational exposure to specific carcinogens such as vinyl chloride can be readily controlled through regulation of relevant concentrations in the workplace; control of occupational exposure to non-specific waste products, or their impact on the wider community, is more difficult to address.
Table 4.5.2. International limit values (2007) for ortho-toluidine
Country Limit value (8 hours) Limit value (short-term)
ppm mg/m3 ppm mg/m3
Belgium 2 8.9 Canada, Québec 2 8.8 Denmark 2 9 4 18 France 2 9 Hungary 0.5 Poland 3 9 Spain 0.2 (skin) 0.89 (skin) Switzerland 0.1 0.5 USA, OSHA 5 22 United Kingdom 0.2 0.89
OSHA, United States Occupational Safety and Health Administration.
310
above – but any such risk is now considered to be best addressed in the context of general food safety legislation [15].
Preventing involuntary exposures Protecting people from injury caused by exposures over which individu- als have little or no control is a par- ticular responsibility of government. Within that broad scope, the pros- pect or proof of cancer causation has prompted a range of legislative measures, depending on the con- text in which relevant exposures may occur.
Occupational exposures Prevention of occupational cancer can be seen in the broader context of avoiding adverse workplace-re- lated health effects due to a broad spectrum of agents. Occupational cancer is wholly preventable by regulatory controls when causation is attributable to a specific chemical or chemicals, as distinct from when increased risk is identified among people engaged in a particular type of work. The adoption of occupation- al exposure limits for carcinogens is a fundamental regulatory approach and generally involves national stan- dards [16]. In Europe, Registration, Evaluation, Authorisation and Re- striction of Chemicals (REACH) reg- ulations are being implemented [17].
Compared with high-income countries, higher levels of exposure to many occupational carcinogens may occur in low- and middle- income countries. The statutory determination of limits for occupa- tional exposure to toxic compounds is being adopted worldwide, as ex- emplified by initiatives in China [18]. Specific information may be gained from the respective regulatory stan- dards themselves. For carcinogens, international standards for indus- trial chemicals are systematically summarized in the relevant IARC Monographs. As an example, for the industrial chemical ortho-toluidine, which is categorized by IARC as Group 1 (carcinogenic to humans) on
the basis of sufficient evidence for causation of bladder cancer in cer- tain workers, exposure limits have been adopted by various countries (Table 4.5.2).
From carcinogenicity data to regulatory control Measures to prevent occupational cancer or other cancers have of- ten been implemented years, if not decades, after unequivocal evidence of carcinogenic risk was available. Many statutory authorities evaluate evidence of carcinogenic- ity, and international collaboration in this field is epitomized by the IARC Monographs. Initially, from the early 1970s, IARC Monographs Programme evaluations were based on epidemiological evidence, when available, together with animal bio- assay data. The evaluation process was expanded to include mechanis- tic data, allowing due recognition of
the likely human hazard from, for example, agents that are carcino- genic in experimental animals but for which epidemiological evidence may be scant. Conversely, bioassay findings may be qualified by biologi- cal data that suggest differences between mechanisms of tumour in- duction in certain rodents and those in humans.
This consideration of the over- all weight of evidence – including that from mechanistic studies – has resulted in IARC evaluations in Group 1 (carcinogenic to humans) or Group 2A (probably carcinogenic to humans) that would previously not have occurred in view of insufficient epidemiological information. As a re- sult, an imperative may be evident to regulatory authorities without the need to await more definitive epide- miology data. Analysing progress and limitations with respect to the next step – from the categorization
Table 4.5.3. Evidence, IARC evaluations, and United States regulations concerning beryllium
Year Event
1930s First industrial uses of beryllium and first reported cases of beryllium-related pneumonitis
1946 Chronic beryllium disease reported for the first time 1949 AEC adopts a 2.0 μg/m3 exposure limit for weapons workers 1970 High lung cancer rates reported among beryllium workers with prior
respiratory illness 1971 OSHA adopts a 2.0 μg/m3 exposure limit 1972 IARC Monographs Volume 1: inadequate evidence in humans; sufficient
evidence in animals 1979 IARC Monographs Supplement 1: Group 2B (possibly carcinogenic to
humans) 1980 Increased lung cancer incidence reported in three studies of beryllium
workers 1980 IARC Monographs Volume 23: limited evidence in humans; Group 2A
(probably carcinogenic to humans) 1987 IARC Monographs Supplement 7: limited evidence in humans confirmed 1989 DOE proposes a 0.5 μg/m3 exposure limit for DOE weapons and clean-up
workers 1993 IARC Monographs Volume 58: sufficient evidence in humans; Group 1
(carcinogenic to humans) 1999 DOE issues a 0.2 μg/m3 exposure limit for DOE weapons and clean-up
workers 2006 California OSHSB adopts a 0.2 μg/m3 exposure limit (8-hour TWA) 2009 ACGIH recommends a 0.05 μg/m3 exposure limit (8-hour TWA) 2012 OSHA exposure limit, 2.0 μg/m3; NIOSH recommended exposure limit,
0.5 μg/m3; EPA ambient air limit, 0.01 μg/m3 (30-day TWA)
ACGIH, American Conference of Governmental Industrial Hygienists; AEC, United States Atomic Energy Commission; DOE, United States Department of Energy; EPA, United States Environmental Protection Agency; NIOSH, United States National Institute for Occupational Safety and Health; OSHA, United States Occupational Safety and Health Administration; OSHSB, Occupational Safety and Health Standards Board; TWA, time-weighted average.
Chapter 4.5 • Legislative and regulatory initiatives 311
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.5
Taxing sugar-sweetened beverages: the Brazilian case Rafael Moreira Claro
Progressively more governments around the world are focusing on ways to use fiscal policies to rectify decades of subsidies for a variety of foods and other goods linked with poor health [1]. WHO has recom- mended such policies for more than a decade as a way to influence food prices and encourage healthy eat- ing. Evidence based on the Brazilian market indicates that taxing sugar- sweetened beverages would be an effective way to control and reduce the consumption of these products [2].
Trends of prices and consump- tion of sugar-sweetened beverages, between 1986 and 2003, obtained in three household budget surveys representative of the 11 major met- ropolitan areas of Brazil, clearly demonstrate how changes in price can affect the consumption of these products (Fig. B4.5.1). As the price of sugar-sweetened beverages de- creased from R$ 5.7 per 1000 kcal to
R$ 3.6 per 1000 kcal, their consump- tion increased significantly from 0.8% to 2.2% of the total calories ac- quired for household consumption. Since the prices of all other foods and beverages remained stable dur- ing this period, the reduction in the price of sugar-sweetened beverages was likely to be even more evident to the population.
The likely impact of taxing sugar- sweetened beverages in Brazil was assessed using national representa- tive data from the early 2000s [2]. A statistically significant reduction in the consumption of these products is expected as a result of increases in price: every 1% increase in price would lead to a 0.84% reduction in consumption in the Brazilian population.
Given that sugar-sweetened beverages are a health hazard re- gardless of their type, initially no caloric sweetened beverage should be exempted from taxation, to avoid
substitution between types of sugar- sweetened beverages [1]. In addi- tion, effective taxation would have to affect the price of these bever- ages on all occasions where they are consumed, in the household or away from it.
Even though taxing of sugar- sweetened beverages has the potential to improve health regard- less of how the revenue is used, the popularity of such an initiative is likely to increase if the revenues are used to support programmes or policies that promote a healthy life- style (such as healthy eating initia- tives in schools) [3]. Data from the most recent national food intake sur- vey, conducted in 2008–2009, show that approximately 6.8 billion litres of sugar-sweetened beverages were consumed by the Brazilian popula- tion during the study period of 1 year (about 750 ml/person/week); thus, an excise tax of 30% per litre (an average increase of R$ 0.3 in the price per litre) would be high enough to reduce consumption significantly and also generate a revenue of R$ 2.2 billion, or almost US$ 1 bil- lion, an amount sufficient to fund investments in promoting a healthy lifestyle or subsidizing healthy foods.
References 1. Brownell KD et al. (2009). N Engl J Med,
361:1599–1605. http://dx.doi.org/10.1056/ NEJMhpr0905723 PMID:19759377
2. Claro RM et al. (2012). Am J Public Health, 102:178–183. http://dx.doi.org/10.2105/ AJPH.2011.300313 PMID:22095333
3. Caraher M, Cowburn G (2005). Public Health Nutr, 8:1242–1249. http://dx.doi.org/ 10.1079/PHN2005755 PMID:16372919
Fig. B4.5.1. Trends in food prices and consumption of sugar-sweetened beverages in major metropolitan areas of Brazil in 1986–1987, 1995–1996, and 2002– 2003. Food prices were adjusted for inflation to represent January 2009 values. Consumption of sugar-sweetened beverages is shown as a percentage of the total calories acquired for household consumption.
0 1 2 3 4 5 6
1986–1987 1995–1996 2002–2003
Price of sugar-sweetened beverages (R$/1000 kcal)
Price of all other foods and beverages (R$/1000 kcal)
Consumption of sugar- sweetened beverages (%)
of carcinogenicity to the implementa- tion of regulatory controls – is inher- ently complicated because it largely identifies national or multinational prerogatives. Progress, or lack of it, is readily evident by considering specific agents.
An international ban on asbestos is in prospect [19], but progress to
this point has taken decades. The categorization of asbestos as carci- nogenic to humans readily followed unequivocal epidemiological find- ings. The first IARC Monograph on asbestos, in 1973, referred to associ- ations between asbestosis and lung cancer that had been reported in 1935, and specified “epidemiological
proof” of the link between exposure to asbestos and increased risks of bronchial carcinoma and mesothe- lioma. Yet, despite more documenta- tion of attributable injury over half a century, regulatory action was pre- cipitated not by specific research findings, nor by authoritative evalua- tion of the available data overall, but
312
by the prospect of litigation involv- ing causation being recognized and compensation being required [20].
Soon after the introduction of beryllium in industrial processes in the 1930s, the first reports of related pneumonitis and chronic beryllium disease were published. Some decades later, the carcino- genicity of beryllium became evi- dent: high lung cancer rates were recorded, in particular among beryl- lium workers with prior respiratory ill- ness. IARC Monograph evaluations of beryllium over four decades have shown a progression, with unequivo- cal animal data being complemented by successively more definitive epi- demiological findings, culminating in a Group 1 evaluation by 1993 (Table 4.5.3). This path has been paralleled by increasingly stringent restrictions – years, and occasion- ally decades, after the relevant sci- entific findings. In some instances, beryllium-induced toxic injury apart from cancer, namely multiple types of respiratory disease, has motivated regulatory action.
Pollution As discussed in Chapter 2.9, air pol- lution causes a minor fraction of lung cancer in high-income countries. Typically, such pollution, which may include carcinogens such as ben- zene, formaldehyde, and 1,3-buta- diene, is subject to statutory control, as exemplified by the United States Clean Air Act legislation [21]. In many jurisdictions, there are regu- latory controls on vehicle emissions [22,23]. The efficacy of regulatory measures to reduce adverse health
effects of atmospheric pollution is being progressively established [24].
While such progress in the USA and similar high-income countries is encouraging, this outcome must be set against the knowledge that two thirds of the global burden of disease attributable to air pollution occurs in the low- and middle-income coun- tries of South-East Asia [25]. The relative lack of regulatory control of atmospheric pollution for much of the world’s population is indicat- ed by the general lack of relevant measurement data outside North America and the European Union. Between 1990 and 2005, a 6% in- crease in global population-weight- ed average ambient concentration of fine particles in the atmosphere occurred, highlighted by increased concentrations in East, South, and South-East Asia and decreases in North America and Europe.
The maximum permissible lev- els of particular water contaminants are specified by regulation in many countries. In relation to cancer control, the immediate and over- whelming concern is arsenic. Other carcinogens recognized as water contaminants include benzene, tri- chloroethylene, and chromium VI. In 1942, a United States standard of 50 μg/l for arsenic was set; in 2001, it was reduced to 10 μg/l, the level that had been adopted by WHO as a guideline for drinking-water in 1992. However, introducing the appropriate regulatory response in developing countries where large populations have much higher concentrations of arsenic in drinking-water, often exceeding 100 μg/l, is complex [26].
Statutory authorities and consumer protection Adverse effects caused by the availability or use of particular com- mercial products are often the re- sponsibility of statutory authorities that have power in relation to, for example, pesticide licensing or con- sumer products, including cosmet- ics. Provisions to protect consumers were invoked in several countries to ban the use of the flame retardant tris(2,3-dibromopropyl) phosphate in children’s sleepwear [27]. In some countries, provisions against mis- leading advertising have served to prevent the marketing of cigarettes using terms such as “light” or “mild”, because of the perception, albeit invalid, that such products present a reduced health risk compared with others on the market [28]. Brazil and the European Union were among the first jurisdictions to act [29].
Typically, the responsibility of statutory authorities, or the outcome of consumer protection legislation, does not involve specification of can- cer control. However, public outrage at perceived inadequate protection against exposures to carcinogens means that the media often tends to focus on involuntary exposures and regulatory inactivity. Accordingly, the community readily recognizes a need for regulatory intervention to prevent or reduce exposure to carcinogens when such measures address involuntary exposures, ex- emplified by those occurring in the workplace or as a result of pollution. However, regulatory measures to in- fluence behaviour may well have the greatest impact on cancer incidence and mortality.
Chapter 4.5 • Legislative and regulatory initiatives 313
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.5
References
1. Appel A (1995). Delaney Clause heads for the history books. Nature, 376:109. h t t p: //d x . d o i . o r g /10 .10 3 8 / 3 7 610 9 a 0 PMID:7603551
2. Stewart BW (2012). Priorities for cancer prevention: lifestyle choices versus unavoidable exposures. Lancet Oncol, 13:e126–e133. http://dx.doi.org/10.1016/ S1470-2045(11)70221-2 PMID:22381935
3. Fong GT, Cummings KM, Borland R et al. (2006). The conceptual framework of the International Tobacco Control (ITC) Policy Evaluation Project. Tob Control, 15 Suppl 3:iii3–iii11. http://dx.doi.org/10.1136/ tc.2005.015438 PMID:16754944
4. Jemal A, Thun MJ, Ries LA et al. (2008). Annual report to the nation on the status of cancer, 1975–2005, featuring trends in lung cancer, tobacco use, and tobacco control. J Natl Cancer Inst, 100:1672– 1694. http://dx.doi.org/10.1093/jnci/djn389 PMID:19033571
5. Mehrotra R, Mehrotra V, Jandoo T (2010). Tobacco control legislation in India: past and present. Indian J Cancer, 47 Suppl 1:75–80. http://dx.doi.org/10.4103/0019- 509X.63870 PMID:20622419
6. Huang J, Zheng R, Emery S (2013). Assessing the impact of the national smok- ing ban in indoor public places in China: evidence from quit smoking related online searches. PLoS One, 8:e65577. http:// dx.doi.org/10.1371/journal.pone.0065577 PMID:23776504
7. Casswell S, Thamarangsi T (2009). Reducing harm from alcohol: call to ac- tion. Lancet, 373:2247–2257. http://dx.doi. org/10.1016/S0140 - 673 6(0 9) 6 0745 - 5 PMID:19560606
8. Davoren SL (2011). Legal interventions to re- duce alcohol-related cancers. Public Health, 125:882–888. http://dx.doi.org/10.1016/j. puhe.2011.09.024 PMID:22036194
9. Giesbrecht N, Stockwell T, Kendall P et al. (2011). Alcohol in Canada: reducing the toll through focused interventions and pub- lic health policies. CMAJ, 183:450– 455. http://dx.doi.org/10.1503/cmaj.100825 PMID:21324848
10. Buller DB, Geller AC, Cantor M et al. (2002). Sun protection policies and environmen- tal features in US elementary schools. Arch Dermatol, 138:771–774. http:// dx.doi.org/10.1001/archderm.138.6.771 PMID:12056958
11. Aulbert W, Parpart C, Schulz-Hornbostel R et al. (2009). Certification of sun pro- tection practices in a German child day- care centre improves children’s sun pro- tection – the ‘SunPass’ pilot study. Br J Dermatol, 161 Suppl 3:5–12. http://dx.doi. org/10.1111/j.1365 -2133.2009.09443.x PMID:19775351
12. Pawlak MT, Bui M, Amir M et al. (2012). Legislation restricting access to indoor tan- ning throughout the world. Arch Dermatol, 148:1006–1012. http://dx.doi.org/10.1001/ archdermatol.2012.2080 PMID:22801924
13. Goulart JM, Wang SQ (2010). Knowledge, motivation, and behavior patterns of the general public towards sun protection. Photochem Photobiol Sci, 9:432– 438. ht tp://dx.doi.org/10.1039/b9pp0 0122k PMID:20354635
14. Pitt JI, Wild CP, Baan RA et al., eds (2012). Improving Public Health through Mycotoxin Control. Lyon: IARC (IARC Scientific Publications Series, No. 158).
15. Merrill RA (1997). Food safety regulation: reforming the Delaney Clause. Annu Rev Public Health, 18:313–340. http://dx.doi. org/10.1146/annurev.publhealth.18.1.313 PMID:9143722
16. Bolt HM, Huici-Montagud A (2008). Strategy of the scientific committee on occupational exposure limits (SCOEL) in the derivation of occupational exposure limits for carcinogens and mutagens. Arch Toxicol, 82:61–64. http://dx.doi.org/10.1007/ s00204-007-0260-z PMID:18008062
17. Milan C, Schifanella O, Roncaglioni A, Benfenati E (2011). Comparison and pos- sible use of in silico tools for carcinogenic- ity within REACH legislation. J Environ Sci Health C Environ Carcinog Ecotoxicol Rev, 29:300–323. http://dx.doi.org/10.1080/105 90501.2011.629973 PMID:22107165
18. Liang Y, Wong O, Yang L et al. (2006). The development and regulation of occu- pational exposure limits in China. Regul Toxicol Pharmacol, 46:107–113. http://dx. d o i.o r g /10.1016 / j.y r t p h. 2 0 0 6 . 0 2 . 0 07 PMID:16624464
19. Sim MR (2013). A worldwide ban on as- bestos production and use: some re- cent progress, but more still to be done. Occup Environ Med, 70:1–2. http://dx.doi. org/10.1136/oemed-2012-101290 PMID: 23248187
20. Gee D, Greenberg M (2001). Asbestos: from ‘magic’ to malevolent mineral. In: Harremoës P, Gee D, MacGarvin M et al., eds. Late Lessons from Early Warnings: The Precautionary Principle 1896 –2000. Copenhagen: European Environmental Agency, pp. 52–63.
21. Woodruff TJ, Axelrad DA, Caldwell J et al. (1998). Public health implications of 1990 air toxics concentrations across the United States. Environ Health Perspect, 106:245–251. http://dx.doi.org/10.1289/ ehp.98106245 PMID:9518474
22. Bahadur R, Feng Y, Russell LM, Ramanathan V (2011). Impact of California’s air pollution laws on black carbon and their implications for direct radiative forcing. Atmos Environ, 45:1162–1167. http://dx.doi.org/10.1016/j. atmosenv.2010.10.054
23. Zhou Y, Wu Y, Yang L et al. (2010). The impact of transportation control mea- sures on emission reductions during the 2008 Olympic Games in Beijing, China. Atmos Environ, 44:285–293. http://dx.doi. org/10.1016/j.atmosenv.2009.10.040
24. Lioy PJ, Georgopoulos PG (2011). New Jersey: a case study of the reduction in urban and suburban air pollution from the 1950s to 2010. Environ Health Perspect, 119:1351–1355. http://dx.doi.org/10.1289/ ehp.1103540 PMID:21622086
25. Brauer M, Amann M, Burnett RT et al. (2012). Exposure assessment for estima- tion of the global burden of disease at- tributable to outdoor air pollution. Environ Sci Technol, 46:652–660. http://dx.doi. org/10.1021/es2025752 PMID:22148428
26. Smith AH, Smith MM (2004). Arsenic drink- ing water regulations in developing coun- tries with extensive exposure. Toxicology, 198:39– 44. http://dx.doi.org/10.1016/j.tox. 2004.02.024 PMID:15138028
27. de Boer JG, Mirsalis JC, Provost GS et al. (1996). Spectrum of mutations in kidney, stomach, and liver from lacI transgenic mice recovered after treat- ment with tris(2,3-dibromopropyl)phos- phate. Environ Mol Mutagen, 28:418– 423. http://dx.doi.org/10.1002/(SICI)1098- 2280(19 9 6)28:4< 418::AID - EM17>3.0. CO;2-I PMID:8991072
28. Anderson SJ, Ling PM, Glantz SA (2007). Implications of the federal court order banning the terms “light” and “mild”: what difference could it make? Tob Control, 16:275–279. http://dx.doi.org/10.1136/tc. 2006.019349 PMID:17652244
29. Borland R, Fong GT, Yong HH et al. (2008). What happened to smokers’ beliefs about light cigarettes when “light/mild” brand de- scriptors were banned in the UK? Findings from the International Tobacco Control (ITC) Four Country Survey. Tob Control, 17:256–262. http://dx.doi.org/10.1136/tc. 2007.023812 PMID:18426868
314
Summary • Hepatitis B virus (HBV), which
causes chronic hepatitis, cirrho- sis, and a large fraction of liver cancer cases, is a very common infection in some areas of the world.
• Highly effective vaccines against HBV and liver cancer have been available since 1982, and most countries include HBV vaccina- tion in their childhood immuniza- tion programmes.
• A group of about 12 human papillomaviruses (HPVs), par- ticularly HPV types 16 and 18, cause most cervical and anal cancers and an important frac- tion of vulvar, vaginal, penile, and oropharyngeal cancers.
• Two prophylactic vaccines con- sisting of empty viral capsids of HPV types 16 and 18 were recently developed. The vac- cines are safe and prevent al- most 100% of anogenital infec- tions and precancerous lesions among previously unexposed in- dividuals, with enormous poten- tial for cancer control worldwide.
• HPV vaccination of adolescent girls is now recommended and is slowly being implemented in low- and middle-income countries,
4.6 4 PREVENTION
Vaccination Rolando Herrero Silvia Franceschi
Andrew J. Hall (reviewer) Eduardo Lazcano Ponce (reviewer) Mark Schiffman (reviewer)
where more than 80% of cervical cancer cases worldwide occur.
A notable fraction of human cancers (16%) are caused by infections and are largely amenable to effective pre- ventive interventions [1]. Among the most important infections associated with cancers are hepatitis B virus (HBV), hepatitis C virus, human pap- illomavirus (HPV), and Helicobacter pylori (see Chapter 2.4).
Chronic infection with HBV is one of the most important causes of liver cancer (hepatocellular carcinoma) worldwide, particularly in highly
endemic areas like sub-Saharan Africa, the Amazon basin, China, the Republic of Korea, and countries in South-East Asia [2]. Vaccines against HBV have been available for several decades and are part of childhood vaccination programmes around the world. The efficacy of the vaccine in preventing chronic HBV infection and liver cancer has now been clearly demonstrated among children and adolescents, and it is expected to reduce incidence of adult hepatocellular carcinoma in many areas when vaccinated popu- lations reach adulthood.
HPV is a very common sexu- ally transmitted virus that causes
Fig. 4.6.1. View from the Chiang Kai-shek Memorial Hall, in Taipei. Taiwan, China, launched a nationwide hepatitis B virus vaccination programme for newborn infants in July 1984.
Chapter 4.6 • Vaccination 315
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.6
cancers of the cervix, anus, vulva, vagina, penis, and oropharynx among individuals in whom HPV infection is not cleared. Recently, highly effective vaccines have be- come available to prevent infection by HPV16 and 18, the types respon- sible for most HPV-related cancers. Vaccine efficacy and cost-effective- ness are maximal among previously unexposed women, and HPV vac- cination is being implemented pro- gressively among adolescent girls in developed and developing countries. The vaccines are efficacious to pre- vent infections and lesions at all the anatomical sites where they have been investigated, and mass vac- cination programmes are expected to reduce incidence and mortality from cancers associated with these viruses in the next few decades [3]. This chapter summarizes some of the epidemiological features of HBV and HPV infections and their associated cancers and discusses the characteristics and performance of vaccines against them.
Hepatitis B virus Hepatitis B virus and liver cancer HBV is a highly contagious DNA virus transmitted by exposure to in- fected blood and other body fluids, including semen and vaginal fluids. The virus is transmitted from mother to infant and from child to child, as well as by unsafe injections, sexual contact, and blood transfusions. Perinatal transmission from infect- ed mothers to their newborn infants or from one child to another is very common in highly endemic areas, and the virus can also be transmitted by fomites [4].
HBV infection is a major global health problem. Worldwide, an es- timated 2 billion people have been infected with HBV, and more than 360 million have chronic liver infec- tions. HBV causes a life-threatening liver infection, namely hepatitis B, which often leads to chronic liver disease. The occurrence of chronic infection is inversely related to the age of acquisition, from 90% among
people infected perinatally to less than 5% among otherwise healthy adults [5]. HBV-driven persistent inflammation, liver necrosis, and re- generative proliferation lead to cir- rhosis and hepatocellular carcinoma. Among people chronically infected, it is estimated that 25% will die from liver disease, including cancer. HBV is responsible for 50–90% of liver cancer cases in high-risk areas [6]. These cancers can also develop without fibrosis because HBV may have a direct carcinogenic effect [4].
Hepatitis B virus vaccine A vaccine against HBV has been avail- able since 1982. The current vaccine is a recombinant HBV surface antigen produced in yeast or mammalian cells into which the HBV surface antigen gene has been inserted using plas- mids. The expressed HBV surface antigen self-assembles into spherical particles that expose the highly im- munogenic “a” determinant. Alum (aluminium phosphate) is generally used as an adjuvant. HBV vaccine, of which the initial dose (of a total of three recommended doses) is given within the first 24 hours after birth, is 95% effective in preventing HBV in- fection and its chronic consequences, even when antibody levels become
undetectable, and was the first vac- cine against a major human cancer [7]. The vaccine has an outstanding record of safety and effectiveness. Since 1982, more than 1 billion dos- es of HBV vaccine have been used worldwide. In many countries where 8–15% of children previously became chronically infected with HBV, vacci- nation has reduced the rate of chronic infection to less than 1% among im- munized children.
Prevention of liver cancer has been demonstrated in children and adolescents in the early vac- cinated cohorts in Taiwan, China, where a nationwide HBV immu- nization programme for newborn infants was initiated in July 1984 [8,9]. Immunization has lowered the prevalence of chronic HBV carriers, incidence of hepatocellu- lar carcinoma, and mortality from fulminant hepatitis [10]; these ef- fects are evident in the birth cohort born in 1981–1984, which received HBV vaccines at preschool ages instead of in infancy. The Republic of Korea also initiated mass HBV vaccination in the early 1980s [11].
A high incidence of liver cancer in Qidong City, China, was attrib- utable to endemic HBV infection and dietary exposure to aflatoxin,
Fig. 4.6.2. A child getting vaccinated in The Gambia. Hepatitis B virus vaccination trials in progress in the country are expected to yield results in the next few years.
316
prompting dietary interventions and the initiation of neonatal HBV vac- cination. As recently reported, liver cancer incidence has fallen dramati- cally: compared with 1980–1983, age-specific liver cancer incidence rates in 2005–2008 had significantly decreased by 14-fold at ages 20– 24 years, and consistent data were reported for other cohorts [12].
In 2010, 179 countries reported inclusion of the HBV vaccine in their national infant immunization pro- grammes. This is a major increase compared with 31 countries in 1992, the year that the World Health Assembly passed a resolution to rec- ommend global vaccination against HBV. Nearly 70% of children born worldwide received three doses of the HBV vaccine (Fig. 4.6.3).
Human papillomaviruses Human papillomaviruses and cancer HPV is considered a necessary cause of cervical cancer, with an
attributable fraction of nearly 100%. Multiple epidemiological studies over the past three decades have con- firmed the etiological role of about 13 HPV types in cervical cancer; HPV16 and 18 are detectable in about 70% of tumours. The natural history and molecular mechanisms involved in cervical carcinogenesis are well un- derstood [13].
In the cervix, HPV is a sexually transmitted infection acquired by most women shortly after initiation of sexual activity. The most com- mon morphological manifestation of HPV infection consists of minor epithelial abnormalities (equivocal and low-grade cellular changes). The initial infection regresses spon- taneously in the majority of women, but among those women where the infection becomes persistent, pre- cancerous lesions (advanced intra- epithelial neoplasia) can develop. If not treated, these lesions can lead to cervical cancer after several years. HPVs have well-identified oncogenes whose products interact with tumour
suppressor gene proteins of the host to induce malignant transformation [14]. Prevention of a necessary cause of disease, by definition, should pre- vent the disease.
An important fraction of other anogenital cancers, including can- cers of the anus, vulva, vagina, and penis, are also HPV-related, and in these tumours HPV16 is the strongly predominant type. Although mo- lecular mechanisms similar to those observed in the cervix are probably involved, there is limited knowledge of the natural history of these tumours [13]. Importantly, and similarly to the cervix, most of these tumours show an important increase among immu- nodeficient individuals. Apart from the lesions in the anogenital tract of both men and women, HPV is impli- cated in a geographically variable and pathologically distinct fraction of cancers of the oropharynx [15], and the incidence of this cancer has re- cently tended to increase in several developed countries. In fact, in the USA, where the increase in incidence
Fig. 4.6.3. Immunization coverage with three doses of hepatitis B virus vaccine in the national infant immunization schedule, 2010.
Chapter 4.6 • Vaccination 317
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.6
Early implementation and monitoring of HPV vaccination in Bhutan Tandin Dorji and Ugyen Tshomo
Bhutan was the first low-income country to introduce vaccination against human papillomavirus (HPV), the necessary cause of cervical cancer. Cervical cancer is the most common cancer in Bhutanese women. To help prevent the disease, the Ministry of Health developed a national vaccination programme. The Ministry’s con- siderations included disease inci- dence, the limited reach of screen- ing, the poor outcomes associated with late diagnosis of disease, and the feasibility of a HPV vaccina- tion programme. For national in- troduction of HPV vaccination, the Ministry recommended immuniza- tion with quadrivalent (HPV6, 11, 16, and 18) vaccine for 12-year-old girls and a “catch-up” campaign for girls aged 13–18 years in the first year of the programme. In 2010, a
pharmaceutical company provided free quadrivalent HPV vaccine for all girls aged 12–18 years, and for the next 5 years the Australian Cervical Cancer Foundation fund- ed the vaccine for 12-year-old girls. Health workers administered the vaccine in schools; girls not at school received the vaccine at health facilities. In 2010, more than 130 000 doses of quadrivalent HPV vaccine were administered, with a three-dose vaccination coverage rate of 92%. There were no seri- ous adverse events, and the vac- cine was well tolerated. The rapid implementation, acceptability, and success of this national HPV vac- cination catch-up programme are attributed to strong political com- mitment and a well-functioning pri- mary health-care system. Bhutan’s experience provides an example
for other low-income countries considering the introduction of HPV vaccination.
Monitoring of HPV vaccination is challenging: it will take two or three decades before its impact on the on- set of cervical cancer will become manifest in cancer registry data. In the near term, the most feasible and informative outcome of vaccination is the decrease in the prevalence of HPV infection and, most notably, the variation in the ratio between HPV types prevented by the current vac- cines and other HPV types in sen- tinel populations of sexually active adolescents and young women. A timely demonstration of the effec- tiveness in Bhutan will sustain the willingness of other low-income countries to continue their efforts. Therefore, in 2012 the Ministry of Health of Bhutan, in collabora- tion with Jigme Dorji Wangchuck National Referral Hospital and IARC, started a study that will regu- larly evaluate HPV prevalence in cervical cells and urine samples of women aged 18 years or older. By 2016, the early impact of vaccination will start to be detectable in women aged 25 years or younger. An expan- sion of cervical screening and a shift from the current cytology-based screening to HPV-based screening in the country is also being consid- ered in the framework of the same project. This investment will improve cervical cancer prevention in older women and could ultimately serve as the basis for long-term vaccina- tion monitoring.
Reference Ladner J et al. (2012). BMC Public Health, 12:370. http://dx.doi.org/10.1186/1471-2458- 12-370 PMID:22621342
Fig. B4.6.1. Staff of the Jigme Dorji Wangchuck National Referral Hospital in Thimphu, Bhutan.
of oropharyngeal cancer is observed mainly among men, the number of HPV-related cancers is expected to be higher in men than in women in the next few decades, attributable
in part to the consideration that screening has greatly lowered rates of cervical cancer in that country [16]. HPV16 is responsible for about 60% of cervical cancer tumours, but
in HPV-associated tumours of other anogenital sites and the oropharynx this viral type is more predominant, responsible for about 90% of the HPV-positive cases.
318
Introduction and monitoring of a national HPV vaccination programme in Rwanda Agnès Binagwaho, Maurice Gatera, and Fidele Ngabo
Cervical cancer is the most com- mon cancer among women in Rwanda, where, before 2011, nei- ther cervical cancer screening nor human papillomavirus (HPV) vac- cination were available in public health facilities. In 2010, however, the Rwandan Ministry of Health de- veloped a National Strategic Plan for the Prevention, Control, and Management of Cervical Lesions and Cancer. A multidisciplinary
committee devised a strategic plan for the preparation, implementa- tion, and evaluation of a national HPV vaccination programme, of which the key aspects included: (i) a public–private partnership with a pharmaceutical company to provide free vaccine for 3 years of HPV vaccinations and offer reduced prices for future doses; (ii) a delivery approach based on school grade; (iii) a strong community involve- ment in identifying and vaccinating girls absent from, or not enrolled in, school; and (iv) a nationwide sen- sitization campaign preceding the delivery of the first dose.
Since 2011, girls enrolled in pri- mary grade six have been receiving the full three-dose course of HPV vaccine, with additional outreach to the small number of 12-year-old girls not at school. During the pro- gramme’s second and third years (2012 and 2013), a “catch-up” phase targeting girls in the third year of secondary school has ensured com- plete coverage of all pre-adolescent and adolescent girls. In 2014 and be- yond, only girls in primary grade six will be vaccinated.
In its first year, Rwanda’s HPV vaccination programme achieved a three-dose vaccination coverage of 93.2% among an estimated 98 762 eligible girls in grade six. Rwanda is the first example of a successful
nationwide delivery of HPV vac- cine in a country eligible for support from the GAVI Alliance, and should motivate other countries to expand their vaccination programmes to include the HPV vaccine, with due customization according to their epidemiological, economic, political, and health system contexts. This is especially timely now that the GAVI Alliance has included the vaccine in its immunization package and the price has been lowered to US$ 5 per vaccine dose for GAVI Alliance- eligible countries.
Rwanda’s HPV vaccination programme is expected to have an important impact on the future bur- den of cervical cancer in Rwanda. In the short term, however, reliable evidence of the effectiveness of the vaccination programme is crucial to help national planners sustain in- vestment in the programme. Thus, a collaboration with IARC is under way to provide timely high-quality data on the impact of the vaccina- tion programme on the prevalence of HPV infection. Such evidence of vaccine effectiveness should also facilitate the introduction of success- ful programmes into other low- and middle-income countries.
Reference Binagwaho A et al. (2012). Bull World Health Organ, 90:623–628. http://dx.doi.org/10.2471/ BLT.11.097253 PMID:22893746
Fig. B4.6.2. A young Rwandan mother and her child.
Human papillomavirus vaccines Two non-infectious subunit vaccines against HPV have been developed, both composed primarily of virus-like particles and produced by expres- sion of the HPV L1 gene in insect cells (bivalent vaccine) or yeast (quadrivalent vaccine). The L1 gene codes for most of the viral capsid, and after producing the L1 protein, it self-assembles as virus-like par- ticles, which are highly antigenic. They produce a strong antibody
response that prevents infection by neutralizing infectious virus in the mucosa at the time of contagion. A comprehensive review has recently been published [3].
The bivalent vaccine includes HPV16 and HPV18 virus-like parti- cles and is produced with a complex adjuvant system (ASO4) consisting of monophosphoryl lipid A and alum. The quadrivalent vaccine, produced with alum adjuvant, includes HPV16 and HPV18 virus-like particles and, in addition, HPV6 and HPV11 virus-
like particles, the viral types respon- sible for most benign genital warts.
Both vaccines are almost 100% effective in preventing cervical HPV infections and precancerous lesions associated with the oncogenic vi- ruses included in the vaccine for- mulations (HPV16 and 18) among women not previously infected, and the vaccines are therefore mainly recommended for adolescent girls before initiation of sexual activity. A summary of demonstrated ef- ficacy of the two cervical cancer
Chapter 4.6 • Vaccination 319
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.6
Table 4.6.1. Key findings from clinical trials of HPV VLP vaccines
Study group Outcome Quadrivalent vaccine Bivalent vaccine
Young women Infection efficacy Proven Proven
CIN2+ efficacy Proven Proven CIN3 efficacy Proven Proven VIN/VaIN 2/3 efficacy Proven Provena Genital warts efficacy Proven Not a target Anal infection efficacy Not provenb Proven Partial cross-protection infection Proven Proven Partial cross-protection CIN2+ Proven Proven Therapeutic efficacy None None Safety No concerns No concerns Mid-adult womenc Infection efficacy Proven Provena
CIN2+ efficacy Proven Not proven Immunogenicity Proven Proven Safety No concerns No concerns Young men Infection efficacy Proven Not proven
Genital warts efficacy Proven Not a target Anal infection efficacy Proven Not proven AIN2+ efficacy Proven Not proven Safety No concerns No concerns Children Infection efficacy Not proven Not proven
Disease efficacy Not proven Not proven Immunogenicity Proven Proven Safety No concerns No concerns
AIN2+, anal intraepithelial neoplasia, grade 2 or worse; CIN, cervical intraepithelial neoplasia; HPV, human papillomavirus; VaIN, vaginal intraepithelial neoplasia; VIN, vulvar intraepithelial neoplasia; VLP, virus-like particle. a Meeting abstract, not yet published. b “Not proven” indicates that no data have been reported. c See comments in the text on cost-effectiveness for this group.
vaccines is presented in Table 4.6.1 [3]. Although there is evidence of protection by both vaccines against HPV infection and disease among women older than 25 years, the benefit–cost ratio is reduced, in part because of the lower incidence of lesions among women infected by HPV later in life [17].
The vaccines also afford limited protection against HPV types phylo- genetically related to HPV16 and 18. Cross-protection against persistent infection of 6 months or more has been reported against HPV31 for the quadrivalent vaccine and against HPV31, 33, 52, 45, and 51 for the bivalent vaccine. However, the sig- nificance of early protection against these additional types is not clear
and needs to be monitored over the long term and against cervical disease. Neither of the vaccines is effective against established infec- tions [18].
The quadrivalent vaccine has been shown to also prevent the majority of vulvar and vaginal HPV infections and lesions associated with HPV16 and 18, as well as genital warts (caused by HPV6 and 11). The bivalent vaccine has been shown to prevent anal HPV16 and 18 infections in young women (Table 4.6.1). Based on cross-sectional studies of women aged 18–24 years who received Pap screening in family planning clinics throughout Australia, there was a de- crease of 77% in prevalence of HPV types targeted by the quadrivalent
vaccine from the period before (2005–2007) to the period after (2009–2010) the vaccine was widely implemented (Fig. 4.6.4) [19]. Among men, the quadrivalent vaccine has been shown to effectively prevent external genital lesions, including genital warts and penile, perianal, or perineal intraepithelial lesions, and among men who have sex with men it has been demonstrated to prevent anal intraepithelial neoplasia.
Trials of the bivalent and quad- rivalent vaccines in terms of infection and lesional end-points have been reviewed [20]. Among HPV-negative women, both vaccines have demon- strated high efficacies against per- sistent HPV infections and cervical cancer precursors. The quadrivalent
320
vaccine is efficacious (> 75% vaccine efficacy) against any of the more se- vere precursors of vulvar, vaginal, and anal cancers.
Both vaccines have demonstrat- ed an excellent safety record in the several years since they were li- censed and have now been incorpo- rated into vaccination programmes in most developed countries. They are recommended for girls before initiation of sexual activity, with ad- ditional “catch-up” programmes for young women in some countries. Although the efficacy has not been shown for younger children, immu- nogenicity bridging studies have shown that the level of the antibody response to the vaccine is stronger in adolescent boys and girls than
in young adult women [21]. In gen- eral, among women and men of all ages, the vaccines induce a strong immune response to both antigens that persists for at least 8 years, without evidence of declining effi- cacy so far.
More than 40 countries have already introduced HPV vaccina- tion, mainly among women [22], with a high uptake in some regions but not in others. It will probably only be possible to determine the extent of attributable reduction in cervical cancer incidence after several decades, particularly be- cause vaccination programmes have only recently been initiated in developing countries (see “Early implementation and monitoring of
HPV vaccination in Bhutan” and “Introduction and monitoring of a national HPV vaccination pro- gramme in Rwanda”). Vaccination of boys to prevent certain cancers in men and to prevent transmission of infection to women is under way in some developed countries. Some areas have implemented vaccina- tion programmes with alternative schedules, including the application of only two doses, based on immu- nogenicity data showing very simi- lar antibody responses, particularly among young girls. In addition, data from a clinical trial in Costa Rica indicated evidence of protection from fewer than three doses of the bivalent vaccine against persistent infections with relevant viral types [23]. In the same trial, reduced prev- alence of oral HPV infection was evident 4 years after vaccination – a result that has implications for the prevention of HPV-associated oropharyngeal cancer, which is in- creasingly common [24].
New vaccines are being devel- oped that could overcome some of the limitations of current HPV vaccines. In particular, additional protection against other HPV types could eventually eliminate the need for screening. In the meantime, modified screening alternatives will have to be implemented in vaccinat- ed populations, and screening ac- tivities will need to continue for the majority of adult women currently alive, who will not benefit from this extraordinary preventive tool.
Fig. 4.6.4. Differences in human papillomavirus (HPV) genoprevalence between pre- vaccine and post-vaccine populations. * P < 0.05 for difference in percentages between groups. CI, confidence interval; excl, excluding; HR-HPV, high-risk HPV.
Chapter 4.6 • Vaccination 321
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.6
References
1. de Martel C, Ferlay J, Franceschi S et al. (2012). Global burden of cancers at- tributable to infections in 2008: a review and synthetic analysis. Lancet Oncol, 13:607–615. http://dx.doi.org/10.1016/S14 70-2045(12)70137-7 PMID:22575588
2. WHO (2006). WHO/UNICEF coverage estimates 1980–2005. Countries having introduced HepB vaccine and infant HepB3 coverage, 2005. Geneva: WHO.
3. Schiller JT, Castellsagué X, Garland SM (2012). A review of clinical trials of human papillomavirus prophylactic vaccines. Vaccine, 30 Suppl 5:F123–F138. http:// dx.doi.org/10.1016/j.vaccine.2012.04.108 PMID:23199956
4. IARC (2012). Biological agents. IARC Monogr Eval Carcinog Risks Hum, 100B: 1– 441. PMID:23189750
5. Hyams KC (1995). Risks of chronicity fol- lowing acute hepatitis B virus infection: a review. Clin Infect Dis, 20:992–1000. http://dx.doi.org/10.1093/clinids/20.4.992 PMID:7795104
6. Chen CJ, Yu MW, Liaw YF (1997). Epidemiological characteristics and risk factors of hepatocellular carcinoma. J Gastroenterol Hepatol, 12:S294–S308. http://dx.doi.org/10.1111/j.1440-1746.1997. tb00513.x PMID:9407350
7. WHO (2009). Hepatitis B vaccines. Wkly Epidemiol Rec, 84:405 – 419. http:// w w w.w h o.i nt / we r/ 2 0 0 9 / we r 8 4 4 0.p d f PMID:19817017
8. Chang MH, Chen CJ, Lai MS et al.; Taiwan Childhood Hepatoma Study Group (1997). Universal hepatitis B vaccination in Taiwan and the incidence of hepatocel- lular carcinoma in children. N Engl J Med, 336:1855–1859. http://dx.doi.org/10.1056/ NEJM199706263362602 PMID:9197213
9. Chang MH, Shau WY, Chen CJ et al.; Taiwan Childhood Hepatoma Study Group (2000). Hepatitis B vaccination and hepa- tocellular carcinoma rates in boys and girls. JAMA, 284:3040–3042. http://dx. doi.org/10.1001/jama.284.23.3040 PMID: 11122592
10. Chiang CJ, Yang YW, You SL et al. (2013). Thirty-year outcomes of the na- tional hepatitis B immunization program in Taiwan. JAMA, 310:974–976. http:// dx.doi.org/10.10 01/jama.2013.276701 PMID:24002285
11. Chen TW (2013). Paths towards hepati- tis B immunization in Republic of Korea and Taiwan, China. Clin Exp Vaccine Res, 2:76–82. http://dx.doi.org/10.7774/cevr. 2013.2.2.76 PMID:23858397
12. Sun Z, Chen T, Thorgeirsson SS et al. (2013). Dramatic reduction of liver can- cer incidence in young adults: 28 year follow-up of etiological interventions in an endemic area of China. Carcinogenesis, 34:1800–1805. http://dx.doi.org/10.1093/ carcin/bgt007 PMID:23322152
13. Moscicki AB, Schiffman M, Burchell A et al. (2012). Updating the natural history of human papillomavirus and anogenital can- cers. Vaccine, 30 Suppl 5:F24–F33. http:// dx.doi.org/10.1016/j.vaccine.2012.05.089 PMID:23199964
14. Doorbar J, Quint W, Banks L et al. (2012). The biology and life-cycle of human papillo- maviruses. Vaccine, 30 Suppl 5:F55–F70. http://dx.doi.org/10.1016/j.vaccine.2012. 06.083 PMID:23199966
15. Gillison ML, Alemany L, Snijders PJ et al. (2012). Human papillomavirus and diseases of the upper airway: head and neck cancer and respiratory papillomato- sis. Vaccine, 30 Suppl 5:F34–F54. http:// dx.doi.org/10.1016/j.vaccine.2012.05.070 PMID:23199965
16. Chaturvedi AK, Engels EA, Pfeiffer RM et al. (2011). Human papillomavirus and rising oropharyngeal cancer inci- dence in the United States. J Clin Oncol, 29:4294– 4301. http://dx.doi.org/10.1200/ JCO.2011.36.4596 PMID:21969503
17. Westra TA, Rozenbaum MH, Rogoza RM et al. (2011). Until which age should wom- en be vaccinated against HPV infection? Recommendation based on cost-effective- ness analyses. J Infect Dis, 204:377–384. ht tp://dx.doi.org/10.10 93/infdis/jir 281 PMID:21742836
18. Hildesheim A, Herrero R, Wacholder S et al.; Costa Rican HPV Vaccine Trial Group (2007). Effect of human papillomavirus 16/18 L1 viruslike particle vaccine among young women with preexisting infection: a randomized trial. JAMA, 298:743–753. http://dx.doi.org/10.1001/jama.298.7.743 PMID:17699008
19. Tabrizi SN, Brotherton JM, Kaldor JM et al. (2012). Fall in human papillomavirus prevalence following a national vaccina- tion program. J Infect Dis, 206:1645–1651. ht tp://dx.doi.org/10.1093/infdis/jis59 0 PMID:23087430
20. Lehtinen M, Dillner J (2013). Clinical tri- als of human papillomavirus vaccines and beyond. Nat Rev Clin Oncol, 10:400– 410. http://dx.doi.org/10.1038/nrclinonc. 2013.84 PMID:23736648
21. Block SL, Nolan T, Sattler C et al.; Protocol 016 Study Group (2006). Comparison of the immunogenicity and reactogenic- ity of a prophylactic quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in male and female adolescents and young adult women. Pediatrics, 118:2135–2145. http:// d x .d o i.o r g /10.15 42 /p e d s . 2 0 0 6 - 0 4 61 PMID:17079588
22. Markowitz LE, Tsu V, Deeks SL et al. (2012). Human papillomavirus vac- cine introduction – the first five years. Vaccine, 30 Suppl 5:F139–F148. http:// dx.doi.org/10.1016/j.vaccine.2012.05.039 PMID:23199957
23. Kreimer AR, Rodriguez AC, Hildesheim A et al.; CVT Vaccine Group (2011). Proof-of- principle evaluation of the efficacy of fewer than three doses of a bivalent HPV16/18 vaccine. J Natl Cancer Inst, 103:1444– 1451. http://dx.doi.org/10.1093/jnci/djr319 PMID:21908768
24. Herrero R, Quint W, Hildesheim A et al.; CVT Vaccine Group (2013). Reduced prevalence of oral human papillomavirus (HPV) 4 years after bivalent HPV vac- cination in a randomized clinical trial in Costa Rica. PLoS One, 8:e68329. http:// dx.doi.org/10.1371/journal.pone.0068329 PMID:23873171
322
Summary • Early detection and treatment
of cancer requires accessible, high-quality health services with adequate human, financial, and technical resources.
• Population-based screening pro- grammes of appropriate quality can reduce morbidity and mor- tality from breast, cervical, and colorectal cancers. Provided ev- idence-based methods are used that have been tested under local conditions, these programmes can offer a more effective alter- native to early symptomatic de- tection of disease.
• Cervical cancer screening may involve visual inspection of the cervix, cervical sampling for cytology, or testing for human papillomavirus infection.
• Screening tests for colorectal cancer include faecal occult blood testing and faecal im- munochemical testing. Flexible sigmoidoscopy and colonosco- py are also used.
• Screening with mammography permits detection of early-stage breast cancers.
• To maximize the benefit and min- imize harm, screening should be
4.7 4 PREVENTION
Screening – principles Lawrence von Karsa Peter B. Dean Silvina Arrossi Rengaswamy Sankaranarayanan
Ahti Anttila (reviewer) Anthony B. Miller (reviewer)
implemented only in organized, population-based programmes with adequate resources for planning and training, identifica- tion and invitation of the target population, and quality assur- ance, including monitoring and evaluation.
• Lack of governmental commit- ment to provide the requisite sustainable resources can be a serious barrier to successful im- plementation of cancer screen- ing programmes.
Detection of cancer in its early stages and the provision of prompt appropriate treatment are impor- tant elements in cancer control. The aim of early detection is to reduce mortality and other serious conse- quences of advanced disease. This goal may be accomplished if earlier treatment improves life expectancy, locoregional control of disease, and quality of life, and/or permits equally effective therapy with fewer side- effects (Fig. 4.7.1). Key to achieving the potential impact of early detec- tion of cancer is universal access to prompt and effective diagnostic and treatment services. Comprehensive, multidisciplinary quality assurance is essential to maintain an appropriate balance between overall benefit and
inherent risk. The concept of early detection of cancer has evolved since the landmark publication by Wilson and Jungner in 1968 [1]. In recent decades the principles es- tablished in that publication have been extended through experience gained from implementation of pop- ulation-based cancer screening pro- grammes [2]. Due to the complexity of the issues that have impacts on the effectiveness and appropriate- ness of early detection of cancer, this chapter provides an overview of key concepts and sources of in- formation and calls attention to not only the opportunities but also the limitations of cancer screening.
Early detection Early detection of symptomatic cases Early diagnosis is particularly rel- evant for cancers of the breast, cer- vix, mouth, larynx, colon, rectum, and skin. Before evidence-based screening methods became avail- able, early detection was achieved by improving diagnostic and thera- peutic methods and by the educa- tion of health-care professionals and the general public to promote early recognition of symptomatic disease. Some early signs of cancer include lumps, sores that fail to heal, abnor- mal bleeding, persistent indigestion,
Chapter 4.7 • Screening – principles 323
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.7
and chronic hoarseness. Increased awareness of cancer warning signs can have a significant impact on the burden of disease, particularly in those health-care systems in low- and middle-income countries in which most patients are currently diagnosed with very-late-stage ma- lignancy [3]. Universal access to adequate diagnosis and treatment is still lacking in many countries and is a serious barrier to the control of symptomatic disease.
Early detection of asymptomatic cancer In high-income countries, popu- lation-based screening has been shown to be effective in reducing cancer-specific mortality, comple- menting early symptomatic detection of breast, cervical, and colorectal cancers [4–6]. The main objective of screening is to discover latent disease among those who are pre- dominantly asymptomatic and to enable adequate treatment before
the cancer poses a more serious threat to the individual and places an additional burden on the com- munity [1]. Early treatment of in- vasive lesions, including surgical removal of early invasive breast cancer or endoscopic resection of early colorectal cancer, can be less detrimental than treat- ment of symptomatic disease [7,8]. Moreover, detection of premalignant
lesions followed by treatment with cryotherapy or the loop electrosur- gical excision procedure for intra- epithelial cervical neoplasia, or by endoscopic excision of colorectal adenomas, prevents progression to cervical and colorectal cancer, re- spectively [5,8,9]. Randomized trials among people at average risk invited to participate in population-based screening have shown a reduction in cervical and colorectal cancer in- cidence and mortality [6,9].
For cervical cancer, the impact of screening is evident when trends in incidence over decades are com- pared across 38 countries [10]. Strong downward trends occurred in the highest-income countries, whereas no clear changes, includ- ing some increases, were recorded in lower-income settings (Fig. 4.7.2). Due to improvements in survival, the full benefit of mammography screen- ing in terms of the number of breast cancer deaths prevented becomes evident only after a follow-up period of more than two decades. In a large randomized trial, follow-up of 133 065 participants over 29 years showed a 31% reduction in breast cancer mor- tality in the group assigned to screen- ing compared with the control group; furthermore, most of the deaths prevented by screening would have occurred after a follow-up period of more than 10 years [11]. Two recent studies have also provided insight into long-term effects of colorectal cancer screening [12]. One report es- timated that screening colonoscopy
Screening
Earlier treatment
Appearance of symptoms Diagnosis
Start of treatment
Serious consequences
Serious consequences
A
B Serious consequences
Detectable by screening
Diagnosis Start of treatment
C
“Usual care”
Earlier diagnosis by better awareness
Diagnosis
Life expectancy
Improved life expectancy
Significantly improved life expectancy
Start of treatment
Time
Early detection of selected cancers through symptoms or by screening
Fig. 4.7.1. (A) There is usually a delay between the appearance of symptoms and the start of treatment of a cancer. Treatment delays or the unavailability of treatment are serious problems worldwide. (B) Earlier diagnosis and treatment of some cancers due to better awareness of symptoms may increase life expectancy and reduce serious consequences of the disease. (C) Still earlier detection and treatment of some asymp- tomatic cancers through screening can significantly increase life expectancy and further reduce serious consequences of the disease (for methods, see Table 4.7.1). Ideally, the delays between the appearance of symptoms, diagnosis, and start of treatment of cancer should be as short as possible. Organized screening programmes also tend to decrease the interval between diagnosis and start of treatment.
Fig. 4.7.2. Comparison of age-standardized incidence trends of cervical cancer, for ages 30–74 years, between countries of northern Europe and countries of Asia/Africa.
324
would have prevented 40% of colo- rectal cancers over a 22-year pe- riod in 88 902 individuals, while the other found that annual faecal occult blood testing reduced mortality by 32% in a 30-year follow-up of 46 551 participants.
A considerably shorter time pe- riod was sufficient to demonstrate the efficacy of colorectal cancer screening using once-only flexible sigmoidoscopy between the ages of 55 and 64 years. After 11.2 years of follow-up of 170 038 participants in a randomized controlled study, colo- rectal cancer incidence and mortal- ity were reduced by 23% and 31%, respectively, in the screening group compared with the control group [13].
A population-based screening programme for detection of asymp- tomatic disease will also detect cases that are symptomatic, but the detec- tion, on average, will result in earlier treatment for the asymptomatic cases than for the cases of symptomatic dis- ease. It is therefore anticipated that the numbers of detected cancers with a favourable prognosis will be higher in a population-based screening pro- gramme of appropriate quality, and over time the effective diagnosis and treatment of these cases will result in a greater decrease in disease-spe- cific mortality than can be achieved by early detection of symptomatic cancer alone. However, early detec- tion may also lead to diagnosis or pro- longation of illness without improving a patient’s prognosis. Randomized controlled trials are therefore used to eliminate lead-time, length, and se- lection bias in the evaluation of the effect of cancer screening [4].
The evidence-based screening methods currently used in population- based cancer screening programmes are shown in Table 4.7.1. Suitable screening tests for cervical cancer include cervical sampling for conven- tional cytology (Pap smear; Fig. 4.7.3) or liquid-based cytology; testing for human papillomavirus (HPV) infec- tion; and visual inspection of the cervix with acetic acid (Fig. 4.7.4). Screening for breast cancer by mam- mography (Fig. 4.7.5) permits early detection of small breast cancers, most of which have not yet metas- tasized and can be successfully treated, although some of these might
never have presented clinically in the subject’s lifetime (overdiagnosis). Bowel cancer screening using faecal occult blood testing, flexible sigmoid- oscopy, or colonoscopy may lead to the diagnosis of colorectal polyps (Fig. 4.7.6), removal of which during colonoscopy may prevent their pro- gression to colorectal cancer. Early colorectal cancer that can be effec- tively treated by endoscopic removal or surgical excision may also be de- tected. Evidence-based programmes for population-based screening of other major cancers, such as can- cers of the ovary, liver, oesophagus, lung, and prostate, have not yet been established.
For high-risk groups, visual screening (Fig. 4.7.7) has been shown to be efficacious in early detection and prevention of deaths from oral cancer [14]. Results of a large randomized controlled trial in the USA showed that screening people at high risk of lung cancer based on age and smok- ing history with low-dose computed tomography can reduce the number of deaths due to lung cancer [15]. The United States Preventive Services Task Force recently announced a draft
Table 4.7.1. Evidence-based screening methods currently used in national or regional cancer screening programmes
Target cancer Screening method
Breast Mammography
Cervical Cytology (conventional and liquid-based) Human papillomavirus testing Visual inspection of the cervix with acetic acid
Colorectal Faecal occult blood testing or faecal immunochemical testing Flexible sigmoidoscopy Colonoscopy
Fig. 4.7.3. Pap smear suggestive of high-grade precancerous lesion: cervical intraepi- thelial neoplasia grade 3 (CIN3). Inflammatory smear contains many parabasal cells with enlarged nuclei with irregular chromatin (black arrow) and some cells with eosinophilic cytoplasm (purple arrow).
Chapter 4.7 • Screening – principles 325
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.7
recommendation for annual screening for lung cancer using this procedure in people at high risk [16].
Determinants of harm and benefit Screening should be conducted only after careful consideration of both harms and benefits. Knowledge has developed about health economic and cost–benefit analysis predicated on the potential health benefits and risks in addition to the financial costs of cancer screening for the people affected [17–19]. Before screen- ing programmes are launched, an analysis of the expected costs and the benefit should be performed us- ing data collected in feasibility and pilot studies in the relevant health- care environment. These analyses should be repeated at intervals in established programmes, typically every 10 years, and before adopt- ing major modifications of existing protocols. The methods applied in cost–benefit analysis are complex, and in interpreting the results, the underlying assumptions as well as the balance between the main ben- efits and harms of screening should be kept in mind.
Despite the considerable poten- tial of cervical and colorectal cancer screening to prevent invasive cancer through detection and treatment of precancerous lesions, the key
measure of benefit for any such screening programme is mortal- ity reduction, rather than increased detection of early-stage disease. A secondary benefit of earlier diagnosis arises because smaller cancers can often be treated with less debilitating procedures and medications.
The primary harm from screening is morbidity and mortality from the procedures for detection and diagno- sis, and the side-effects of treatment
prompted by screening. Exposure to these risks in the absence of any direct health benefit is of particular concern. Such a situation may arise due to false-positive tests or procedures, or from overdiagnosis, i.e. diagnosis through screening of cancers that would not have been detected symp- tomatically during the lifetime of an individual. Unfortunately, overdiag- nosed cancers cannot be individually identified; the likely number of such cases may be estimated after many years of follow-up in a randomized screening trial where screening was never offered to the relevant controls. A related potential harm of screening is overtreatment, which is not limited to those cases that are overdiag- nosed. Overtreatment may occur if less-intensive treatment protocols appropriate to early-stage lesions are not available or are not followed.
Quality assurance plays an im- portant role in keeping overdiagnosis and overtreatment in an appropriate range. Studies in Europe address- ing overtreatment in breast cancer screening have demonstrated that mastectomy rates decreased mark- edly after the introduction of popula- tion-based mammography screening
Fig. 4.7.4. Outcomes of visual inspection of the cervix with acetic acid: (A) negative, (B) positive. The circumorificial acetowhite lesion on the right (B) is suggestive of cervical intraepithelial neoplasia.
Fig. 4.7.5. (A) A normal oblique view mammogram of the left breast of an asymptomatic 57-year-old woman. (B) A small invasive cancer detected at screening 2 years later. This cancer could not be detected with palpation even after it had been detected with mammography.
A B
326
[20,21]. Two recent reviews estimated that for every one or two overdiag- nosed cases, at least one death due to breast cancer was avoided, a balance between benefit and harm considered to be appropriate [22,23].
The potential for overdiagnosis is substantial when screening for prostate cancer by determining se- rum prostate-specific antigen levels [24]. Whereas data from the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial in the USA [25], unlike results from the European Randomized Study of Screening for Prostate Cancer [26], have not indi- cated a disease-specific mortality reduction, both trials suggest that systematic screening would result in significant levels of overdiagno- sis. Due to uncertainties about the benefits and the risks, population- based prostate cancer screening is not supported by currently available evidence.
Since estimation of overdiagnosis requires the application of complex statistical methods,all such estimates are subject to uncertainty. Whereas cancer-registry-based studies tend to
be imprecise and underestimate the benefits of screening, calculations of the mortality benefit based on precise individual data tend to be highly ac- curate. Although many people con- sider the benefit of an averted cancer death to greatly outweigh the risks of overdiagnosis and overtreatment, strict quality assurance practices must be maintained to minimize all risks associated with screening.
Organization of screening People eligible for population-based screening are defined by age and/or gender, as well as the cancer bur- den for the relevant tumour type. In most countries breast cancer screening is offered to normal-risk women from the age of 40–50 years and extending to the age of 70– 75 years, typically at 2-year inter- vals. Cervical cancer screening is generally offered to women from the age of 25–30 years and extending to the age of 60–65 years. The recom- mended interval commonly varies between 3 and 5 years. In low- and middle- income countries, cervical
screening may be provided in a nar- rower age range or only once or a few times in a lifetime. Colorectal cancer screening is generally of- fered to women and men from the age of 50 years and extending to the age of 74 years. Depending on the screening protocol, which may use faecal occult blood testing, fae- cal immunochemical testing, flexible sigmoidoscopy, or colonoscopy, the recommended interval commonly varies between 2 and 10 years, and endoscopic screening may be performed only once in a lifetime [27,28].
Screening of high-risk groups is generally not considered a pop- ulation-based strategy since only a small portion of the total population is affected [3]. For evidence-based screening of high-risk groups, a pro- grammatic approach can be recom- mended, provided the quality and the cost-effectiveness of the screen- ing process are assured.
Organized programmes For a programme to be considered an organized screening programme, a public cancer screening policy documented in a law, or an official regulation, decision, directive, or recommendation is required. The policy should define, as a minimum, the screening protocol and repeat interval, and determinants of eligi- bility for screening. The screening examinations should be financed by public sources (apart from a pos- sible co-payment in high-income countries). Provision of screening in organized programmes is recom- mended because these include an administrative structure responsible for service delivery, quality assur- ance, and evaluation [29].
The rules and regulations appli- cable to an organized programme generally provide for a team at the national or regional level that is responsible for implementa- tion, i.e. for coordinating the de- livery of the screening services, maintaining the requisite quality, and reporting on performance and results. Such organizational and managerial elements generally
Fig. 4.7.6. A colonoscopy performed as part of the Lampang Province colorectal cancer screening programme in Thailand. Inset at lower right shows a large bowel polyp being removed during the colonoscopy.
Chapter 4.7 • Screening – principles 327
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.7
provide for supervision and monitor- ing of most steps in the screening process, as well as comprehensive guidelines and rules defining stan- dard operating procedures. In addi- tion, a quality assurance structure is required, and a means of ascer- taining the population burden of the disease must be available so that the programme can be evaluated [4].
Population-based programmes Population-based screening pro- grammes identify and individually invite each person in the eligible population to attend each round of screening. Individual invitations aim to give each person in the eli- gible population an equal chance of benefiting from screening and to thereby reduce health-care inequalities [29]. The invitations should be supported by effective communication for groups with lim- ited access to screening, such as less-advantaged socioeconomic groups. This communication, in turn, should support an informed deci- sion about participation, based on information about the risks and ben- efits of screening [30,31]. Invitations indicating a date, time, and place for the screening examination have been most effective in high-resource settings. These invitations have usu- ally been sent by letter, often followed up with reminders to non-respond- ers. Recent experience with call centres and electronic messaging
via mobile phones suggests that such methods of communication warrant further investigation [30].
Opportunistic programmes “Opportunistic” screening is a less cost-effective form of screening and is less amenable to quality as- surance than is population-based screening. Such screening relies on a health-care provider taking the initiative to offer screening or to en- courage individuals to participate in a screening programme, or to under- take screening outside the context of any programme (so-called wild screening). In high-resource set- tings, opportunistic screening has been associated with lower overall impact of screening and lower par- ticipation rates [32]. In addition, the total volume of resources consumed tends to be significantly higher with opportunistic screening. This leads to overuse of screening resources by part of the target population and underuse by, or lack of coverage of, the rest of the population [33,34].
Importance of quality assurance The importance of quality assur- ance in cancer screening goes beyond the need to ensure that any medical intervention is per- formed adequately, efficiently, and with minimum risk and maximum benefit. At any given time, only a small number of people invited to attend population-based screening
(usually < 1%) will have undetected cancer or precancerous lesions that can be treated effectively. Therefore, only a small number of people can derive a direct health benefit from attending screening. However, all participants are exposed to the risks of screening, and although these are slight, the risks add up. To achieve the benefit of cancer screening, quality must be optimal at every step in the screening process, in- cluding identification and personal invitation of each eligible individual; performance of the screening test, examination, or procedure; diagnos- tic work-up of people with detected abnormalities; and, when neces- sary, treatment, surveillance, and aftercare (Fig. 4.7.8) [6]. In practice, the process of screening is much more complex than the schematic diagram in Fig. 4.7.8 suggests. For example, in a quality-assured colo- rectal cancer screening programme using a faecal occult blood testing kit, receipt of the test result indicat- ing whether diagnostic work-up is warranted involves at least five dif- ferent activities. Diagnostic work-up and clinical management of the le- sions detected in screening require multidisciplinary teamwork and can be complicated. That is why orga- nized, quality-assured screening programmes follow comprehen- sive quality assurance guidelines with evidence-based standards, procedures, and protocols of best practice that are continuously being improved [6,7,35].
Evaluation of the impact of a screening programme on cancer mortality is a long-term objective of quality assurance. The actual effect cannot be exactly quantified because mortality in the screened population would have to be compared with mor- tality in the same population in the absence of screening. To estimate the effect, various study designs have been used, such as comparing screened and matched unscreened populations, time trends in mortality, case–control studies, and modelling studies. However, changes attribut- able to improved awareness and treatments cannot be completely
Fig. 4.7.7. Early-stage oral cancer detected by visual screening.
328
disentangled from the outcomes of screening programmes. Evidence on the effectiveness of population-based cancer screening is predominantly available from programmes in high- income countries.
To fully evaluate the impact of a screening programme, the potential harms mentioned above must also be determined, such as those due to false-positive tests and overdi- agnosis. Since the full impact of a screening programme will not be known for many years after its ini- tiation, it is also necessary to pro- actively monitor and continuously improve performance of all activi- ties in the screening process from the outset, i.e. long before the im- pact is actually discernible. That will
permit the timely recognition of steps that must be taken to continuously improve communication, organiza- tion, professional performance, and equipment.
Low coverage of the target popu- lation substantially limits the effec- tiveness and the cost-effectiveness of any programme. Health-care pro- fessionals, responsible authorities, and governments should also be aware of this and other constraints that may limit the achievable impact of cancer screening programmes. A common pitfall is the adoption of non-validated working models and methods. Of prime importance is the availability and sustainability of adequate human, financial, and tech- nical resources to ensure that the
programme achieves an appropriate balance between benefit and harm.
Outlook It may take several years to imple- ment a population-based cancer screening programme, from the beginning of planning to completion of roll-out across an entire country or region. Determinants of successful implementation of population-based cancer screening programmes have been developed for application to screening for any target cancer. Insights into the implementation of such programmes are provided in Chapter 4.8, which illustrates the need for comprehensive quality as- surance in the establishment and de- livery of multidisciplinary screening services at all resource levels. It also shows the importance of universal access to prompt and effective diag- nostic and treatment services, which is a common goal in all attempts to improve cancer control through early detection.
References
1. Wilson JMG, Jungner G (1968). Principles and Practice of Screening for Disease. Geneva: WHO. Available at http://whqlib doc.who.int/php/WHO_PHP_34.pdf/.
2. Lansdorp-Vogelaar I, von Karsa L (2012). European guidelines for quality assur- ance in colorectal cancer screening and diagnosis. First edition - Introduction. Endoscopy, 44 Suppl 3:SE15–SE30. http:// dx.doi.org/10.1055/s-0032-1308898 PMID: 23012118
3. WHO (2007). Cancer Control: Knowledge into Action. WHO Guide for Effective Programmes. Module 3: Early Detection. Geneva: WHO. Available at http://www.who. int/cancer/publications/cancer_control_ detection/en/index.html.
4. IARC (2002). IARC Handbooks of Cancer Prevention, Vol. 7: Breast Cancer Screening. Lyon: IARC.
5. IARC (2005). IARC Handbooks of Cancer Prevention, Vol. 10: Cervix Cancer Screening. Lyon: IARC.
6. von Karsa L, Patnick J, Segnan N et al.; European Colorectal Cancer Screening Guidelines Working Group (2013). European guidelines for quality assurance in colorectal cancer screening and diagno- sis: overview and introduction to the full sup- plement publication. Endoscopy, 45:51–59. http://dx.doi.org/10.1055/s-0032-1325997 PMID:23212726
7. Perry N, Broeders M, de Wolf C et al. (2008). European guidelines for qual- ity assurance in breast cancer screening and diagnosis. Fourth edition - Summary document. Ann Oncol, 19:614–622. http:// d x .d o i.o r g /10 .10 9 3 /a n n o n c /m d m 4 81 PMID:18024988
8. Segnan N, Patnick J, von Karsa L, eds (2010). European Guidelines for Quality Assurance in Colorectal Cancer Screening and Diagnosis - First Edition. Luxembourg: European Commission, Publications Office of the European Union.
9. Sankaranarayanan R, Nene BM, Shastri SS et al. (2009). HPV screening for cervi- cal cancer in rural India. N Engl J Med, 360:1385–1394. http://dx.doi.org/10.1056/ NEJMoa0808516 PMID:19339719
10. Vaccarella S, Lortet-Tieulent J, Plummer M et al. (2013). Worldwide trends in cervi- cal cancer incidence: Impact of screen- ing against changes in disease risk fac- tors. Eur J Cancer, 49:3262–3273. http:// dx.doi.org/10.1016/j.ejca.2013.04.024 PMID:23751569
Fig. 4.7.8. The cancer screening process.
Chapter 4.7 • Screening – principles 329
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.7
11. Tabar L, Vitak B, Chen TH et al. (2011). Swedish two-county trial: impact of mam- mographic screening on breast cancer mortality during 3 decades. Radiology, 260:658–663. http://dx.doi.org/10.1148/ radiol.11110469 PMID:21712474
12. Levin TR, Corley DA (2013). Colorectal- cancer screening – coming of age. N Engl J Med, 369:1164–1166. http://dx.doi.org/ 10.1056/NEJMe1308253 PMID:24047066
13. Atkin WS, Edwards R, Kralj-Hans I et al. (2010). Once-only flexible sigmoidoscopy screening in prevention of colorectal can- cer: a multicentre randomised controlled trial. Lancet, 375:1624–1633. http://dx.doi. org /10.1016/S0140 - 673 6(10) 6 0551-X PMID:20430429
14. Sankaranarayanan R, Ramadas K, Somanathan T et al. (2013). Long term ef- fect of visual screening on oral cancer in- cidence and mortality in a randomized trial in Kerala, India. Oral Oncol, 49:314–321. http://dx.doi.org/10.1016/j.oraloncology. 2012.11.004 PMID:23265945
15. Church TR, Black WC, Aberle DR et al.; National Lung Screening Trial Research Team (2013). Results of initial low-dose computed tomographic screening for lung cancer. N Engl J Med, 368:1980–1991. http://dx.doi.org/10.1056/NEJMoa1209120 PMID:23697514
16. U.S. Preventive Services Task Force (2013). Screening for Lung Cancer: Draft Recommendation Statement. AHRQ Publication No. 13-05196-EF-3. Available at http://www.uspreventiveservicestask force.org/uspstf13/lungcan/lungcandraf trec.htm.
17. Wilschut JA, Habbema JD, van Leerdam ME et al. (2011). Fecal occult blood testing when colonoscopy capacity is limited. J Natl Cancer Inst, 103:1741–1751. http://dx.doi. org/10.1093/jnci/djr385 PMID:22076285
18. de Kok IM, van Rosmalen J, Dillner J et al. (2012). Primary screening for human papillomavirus compared with cytology screening for cervical cancer in European settings: cost effectiveness analysis based on a Dutch microsimulation model. BMJ, 344:e670. http://dx.doi.org/10.1136/bmj.e6 70 PMID:22391612
19. Heijnsdijk EA, Wever EM, Auvinen A et al. (2012). Quality-of-life effects of prostate- specific antigen screening. N Engl J Med, 367:595–605. http://dx.doi.org/10.1056/NEJ Moa1201637 PMID:22894572
20. Paci E, Duffy SW, Giorgi D et al. (2002). Are breast cancer screening pro- grammes increasing rates of mastecto- my? Observational study. BMJ, 325:418. http://dx.doi.org/10.1136/bmj.325.7361.418 PMID:12193357
21. Lawrence G, Kearins O, Lagord C et al. (2011). The Second All Breast Cancer Report. London: National Cancer Intelligence Network. Available at http://www.ncin.org. uk/view.aspx?rid=612
22. Paci E; EUROSCREEN Working Group (2012). Summary of the evidence of breast cancer service screening outcomes in Europe and first estimate of the benefit and harm balance sheet. J Med Screen, 19 Suppl 1:5–13. http://dx.doi.org/10.1258/ jms.2012.012077 PMID:22972806
23. Marmot MG, Altman D, Cameron D et al.; Independent UK Panel on Breast Cancer Screening (2012). The Benefits and Harms of Breast Cancer Screening: An Independent Review. Cancer Research UK and the Department of Health (England). Available at www.cruk.org/breast screeningreview.
24. Draisma G, Etzioni R, Tsodikov A et al. (2009). Lead time and overdiagnosis in prostate-specific antigen screening: im- portance of methods and context. J Natl Cancer Inst, 101:374–383. http://dx.doi. org/10.1093/jnci/djp001 PMID:19276453
25. Andriole GL, Crawford ED, Grubb RL 3rd et al.; PLCO Project Team (2012). Prostate cancer screening in the ran- domized Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial: mortality results after 13 years of follow-up. J Natl Cancer Inst, 104:125–132. http://dx.doi. org/10.1093/jnci/djr500 PMID:22228146
26. Schröder FH, Hugosson J, Carlsson S et al. (2012). Screening for prostate cancer decreases the risk of develop- ing metastatic disease: findings from the European Randomized Study of Screening for Prostate Cancer (ERSPC). Eur Urol, 62:745–752. http://dx.doi.org/10.1016/j.eur uro.2012.05.068 PMID:22704366
27. von Karsa L, Anttila A, Ronco G et al. (2008). Cancer Screening in the European Union: Report on the Implementation of the Council Recommendation on Cancer Screening - First Report. Luxembourg: European Communities. Available at http:// ec.europa.eu/health/ph_determinants/ge netics/documents/cancer_screening.pdf.
28. International Cancer Screening Network (2013). Cancer Sites. National Cancer Institute. Available at http://appliedresearch. cancer.gov/icsn/sites.html.
29. Karsa LV, Lignini TA, Patnick J et al. (2010). The dimensions of the CRC problem. Best Pract Res Clin Gastroenterol, 24:381–396. http://dx.doi.org/10.1016/j.bpg.2010.06.004 PMID:20833343
30. Austoker J, Giordano L, Hewitson P et al. (2012). European guidelines for quality as- surance in colorectal cancer screening and diagnosis. First edition - Communication. Endoscopy, 44 Suppl 3:SE164–SE185. http://dx.doi.org/10.1055/s-0032-1309809 PMID:23012120
31. von Karsa L (1995). Mammography screening – comprehensive, population- based quality assurance is required! [in German] Z Allgemeinmed, 71:1863–1867.
32. Malila N, Senore C, Armaroli P (2012). European guidelines for quality assur- ance in colorectal cancer screening and diagnosis. First edition - Organisation. Endoscopy, 44 Suppl 3:SE31–SE48. http://dx.doi.org/10.1055/s-0032-1309783 PMID:23012121
33. Arbyn M, Anttila A, Jordan J et al. (2010). European guidelines for quality assurance in cervical cancer screening. Second edi- tion - Summary document. Ann Oncol, 21:448– 458. http://dx.doi.org/10.1093/ann onc/mdp471 PMID:20176693
34. Anttila A, von Karsa L, Aasmaa A et al. (2009). Cervical cancer screening policies and coverage in Europe. Eur J Cancer, 45:2649–2658. http://dx.doi.org/10.1016/j. ejca.2009.07.020 PMID:19699081
35. Cancer Care Ontario (2012). Colorectal cancer screening. Surveillance guide- lines. Available at https://www.cancer care.on.ca/pcs/screening/coloscreening/ cccstandardsguidelines/.
Websites International Cancer Screening Network: http://appliedresearch.cancer.gov/icsn/
Screening Group, Section of Early Detection and Prevention, IARC: http://www.iarc.fr/en/ research-groups/SCR/index.php
330
Summary • Global implementation of popu-
lation-based cancer screening reveals heterogeneity, and broad knowledge is restricted to three cancer types: breast, cervical, and colorectal cancer.
• In many high-income countries and some middle-income coun- tries, population-based screen- ing programmes for breast and cervical cancer have been estab- lished for decades, or shorter pe- riods, and some have achieved significant reductions in cancer- specific mortality. Colorectal cancer screening programmes have been introduced more recently.
• There are few population-based screening programmes in low- and middle-income countries.
• Screening programmes require political commitment, engage- ment of civil society, competent oversight, and adequate, sus- tainable resources.
• Programme management should have the authority to coordinate all activities essential for provi- sion of screening services, in- cluding training, documentation, monitoring, and other aspects of quality assurance.
4.8 4 PREVENTION
Screening – implementation Lawrence von Karsa You-Lin Qiao Kunnambath Ramadas Namory Keita Silvina Arrossi Peter B. Dean Nada Al Alwan Rengaswamy Sankaranarayanan
Ahti Anttila (reviewer) Anthony B. Miller (reviewer)
• Programmes should be imple- mented gradually, tailoring ex- pansion of the screening service to the capacity of the health-care system and with optimal accessi- bility for less-advantaged groups in the target population.
Implementation of quality-assured population-based cancer screen- ing programmes affects very large numbers of people in the eligible age range (see Chapter 4.7). The pro- cess requires many years, beginning with the development and testing of a comprehensive quality assurance system to ensure the provision of cost-effective, affordable, and ac- ceptable services for the entire target population. Carefully planned, long- term investment of time and financial resources is justified because the impact of a population-based pro- gramme goes far beyond the ser- vices provided. Screening can stim- ulate health systems development and raises the level of awareness of cancer symptoms among health- care professionals and the general public. A quality-assured population- based screening programme raises the standards of cancer diagnosis and treatment throughout the medi- cal community. Professionals trained to meet the standards of the screen- ing programme may also care for
patients outside the programme. These factors, in combination, lead to an overall improvement in the level of cancer care [1].
Criteria for successful implementation of screening Successful implementation of pop- ulation-based cancer screening programmes requires expertise in areas relevant to other prevention activities, such as motivation, com- munication, and reinforcement of efforts designed to prevent cancer [2]. The ability to mobilize large num- bers of health-care professionals, other stakeholders, and the target population itself in collective actions focused on a common goal is cru- cial to the success of any screening programme [3].
Programme implementation be- gins with a planning and feasibility phase, which includes large-scale pilot studies to evaluate whether screening conducted as planned fulfils the key performance targets and is likely to be cost-effective. It may take several years to collect suf- ficient data for valid conclusions to be drawn about routine implementa- tion and to make any necessary ad- justments to the screening protocol warranted by the pilot results. Pilot studies are also needed to provide information on the cost-effective- ness of screening in the context of
Chapter 4.8 • Screening – implementation 331
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.8
an individual country. After success- ful pilot studies, gradual, quality- controlled roll-out of the programme across the country begins, with the health-care system controlling the pace of programme expansion, with particular reference to specialized training of staff and investment in in- frastructure. The entire process of programme implementation rarely takes less than 10 years [1].
Certain conditions are essen- tial to assure quality in population- based screening programmes. These include long-term political commitment and competent over- sight, adequate and sustainable resources, engagement of the com- munity, international collaboration, and autonomous programme man- agement with effective coordination of all activities, particularly commu- nication and training, organizational development, and control of pro- gramme resources [4]. Programme resources include a dedicated budget and staff, computerized in- formation systems, and registries for cancer, population, and screen- ing, all based on individual data [5]. Successful implementation of effective screening programmes
requires national investment [6]. In a fully established programme, the proportion of the expenditure devoted to quality assurance in a high-resource setting should be no less than 10–20% [7]. A substantial proportion of these resources are required for well-organized informa- tion systems, such as those used by cancer and screening registries, to provide reports on test results for participants and health profes- sionals, to confirm that appropriate action follows a positive screening test, and to monitor performance and evaluate the impact of any screening programme [4].
Setting priorities in implementing screening programmes Given the substantial resources required for successful implemen- tation of population-based cancer screening programmes, efforts to improve early detection of cancer should be integrated into national comprehensive cancer control plans that establish overall priorities in the health-care agenda and take into ac- count all relevant activities, such as
national or regional programmes for reproductive health [8,9].
In most high-income countries, the burden of cancer has resulted in coordinated efforts to implement population-based screening for all of the tumour types for which evi- dence-based methods are currently established (breast, cervical, and colorectal cancer). In other coun- tries, due to the comparatively long period required to successfully es- tablish population-based screening programmes, time trends in cancer incidence and mortality should be taken into account when prioritizing potential target cancers for screen- ing [1]. On this basis, cancers cur- rently suitable for screening in low- and middle-income countries include, for example, cervical can- cer in sub-Saharan Africa, India, and Latin America; breast cancer in Latin America, the Middle East, and some countries in Asia; and colorec- tal cancer in countries in transition, such as Brazil and China. Before a commitment is made to the roll- out of a screening programme, the potential cost-effectiveness of the programme in the country should be taken into account. If pilot studies indicate that the cost per year of life saved by a given intervention is less than the per capita gross national product, a screening programme can be considered to be highly cost-ef- fective in that setting [10].
The same principles of qual- ity assurance for population-based screening apply to screening of high-risk groups or to early detection and treatment of symptomatic dis- ease where significant proportions and numbers of patients are diag- nosed with very-late-stage disease. Priority should be given to providing adequate resources for diagnosis and treatment of currently increas- ing levels of symptomatic disease, for example through training of com- petent staff. When access to prompt diagnosis and treatment is available, alternative strategies of early detec- tion can be considered in countries where low incidence rates do not yet justify population-based screen- ing programmes for asymptomatic
Fig. 4.8.1. Immunochemical faecal occult blood testing kit used in the Lampang Province colorectal cancer screening programme in Thailand.
332
people. In Iraq, for example, recent surveys show significant knowledge gaps in the educated population about the relative importance of breast cancer in the Iraqi population and the performance of breast self- examination, suggesting a potential to increase early detection of breast cancer through better awareness [11]. Morocco is currently establish- ing universal access to comprehen- sive cancer diagnosis, treatment, and palliative care according to the 2010–2019 national cancer control plan. This includes a nationwide early detection programme for clini- cally detectable breast cancer [12].
Screening programmes in high-income countries The first population-based can- cer screening programmes were launched in middle- and high-income countries in the 1960s to 1980s, for cervical cancer screening. These programmes were based on con- ventional cytology, and many led to reductions of 50–80% in cervical cancer mortality within two to three decades of implementation [13]. In recent years, primary cervical cancer
screening by human papillomavirus (HPV) testing has been established as more accurate and effective. HPV testing is therefore expected to be- come the preferred screening test for cervical cancer in the medium and long term [14,15].
The first population-based breast screening programmes based on mammography were initiated in Europe, Canada, and Australia in the late 1980s after randomized controlled trials showed the effi- cacy of screening [16,17]. By 2012, population-based breast screening programmes were running or be- ing established in 24 countries in the European Union and in several countries in Asia and the Americas. Screen-film mammography is be- ing increasingly replaced by digi- tal mammography [18]. Now that screening has been performed for more than two decades in several population-based programmes in Europe, methodologies used to estimate the impact of screening and the level of overdiagnosis have been evaluated using data from ser- vice screening programmes [19]. Registry studies analysing popula- tion breast cancer mortality rates
over time underestimate the impact of screening on breast cancer mor- tality because of their inability to ex- clude deaths in women with breast cancer diagnosed before invitation to screening, and due to the inabil- ity to discriminate between data of screened and unscreened women during the gradual implementation of a screening programme in a country or region [20].
Estimates of the impact of ser- vice screening in Europe have re- cently been obtained from a review of all available observational studies from which sufficient longitudinal in- dividual data were available, directly linking a woman’s screening history to her cause of death. The authors considered that the most reliable es- timates of reduction in breast cancer mortality were 25–31% for women invited for screening and 38–48% for women actually screened [21].
Reviews of all randomized clinical trials that evaluated annual or bien- nial guaiac-based faecal occult blood testing indicate that trial participants allocated to screening had a 16% lower relative risk of colorectal cancer mortality [22]. Population-based colo- rectal cancer screening programmes in developed countries are recent, and are still evolving in most coun- tries. Acceptable and desirable levels of performance have recently been developed for the population-based programmes in Europe based on re- sults achieved in randomized trials
Fig. 4.8.2. Women lining up at Kanungu Health Centre IV in Uganda to receive HIV counselling and cervical cancer screening, and to learn more about available family planning services.
Fig. 4.8.3. Augusta Victoria Hospital in Jerusalem operates this mobile mammog- raphy unit, which strives to identify women with breast cancer at an early stage so that they may be referred to the hospital for treatment, with better outcomes.
Chapter 4.8 • Screening – implementation 333
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.8
and routine programmes [23]. Efforts are under way to improve programme performance in many countries [24].
Screening programmes in upper-middle-income countries Opportunistic, large-scale cervical cancer screening has been conduct- ed in some upper-middle-income countries for several years. The resulting impact on cervical cancer incidence and mortality has been limited, due to poor coverage and lack of quality assurance in cytol- ogy screening, suboptimal adher- ence by screen-positive women to further diagnosis and treatment, and lack of information systems to moni- tor progress and assess impact. In Latin America, this has led to initia- tives to reassess the approaches to cervical cancer control, taking into account opportunities to improve organization and quality assurance, as well as the introduction of HPV testing for cervical cancer screen- ing, and HPV vaccination [25]. Self- sampling has also been investigat- ed and shows potential to increase cervical screening effectiveness by increasing participation, especially in women who are difficult to reach or who do not regularly attend cervi- cal cancer screening [26]. In recent years, many of these programmes have been reorganized, and HPV testing is being integrated into rou- tine screening in a phased manner in countries such as Argentina and Mexico. Sporadic and opportunistic mammography screening is wide- spread, but population-based breast cancer screening programmes have yet to evolve in many upper-middle- income countries. Colorectal cancer screening is less widespread; a na- tional programme exists in Uruguay.
Screening programmes in lower-middle-income and low-income countries Cancer screening programmes are operational in very few lower-middle- income and low-income countries in Africa, Asia, Central America, and the Caribbean, despite the high risk
of cervical cancer and the increasing incidence of breast cancer in these regions. Limited resources and inad- equate development of health ser- vices currently preclude introducing screening programmes in most of these countries. The challenges to introducing cytology screening and to improving treatment of screen- positive women, as well as the lim- ited impact of Pap smear screening in upper-middle-income countries, have prompted the search for alter- native screening methods and algo- rithms for cervical cancer control in low-resource settings. HPV testing and visual inspection of the cervix with acetic acid have been evalu- ated as alternative methods, and single-visit approaches, involving diagnosis and treatment of screen- positive women in the same sitting as screening, have been explored to improve compliance with treatment. Recent results from such studies have prompted the introduction of visual inspection of the cervix with acetic acid screening programmes in Bangladesh [27], Tamil Nadu state in India, Thailand [28], and Zambia [29], as well as demonstration programmes in 43 counties of 31
provincial administrative areas in China and in several countries in sub-Saharan Africa (Fig. 4.8.4) and Central America. Implementation of acetic acid-based screening may improve development of screening infrastructure, which may, in turn, facilitate the introduction of HPV screening when simple and af- fordable viral tests become widely available.
Limited resources and inad- equate development of health ser- vices preclude introducing mam- mography screening programmes in most lower-middle-income and low- income countries. Improving breast awareness may facilitate earlier clin- ical diagnosis among symptomatic women in such settings, but these efforts may be counterproductive if the women cannot obtain timely di- agnostic and therapeutic services. For the same reason, systematic ap- proaches to early detection based on breast examination and imaging methods are required in countries that have an increasing burden of breast cancer but that currently lack adequate diagnostic and therapeu- tic services [30,31]. Population- based colorectal cancer screening
Fig. 4.8.4. Women waiting to be screened during a cervical cancer prevention campaign organized by IARC and Conakry University Hospital, in the Republic of Guinea.
334
does not exist in any low-income or lower-middle-income country, with the exception of Thailand, where a pilot colorectal cancer screening programme has been introduced as a forerunner to the phased national expansion.
Outlook Breast, cervical, and colorectal cancer screening programmes have been improved globally through re- search in terms of quality inputs, efficiency, and effectiveness. New research findings have catalysed the planning and organization of new screening programmes in some countries [32]. In the foreseeable fu- ture, HPV vaccination programmes seem likely to reduce the prevalence of this viral infection and related pre- cancerous lesions, necessitating more sensitive and objective screen- ing tests [33]. Research has indicat- ed the efficacy of mammography and faecal occult blood screening and paved the way for population-based screening programmes. Screening approaches for other tumour types, such as lung, ovarian, oesophageal, stomach, and prostate cancer, are currently being investigated in re- search settings (Fig. 4.8.5), and in the case of gastric cancer also in the context of a population-based cancer screening programme in the Republic of Korea [34]. Unless these initiatives prove their efficacy, feasibility, and cost-effectiveness in those settings, population-based programmes are unlikely to be es- tablished and widely implemented.
Population-based screening programmes for breast, cervical, and colorectal cancer have been introduced as part of cancer con- trol in many high-income countries. Population-based cancer screening programmes do not exist in most
low- and middle-income countries, where cancers are mainly detected through very-late-stage symptoms and many are not reported. Recent research may lead to new approach- es to early detection and treatment using improved awareness of symp- tomatic disease and population- based screening of asymptomatic people.
Success in decreasing the bur- den of cancer will depend on the acceptance by the population of a screening programme that is an- chored in a comprehensive and well-structured cancer control pro- gramme, and on the capacity of the health-care system to make efficient
and appropriate diagnostic and ther- apeutic services universally avail- able for the cases detected early. Provision of adequate resources will be decisive.
International cooperation can enable countries to avoid common pitfalls in the implementation of screening programmes and other early detection programmes, and to share knowledge about successful methods and approaches. Sharing of expertise may enable a country to implement programmes more suc- cessfully and to avoid unnecessary expenditure and delay.
Fig. 4.8.5. A woman being screened for stomach cancer at the Osaka Cancer Prevention and Detection Center, in Japan, in research directed towards the development of a population-based protocol for this tumour type.
Chapter 4.8 • Screening – implementation 335
4 P
R E
V E
N T
IO N
C H
A P
T E
R 4
.8
References
1. Karsa LV, Lignini TA, Patnick J et al. (2010). The dimensions of the CRC problem. Best Pract Res Clin Gastroenterol, 24:381– 396. http://dx.doi.org/10.1016/j.bpg.2010. 06.004 PMID:20833343
2. Hill D, Dixon H (2010). Achieving behav- ioural changes in individuals and popula- tions. In: Elwood JM, Sutcliffe SB, eds. Cancer Control. Oxford: Oxford University Press, pp. 43–61.
3. Hanleybrown F, Kania J, Kramer M (2012). Channeling change: making collective impact work. Stanford Social Innovation Review, 1–8. Available at http://www.ssire view.org/blog/entry/channeling_change_ making_collective_impact_work/.
4. Lynge E, Törnberg S, von Karsa L et al. (2012). Determinants of successful im- plementation of population-based cancer screening programmes. Eur J Cancer, 48:743–748. http://dx.doi.org/10.1016/j.ejca. 2011.06.051 PMID:21788130
5. von Karsa L, Arrossi S (2013). Development and implementation of guidelines for qual- ity assurance in breast cancer screening: the European experience. Salud Publica Mex, 55:318–328. PMID:23912545
6. National Colorectal Screening Programme (2011). International Peer Review Panel Report of Quality Assurance Standards, 10 –11 March 2011. NCSS/PUB/Q-3 Rev 2. Dublin, Ireland: National Cancer Screening Service. Available at http://www.cancer screening.ie/publications/Colorectal_Peer_ Review_Report_Rev_2.pdf.
7. von Karsa L, Suonio E, Lignini TA et al., eds (2012). Current Status and Future Directions of Breast and Cervical Cancer Prevention and Early Detection in Belarus. Cancer Control Assessment and Advice Requested by the Belarus Ministry of Health. Report of Expert Mission to Minsk, Belarus, 15–18 February 2011. Lyon: IARC. Available at http://www.iarc.fr/en/publications/pdfs- online/wrk/wrk6/Belarus_Report.pdf.
8. WHO (2006). Cancer Control: Knowledge into Action. WHO Guide for Effective Programmes. Module 1: Planning. Geneva: WHO. Available at http://www.who.int/can cer/publications/cancer_control_planning/ en/index.html.
9. WHO (2007). Cancer Control: Knowledge into Action. WHO Guide for Effective Programmes. Module 3: Early Detection. Geneva: WHO. Available at http://www.who. int/cancer/publications/cancer_control_ detection/en/index.html.
10. Ginsberg GM, Lauer JA, Zelle S et al. (2012). Cost effectiveness of strategies to combat breast, cervical, and colorectal cancer in sub-Saharan Africa and South East Asia: mathematical modelling study. BMJ, 344:e614. http://dx.doi.org/10.1136/ bmj.e614 PMID:22389347
11. Alwan NA, Al-Attar WM, Eliessa RA et al. (2012). Knowledge, attitude and prac- tice regarding breast cancer and breast self-examination among a sample of the educated population in Iraq. East Mediterr Health J, 18:337–345. PMID:22768695
12. Lalla Salma Association Against Cancer (2009). National Cancer Prevention and Control Plan, 2010 –2019: Strategic axes and measures. Available at http://www. nccp-uicc.org/sites/default /files/plans/ Morocco%20National-Cancer-Axes-and- Measures.pdf.
13. Hakama M, Chamberlain J, Day NE et al. (1985). Evaluation of screening pro- grammes for gynaecological cancer. Br J Cancer, 52:669–673. http://dx.doi. org/10.1038/bjc.1985.241 PMID:4063143
14. Arbyn M, Ronco G, Anttila A et al. (2012). Evidence regarding human papilloma- virus testing in secondary prevention of cervical cancer. Vaccine, 30 Suppl 5:F88–F99. http://dx.doi.org/10.1016/j.vac cine.2012.06.095 PMID:23199969
15. Cuzick J, Bergeron C, von Knebel Doeberitz M et al. (2012). New technologies and pro- cedures for cervical cancer screening. Vaccine, 30 Suppl 5:F107–F116. http:// dx.doi.org/10.1016/j.vaccine.2012.05.088 PMID:23199953
16. Shapiro S, Strax P, Venet L (1971). Periodic breast cancer screening in reduc- ing mortality from breast cancer. JAMA, 215:1777–1785. http://dx.doi.org/10.1001/ jama.1971.03180240027005 PMID:5107709
17. Tabár L, Fagerberg CJ, Gad A et al. (1985). Reduction in mortality from breast cancer after mass screening with mammogra- phy. Randomised trial from the Breast Cancer Screening Working Group of the Swedish National Board of Health and Welfare. Lancet, 1:829–832. http://dx.doi. org/10.1016/S0140 - 6736(85) 9220 4 - 4 PMID:2858707
18. Timmers JM, den Heeten GJ, Adang EM et al. (2012). Dutch digital breast cancer screening: implications for breast cancer care. Eur J Public Health, 22:925–929. http://dx.doi.org/10.1093/eurpub/ckr170 PMID:22158996
19. Paci E; EUROSCREEN Working Group (2012). Summary of the evidence of breast cancer service screening outcomes in Europe and first estimate of the benefit and harm balance sheet. J Med Screen, 19 Suppl 1:5–13. http://dx.doi.org/10.1258/ jms.2012.012077 PMID:22972806
20. Moss SM, Nyström L, Jonsson H et al.; Euroscreen Working Group (2012). The im- pact of mammographic screening on breast cancer mortality in Europe: a review of trend studies. J Med Screen, 19 Suppl 1:26–32. http://dx.doi.org/10.1258/jms.2012.012079 PMID:22972808
21. Broeders M, Moss S, Nyström L et al.; EUROSCREEN Working Group (2012). The impact of mammographic screening on breast cancer mortality in Europe: a review of observational studies. J Med Screen, 19 Suppl 1:14–25. http://dx.doi.org/10.1258/ jms.2012.012078 PMID:22972807
22. Hewitson P, Glasziou P, Watson E et al. (2008). Cochrane systematic review of colo- rectal cancer screening using the fecal oc- cult blood test (hemoccult): an update. Am J Gastroenterol, 103:1541–1549. http://dx.doi. org/10.1111/j.1572- 0241.2008.01875.x PMID:18479499
23. von Karsa L, Patnick J, Segnan N (2012). European guidelines for quality assur- ance in colorectal cancer screening and diagnosis. First edition - Executive sum- mary. Endoscopy, 44 Suppl 3:SE1–SE8. http://dx.doi.org/10.1055/s-0032-1309822 PMID:23012113
24. Benson VS, Atkin WS, Green J et al.; International Colorectal Cancer Screening Network (2012). Toward standardizing and reporting colorectal cancer screening in- dicators on an international level: The International Colorectal Cancer Screening Network. Int J Cancer, 130:2961–2973. h t t p: / / d x . d o i . o r g / 10 .10 0 2 / i j c . 2 6 310 PMID:21792895
25. Herrero R (2012). A new era begins for cervical cancer control. HPV Today. Latin America Special Issue, 27:2. Available at www.g-o-c.org/uploads/12nov_hpvtoday. pdf.
26. Hernández-Ávila M, Lazcano-Ponce E, Cruz-Valdés A et al. (2012). Self-sampled vaginal testing to determine HPV DNA: ap- propriate technology for women who do not regularly attend cervical screening. HPV Today. Latin America Special Issue, 27:7. Available at www.g-o-c.org/uploads/12 nov_hpvtoday.pdf.
27. Nessa A, Hussain MA, Rahman JN et al. (2010). Screening for cervical neoplasia in Bangladesh using visual inspection with acetic acid. Int J Gynaecol Obstet, 111:115– 118. http://dx.doi.org/10.1016/j.ijgo.2010. 06.004 PMID:20674919
28. Chumworathayi B, Blumenthal PD, Limpaphayom KK et al. (2010). Effect of single-visit VIA and cryotherapy cervi- cal cancer prevention program in Roi Et, Thailand: a preliminary report. J Obstet Gynaecol Res, 36:79–85. http://dx.doi. org/10.1111/j.1447- 0756.2009.01089.x PMID:20178531
29. Pfaendler KS, Mwanahamuntu MH, Sahasrabuddhe VV et al. (2008). Management of cryotherapy-ineligible women in a “screen-and-treat” cervical cancer prevention program targeting HIV- infected women in Zambia: lessons from the field. Gynecol Oncol, 110:402–407. http:// dx.doi.org/10.1016/j.ygyno.2008.04.031 PMID:18556050
336
30. Mittra I, Mishra GA, Singh S et al. (2010). A cluster randomized, controlled trial of breast and cervix cancer screening in Mumbai, India: methodology and interim results after three rounds of screening. Int J Cancer, 126:976–984. http://dx.doi. org/10.1002/ijc.24840 PMID:19697326
31. Sankaranarayanan R, Ramadas K, Thara S et al. (2011). Clinical breast examina- tion: preliminary results from a cluster randomized controlled trial in India. J Natl Cancer Inst, 103:1476–1480. http://dx.doi. org/10.1093/jnci/djr304 PMID:21862730
32. Sankaranarayanan R, Sauvaget C, Ramadas K et al. (2011). Clinical trials of cancer screening in the developing world and their impact on cancer health- care. Ann Oncol, 22 Suppl 7:vii20–vii28. http://dx.doi.org/10.1093/annonc/mdr422 PMID:22039141
33. Sankaranarayanan R, Nene BM, Shastri SS et al. (2009). HPV screening for cervi- cal cancer in rural India. N Engl J Med, 360:1385–1394. http://dx.doi.org/10.1056/ NEJMoa0808516 PMID:19339719
34. Kim Y, Jun JK, Choi KS et al. (2011). Overview of the national cancer screen- ing programme and the cancer screening status in Korea. Asian Pac J Cancer Prev, 12:725–730. PMID:21627372
An evolutionary foundation for cancer control 337
“Nothing in biology makes sense except in the light of evolution” — T. Dobzhansky (1973)
Summary
Epidemiologists, cell and molecular biologists, geneticists and geno- mics experts, and oncologists can all lay claim to significant contri- butions towards the advancement of our understanding of cancer and its causation and treatment. But the stark fact remains that the worldwide burden of cancer re- mains huge and is escalating, and advanced or metastatic disease is mostly intransigent to treatment. Billions of dollars are riding on the premise that personalized medicine and targeted therapy will come to the rescue. But do we really have an adequate grasp of the underly- ing biology? Do we have a coherent framework for accommodating and rationalizing all the multilayered complexity that exists? Here, I ad- vance the argument that cancer is a complex adaptive system and that its causation, stepwise emergence, and therapeutic resistance all fol- low an evolutionary biology logic. The implications of this perspective for cancer control are considerable.
Darwinian evolution by natural se- lection is the fundamental law of biology and provides the frame- work for an understanding of the totality of the living world, from extremophilic species to human di- versity. Modern genomics provides a striking validation of evolutionary
An evolutionary foundation for cancer control
Melvyn “Mel” Greaves is a
professor of cell biology at the Institute
of Cancer Research in London; he
established the Leukaemia Research
Fund Centre there in 1984. Professor
Greaves trained in zoology and
obtained his Ph.D. in immunology at
University College London, before
focusing his research on cancer
and leukaemia. His pioneering work
includes developing new methods for
biological classification of leukaemias,
identifying that acute lymphoblastic
leukaemia (ALL) develops in the
womb, and identifying cancer stem
cells that cause ALL. His research
has greatly contributed to the
dramatic reduction in mortality from
childhood leukaemia in the past 30
years. His major research goals
include confirming the role that
common childhood infections play in
the development of leukaemia and
identifying the major causal factors of
the disease. Professor Greaves is also
a successful popular science writer
and encourages science writing
that is aimed at both scientists
and the general public.
Mel Greaves
Fig. P4.1. Speciation of living organisms: the ancestral tree. Charles Darwin’s spe- ciation “tree” from his notebook B, 1837. Founder 1 gives rise, via a branching ar- chitecture, to variants A, B, C, and D.
338
principles, by revealing phylogenet- ic relationships of species and sig- natures of prior, positive selection. How odd, then, that medical prac- tice, which deals with dysfunctional biology, should have, until very re- cently, remained impervious to this principle [1]. How can we under- stand emergent infectious diseas- es, antibiotic resistance, modern chronic diseases, drug resistance in cancer, and our vulnerability to cancer without reference to how our biology has been sculpted by evolutionary processes? In this Perspective, I summarize the ar- gument that evolutionary biology provides a coherent framework that recognizes and rationalizes the in- herent, multilayered complexity of
cancer. Proximate causes and a gene-centric view of cancer can be placed in context. Our apparent vulnerability to cancer and our lim- ited success in treating advanced disease can be seen in a different light. And, critically, an evolutionary biology paradigm can contribute novel ideas to the challenge of how we might best control cancer.
For the historical context of these ideas and more detailed data and argument, see [2–8].
Evolutionary determinants of vulnerability to cancer Evolution operates by stochastic, or random, genetic variation that provides the substrate for selec- tion in the context of environmental
selective pressures. Winners who survive and reproduce have the best adaptive phenotypes or fitness in relation to the prevailing selective forces. This is very different from intelligent design in that evolution can only select from random varia- tion in prior forms, or from the best of what’s available. This is a short- term fix strategy with “no eyes to the future” [9]. What is much less ap- preciated is the baggage, liabilities, and limitations that this inevitably entails (Table P4.1). These consid- erations can help rationalize the counterintuitive or paradoxical role of endogenous processes in can- cer, the ubiquity of mutant clones and premalignant lesions, and the
Table P4.1. Evolutionary determinants of vulnerability to cancer
Principle Liability Impact in cancer
Principle 1. Evolution just selects from the best available variants.
Limitations or imperfections of design, function, and specificity.
Error-prone replication. Error-prone DNA repair. Stem cells have inherent malignant potential: extensive, replicative, potential migratory phenotype, and quiescence option. Beneficial adaptive responses: inflammation, wound healing, and angiogenesis facilitate cancer promotion. Intrinsically mutagenic enzymes with off-target substrates: RAGs (lymphoid cancers), APOBECs/AID (multiple cancers/ lymphoid cancer).
Principle 2. Evolution has “no eyes to the future”; it just selects what fits best, now.
When environmental circumstances change, previously beneficial phe- notypes become mismatched with our genetic heritage.
Contemporary lifestyles, especially in affluent or “developed” societies, are mismatched with human adaptation to more ancient environments:
High UVB exposure versus under-melanized skin: skin cancers. Delayed first pregnancy, minimal breast feeding, estrogen-fuelling diets versus non-seasonal estrus: breast cancer, ovarian cancer. Diminished infection in infancy versus programmed anticipation of infection in the immune system: childhood acute lymphoblastic leukaemia.
Principle 3. Natural selection can only operate on phenotypes that have impacts on survival and reproduction.
There is little or no evolutionary resilience for deleterious phenotypes in the post- reproductive period. Adaptive features that are beneficial for reproductive life can later impose a trade-off or penalty (= antagonistic pleiotropy).
Our adaptive phenotypes are optimized for reproductive life versus extensive (decades) post-reproductive longevity: time for accidents to happen (via liabilities 1 and 2 above) (= age-associated cancer incidence).
Principle 4. Natural selection will always happen when replicating units have diversity and there is competition for resources and selective pressures that influence fitness for reproduction and that vary in space and time.
Clonal escape. Progressive selection of more robust or resilient clonal variants, metastasis, and drug resistance.
AID, activation-induced deaminase; APOBEC, apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like; RAGs, recombination activating genes; UVB, ultraviolet B.
An evolutionary foundation for cancer control 339
high frequency of cancer in ageing humans in modern societies.
Evolutionary considerations may also help us to understand why inher- ent susceptibility to particular can- cers, for example breast or prostate cancer, is highly variable within hu- man populations. Recent genome- wide association studies have pro- vided an audit of inherited gene variants that increase risk [10]. Highly penetrant mutant alleles, such as BRCA1 and BRCA2, exist at relative- ly high frequencies in certain popula- tions. This is best rationalized not by any historical, adaptive benefit they endowed but from founder effects, i.e. an origin in an individual (2000– 3000 years ago for BRCA1) who survived a population bottleneck [11]. More common allelic variants are as- sociated with much lower risk levels (odds ratios, 1.01–1.3) [10]. But why should such deleterious variants be common? A significant fraction of the variants that affect breast and pros- tate cancer are in genes that encode proteins in estrogen or androgen
metabolism and signalling and re- sponse pathways [12–14]. One possi- bility is that the relatively high frequen- cy of these alleles reflects prior positive selection because of a beneficial effect on enhancing fertility [15]. Genomic signatures can reveal whether genes have been subject to positive, adap- tive selection during human evolution and the approximate time frames in- volved [16].
Clonal evolution and cellular adaptability Cancer is a complex adaptive sys- tem with emergent properties,
some of which are both consistent with and deleterious to the host – metastasis and drug resistance in particular. Cellular adaptability de- rives from two principal sources: genetic or mutational diversity, and epigenetic or phenotypic plasticity. Both are critical in cancer.
For the past 3 billion years, when- ever replicating entities – molecules, cells, or individuals – have diversi- fied and were confronted with a chal- lenging or competitive environment, evolution by natural selection has occurred (principle 4 in Table P4.1) [17]. This is a, or even the, hallmark feature of cancer, at the somatic cell level – although curiously not listed as such [18]. Those 10 cellular fea- tures that are considered cancer hallmarks, including angiogenesis, evading cell death, and immune evasion [18], all enhance fitness for survival and/or reproduction – the ultimate register of evolutionary suc- cess. The idea that cancer involves sequential genetic changes in cells was evident from chromosome stud- ies for many years [19,20] but was first clearly posited as the clonal evolution paradigm by Peter Nowell in 1976 [2]. Since then, technical ad- vances in genetics, genomics, cell sorting, microscopy, and polymer- ase chain reaction (PCR)-based as- says, culminating in single-cell ge- nomics [21–23], have validated this concept and revealed the extensive and dynamic intraclonal genetic di- versity that exists in cancer [7]. The parallels with Darwinian evolution- ary speciation in ecosystems, miss- ing from the original vision [2], are now explicitly recognized [7].
Every patient’s cancer has a unique evolutionary trajectory or phylogenetic tree, with a branching
Fig. P4.2. Speciation of a cancer clone. Diagram showing the phylogenetic tree or evolutionary relationships of seven genetically diverse subclones of acute lymphoblastic leukaemia (ALL), all derived from a single common ancestral cell. All cells have been interrogated for the eight mutations listed on the left. Those mutations were determined by exome sequencing and single-nucleotide polymorphism arrays on whole leukaemic sample B1 (7.1%) normal cells. (Data from [23] using microfluidic-based multiplex quantitative polymerase chain reaction for single-cell genetic analysis.) 0C, loss of both copies; 1C, loss of one copy; het, heterozygous; homo, homozygous.
Table P4.2. Cancer cells’ evolutionary resilience and routes to therapeutic escape
Component Basis of resistance/adaptability [source]
Genetic diversity Drug (or immunological) targets or components of cellular response mutate [61]. Drug target (e.g. mutant protein) is segregated in subclone.a
Epigenetic plasticity
Cells bypass redundant signalling pathways blocked by drugs [45]. Stem cells adopt quiescent “resistant” state [62,63].
a Suggested by clonal variegation of mutations [24,25].
340
clonal architecture, as long recog- nized in ecological speciation (Figs P4.1, P4.2). Some mutations are consistently present in all cells at the base or trunk of the phyloge- netic tree, suggesting that they may be founder or initiating lesions, for example ETV6-RUNX1 mutations in acute lymphoblastic leukaemia (ALL) [23,24] and von Hippel–Lindau (VHL) gene mutations in renal carci- nomas [25]. Genetically distinct sub- clones of cancer frequently exist in distinctive, topographically different regions of tissues [25,26]. This is anticipated from evolutionary prin- ciples (consider Galapagos Islands finches) and poses considerable problems for traditional approaches to biopsy-based sampling and prog- nostics, as well as for biopsy-based genomics [27]. Although only a few studies have integrated serial sam- ples from primary tumours, metasta- ses, and post-therapeutic relapses or recurrences [8,24,28], the pat- terns that emerge are of dynamic shifts in clonal architecture, with minor subclones emerging after se- lective sweeps (Fig. P4.2). Dominant clones in relapse [29], drug-resistant recurrence [30,31], or metastasis [28] emerge from previously minor subclones whose existence can be traced back to primary samples.
The evolutionary dynamics of cancer are extremely variable, with intervals between initiation and di- agnosis of 1–50 years. The pace of evolution is often characterized by long periods of stasis or slow change, with occasional abrupt or catastrophic changes [8,32], par- alleling punctuated equilibrium in ecological evolution [33]. Oncogenic mutations are extremely common in all of us; premalignant lesions are probably ubiquitous, but only a frac- tion of them evolve to fully fledged malignancy [34]. The implication of this modest evolutionary penetration is that restraints are largely effective and promotional factors are likely critical. The challenge for epidemi- ologists is taking this protracted and “sluggish” natural history into ac- count, especially with respect to the timing of causative exposures. For
those interested in causation, cancer genomics may provide some vital clues. Although mutations are sto- chastic in origin, sequencing reveals mutational signatures that are indic- ative of underlying mechanisms of mutation and, in some cases, reflect the nature of genotoxic exposures [35–37].
Clonal evolution, as with specia- tion in ecology, is contingent upon ecosystem (tissue-derived or fil- tered) selective pressures, modu- lated by inherited susceptibility and chance (Fig. P4.3). Chance is all-pervasive in evolutionary biolo- gy and cancer because mutations are stochastic with respect to the functions encoded by genes [3]. The emergence of subclones with
particular putative “driver” muta- tions can be viewed as an adaptive response to selective pressures, of which environmental and therapeu- tic exposures are clear examples. Intense therapeutic selective pres- sure often has the exact opposite effect to that desired: the positive selection of previously silent and innocuous subclones. Metastases are likely to require adaptation to ec- topic tissue, fermenting further evo- lution. Adaptation of cancer cells via beneficial mutations involves both gain-of-function mutants (activat- ing oncogenes) and loss-of-function mutants (deleted tumour suppressor genes), similar to bacteria adapting to adverse conditions [38]. The com- plex mutation profiles of individual
Fig. P4.3. Components of cancer clone evolution. The different colours represent genetically distinct subclones. The roulette wheel symbolizes the role of chance in all components. Endogenous selective pressures include inflammation, anoxia, and metabolic stress [57].
Table P4.3. Evolutionary parameters are predictive of cancer progression and clinical outcome
Evolutionary parameter Clinical outcome [source]
Measures of intraclonal genetic diversity (substrate for selection)
Progression of Barrett oesophagus [64] and chronic lymphocytic leukaemia [65]. Survival in head and neck cancer [66].
Burden of stem cells (units of selection)
Progression and outcome in multiple cancers ([67] and references therein; [68]).
Ecosystem diversity (selective pressures)
Survival in breast cancer [69]. Mathematical modelling [70].
An evolutionary foundation for cancer control 341
cancer cells reflect combinatorial impacts on cellular fitness and, in some cases, epistatic interactions of mutations, which occur in evolu- tion in general [39]. Genetic insta- bility and chromosomal aneuploidy, common features in cancer, are other adaptive tactics that acceler- ate the pace of diversification and increase the probability of beneficial
mutations arising. Bacteria use the same evolutionary trick under chal- lenging conditions: blind, but more frequent, spinning of the roulette wheel (Fig. P4.3).
Epigenetic plasticity The adaptive resilience of cancer cells is not solely dependent on mutation. Cells within individual
genetically homogeneous clones have epigenetic plasticity and di- verse alternative phenotypes that are spatiotemporally expressed (Fig. P4.4A) [40–42]. This facili- tates evasion of a therapeutic chal- lenge (Table P4.2). The adaptability of cells is further evidenced at the single-cell level. Each cell has an extensive, complex signalling net- work that is highly dynamic, robust, and adaptable (Fig. P4.4B) [43]. In cancer, all functionally relevant mu- tations [44] effectively corrupt these networks, resulting in dysregulation or resetting to a new steady state (e.g. “on” for proliferation). The critical nodes in the signalling net- works of normal cells, as well as cancer cells, have built-in redun- dancy, which provides yet another adaptive route to clonal escape in cancer, for example from targeted therapy (Table P4.2) [45]. The in- trinsic adaptability or epigenetic plasticity of cells is itself an ancient evolutionary legacy reflecting built- in safeguards to combat adverse circumstances, particularly for stem cells in more complex long-lived or- ganisms. Cancer stem cells “hunker down” or adopt a dormancy (or qui- escent) status under chemothera- peutic challenge [46], an adaptive tactic of considerable evolutionary antiquity [47]. Those cells within subclones that have self-renewal or stem cell competence [48] are crucial to both clonal progression and adaptability as they are, in evo- lutionary terms, the “units of selec- tion” [49] by virtue of their capacity for extensive or unlimited self-re- newal and proliferation [50]. In ac- cordance with a status of “units of selection”, cancer stem cells within individual patients are genetically diverse [24] as well as phenotypi- cally plastic [41].
Can we predict and thwart the evolutionary resilience of cancer? Despite the enthusiasm and some successes for genome-guided per- sonalized medicine and targeted therapeutics in cancer, it is evident
Fig. P4.4. Epigenetic plasticity of cancer cells. (A) Illustration of the variation of phenotype within a single subclone. C, cell; D+, differentiating cells; Q, quiescent (out- of-cycle) cells. (B) Depiction of the complex and dynamic signalling networks of an individual cell [43]. Dense coloured regions indicate critical signalling routes. Pathways involved are indicated around the outside of the cell. These data derive from genetic (mutational) analysis of gene interactions in yeast cells. We currently have no such wiring diagram for human cells. The figure underestimates complexity by showing interaction networks under single, steady-state conditions. The reality is considerably more dynamic and complex.
B
A
Q D+ D+
C
C C
Self-renewing stem cells
Phenotypic variation
Intraclonal diversity
1 cell
1 subclone
342
that the adaptive resilience of cancer cells usually wins out, a pyrrhic vic- tory for the genetic and epigenetic adaptability of cells (Table P4.2). Evolutionary principles of natural selection suggest that the likeli- hood of therapeutic failure should be predictable by quantitative mea- sures of the major drivers of clonal diversity and selection. In strains of microbial species, mutation rates are predictive of the probability of drug-resistant mutants [51]. The same has long been anticipated for cancer, but here mutation rates are difficult to measure because of dy- namic changes and topography of subclones. However, other evolu- tionary parameters that should be predictive of the likelihood of mu- tation-based drug resistance can be assayed and quantified: genetic diversity within the clone, the size of the selectable stem cell compart- ment, and the diversity of the eco- system. Recent data confirm these expectations (Table P4.3), with con- siderable implications for instiga- tion of tests that can predict cancer progression of early lesions or ther- apeutic resistance. The implica- tions for treatment itself are major.
We are pollarding (or pruning) the major branches of the cancer clone phylogenetic tree, largely in igno- rance of its architecture and intrin- sic resilience. The common conse- quence, post-therapeutically, is the illusory success of transient tumour shrinkage, followed by florid regen- eration as residual minor clones adapt to the new circumstances.
Given the extraordinary adapt- ability of cells, the sheer number of proliferative cycles (1011 per day in the small intestine and bone mar- row), and the inherent risk of mu- tation and the ubiquity of covert tumours [49], it is perhaps surpris- ing that we don’t all have malignant cancer at an early age [52]. The rea- son is that the success of multicel- lularity as an evolutionary innovation some 600 million years ago required securing the maintenance of tissue integrity with multiple restraints on clonal expansion. This included the innovation of gene functions that can also operate as tumour suppressors [52–54]. And, despite the indulgenc- es of human behaviour, these re- straints remain largely successful up to and during our evolutionarily “use- ful” reproductive lifetime. But the
challenging reality is that approxi- mately one in three individuals will receive a diagnosis of life-threat- ening cancer. From an evolutionary perspective, the question that needs to be posed is: “How can we best thwart the evolutionary resilience of cancer?” The same issue is cen- tral to the control of drug-resistant tuberculosis, malaria, and HIV [55]. Generally, we already know what to do for cancer (Fig. P4.5). The prob- lem is where to put our efforts and resources, and what particular tac- tics to adopt. The best way to stop evolution is clearly to prevent it start- ing in the first place. But once it’s up and running, there are strategies available that are distinct from cur- rent or conventional practice, some of which exploit cancer’s evolution- ary features and ecological depend- encies (Fig. P4.5) [56–60]. To do this effectively would require consid- erable refocusing of current cancer research and therapeutic priorities. And a recognition of cancer’s indel- ible, evolutionary character.
This Perspective is dedicated to the memory of Professor Pat Buffler.
Fig. P4.5. How can we thwart cancer’s evolutionary resilience? Numbers on the left indicate the percentage of cancers that might be prevented or controlled by implementing these three approaches. The proportion of 75% for prevention is based on original estimates by Doll and Peto [71]. (1) Vaccines for infection-associated cancers: human papillomavirus, Epstein–Barr virus, hepatitis B virus, hepatitis C virus, Helicobacter pylori, etc. (2) Smoking, chronic or intermittent, and exposure to intensive ultraviolet B radiation. (3) Diet and exercise. (4) Surgery for some premalignant lesions (skin, colon, cervix); non-steroidal anti-inflammatory drugs (e.g. aspirin) for premalignant lesions of the gastrointestinal tract and others [72]. (5) As for HIV [52], but applied to cancer [73,74]. (6) See [75]. (7) Molecules that cancer cells invariably depend upon or are addicted to, for example activated MYC, heat shock proteins, self- renewal signalling. (8) Founder lesion in base or trunk of phylogenetic tree [24,25,76]. Target in every cancer cell, for example ABL kinase targeting with imatinib in chronic myeloid leukaemia [77]. (9) Targeting features of self-renewal signatures of cancer stem cells that are different from those of normal stem cells [63,78,79]. (10) Anti- angiogenesis, anti-inflammatories, modest-dose cytostatic versus intensive cytotoxic therapy; see [58–60].
An evolutionary foundation for cancer control 343
References 1. Stearns SC, Nesse RM, Govindaraju DR,
Ellison PT (2010). Evolution in health and medicine Sackler colloquium: evolutionary perspectives on health and medicine. Proc Natl Acad Sci U S A, 107 Suppl 1:1691– 1695. http://dx.doi.org/10.1073/pnas.0914 475107 PMID:20133821
2. Nowell PC (1976). The clonal evolution of tu- mor cell populations. Science, 194:23–28. http://dx.doi.org/10.1126/science.959840 PMID:959840
3. Greaves M (2000). Cancer: The Evolutionary Legacy. Oxford: Oxford University Press.
4. Crespi B, Summers K (2005). Evolutionary biology of cancer. Trends Ecol Evol, 20:545– 552. http://dx.doi.org/10.1016/j.tree.2005. 07.007 PMID:16701433
5. Merlo LMF, Pepper JW, Reid BJ, Maley CC (2006). Cancer as an evolutionary and ecological process. Nat Rev Cancer, 6: 924–935. http://dx.doi.org/10.1038/nrc2013 PMID:17109012
6. Greaves M (2007). Darwinian medicine: a case for cancer. Nat Rev Cancer, 7:213–221. http://dx.doi.org/10.1038/nrc2071 PMID: 17301845
7. Greaves M, Maley CC (2012). Clonal evo- lution in cancer. Nature, 481:306–313. ht tp: //dx.doi.org /10.103 8 /nature10762 PMID:22258609
8. Yates LR, Campbell PJ (2012). Evolution of the cancer genome. Nat Rev Genet, 13: 795–806. http://dx.doi.org/10.1038/nrg3317 PMID:23044827
9. Williams GC (1966). Adaptation and Natural Selection. Princeton, NJ: Princeton University Press.
10. Fletcher O, Houlston RS (2010). Architecture of inherited susceptibil- ity to common cancer. Nat Rev Cancer, 10:353–361. http://dx.doi.org/10.1038/nrc 2840 PMID:20414203
11. Szabo CI, King M-C (1997). Population genetics of BRCA1 and BRCA2. Am J Hum Genet, 60:1013–1020. PMID:9150148
12. Low YL, Li Y, Humphreys K et al. (2010). Multi-variant pathway association analy- sis reveals the importance of genetic determinants of estrogen metabolism in breast and endometrial cancer suscep- tibility. PLoS Genet, 6:e1001012. http:// dx.doi.org/10.1371/journal.pgen.1001012 PMID:20617168
13. Eeles RA, Kote-Jarai Z, Giles GG et al.; UK Genetic Prostate Cancer Study Collaborators; British Association of Urological Surgeons’ Section of Oncology; UK ProtecT Study Collaborators (2008). Multiple newly identified loci associ- ated with prostate cancer susceptibil- ity. Nat Genet, 40:316–321. http://dx.doi. org/10.1038/ng.90 PMID:18264097
14. Ghoussaini M, Fletcher O, Michailidou K et al.; Netherlands Collaborative Group on Hereditary Breast and Ovarian Cancer (HEBON); Familial Breast Cancer Study (FBCS); Gene Environment Interaction of Breast Cancer in Germany (GENICA) Network; kConFab Investigators; Australian Ovarian Cancer Study Group (2012). Genome-wide association analysis identi- fies three new breast cancer susceptibility loci. Nat Genet, 44:312–318. http://dx.doi. org/10.1038/ng.1049 PMID:22267197
15. Greaves M (2002). Cancer causation: the Darwinian downside of past success? Lancet Oncol, 3:244–251. http://dx.doi.org/ 10.1016/S1470-2045(02)00716-7 PMID: 12067687
16. Harris EE, Meyer D (2006). The molecu- lar signature of selection underlying hu- man adaptations. Am J Phys Anthropol, 131 Suppl 43:89–130. http://dx.doi.org/ 10.1002/ajpa.20518 PMID:17103426
17. Buss LW (1987). The Evolution of Individuality. Princeton, NJ: Princeton University Press.
18. Hanahan D, Weinberg RA (2011). Hallmarks of cancer: the next genera- tion. Cell, 144:646–674. http://dx.doi.org/ 10.1016/j.cell.2011.02.013 PMID:21376230
19. Boveri T (1929). The Origin of Malignant Tumors. London: Baillière, Tindall & Cox.
20. Sandberg AA, Hossfeld DK (1970). Chromosomal abnormalities in human neoplasia. Annu Rev Med, 21:379–408. http://dx.doi.org/10.1146/annurev.me.21.0 20170.002115 PMID:4247449
21. Zong C, Lu S, Chapman AR, Xie XS (2012). Genome-wide detection of sin- gle-nucleotide and copy-number vari- ations of a single human cell. Science, 338:1622–1626. http://dx.doi.org/10.1126/ science.1229164 PMID:23258894
22. Baslan T, Kendall J, Rodgers L et al. (2012). Genome-wide copy number analysis of single cells. Nat Protoc, 7:1024–1041. http://dx.doi.org/10.1038/nprot.2012.039 PMID:22555242
23. Potter NE, Ermini L, Papaemmanuil E et al. (2013). Single cell mutational profiling and clonal phylogeny in cancer. Genome Res, [epub ahead of print]. http://dx.doi. org/10.1101/gr.159913.113 PMID:24056532
24. Anderson K, Lutz C, van Delft FW et al. (2011). Genetic variegation of clonal archi- tecture and propagating cells in leukae- mia. Nature, 469:356–361. http://dx.doi. org/10.1038/nature09650 PMID:21160474
25. Gerlinger M, Rowan AJ, Horswell S et al. (2012). Intratumor heterogene- ity and branched evolution revealed by multiregion sequencing. N Engl J Med, 366:883–892. http://dx.doi.org/10.1056/ NEJMoa1113205 PMID:22397650
26. Sottoriva A, Spiteri I, Piccirillo SGM et al. (2013). Intratumor heterogeneity in human glioblastoma reflects cancer evolution- ary dynamics. Proc Natl Acad Sci U S A, 110:4009–4014. http://dx.doi.org/10.1073/ pnas.1219747110 PMID:23412337
27. Swanton C (2012). Intratumor heteroge- neity: evolution through space and time. Cancer Res, 72:4875– 4882. http://dx.doi. o r g /10.115 8 / 0 0 0 8 - 5 472 .CA N -12-2 217 PMID:23002210
28. Yachida S, Jones S, Bozic I et al. (2010). Distant metastasis occurs late during the genetic evolution of pancreatic cancer. Nature, 467:1114–1117. http://dx.doi.org/ 10.1038/nature09515 PMID:20981102
29. Mullighan CG, Phillips LA, Su X et al. (2008). Genomic analysis of the clonal origins of relapsed acute lymphoblastic leukemia. Science, 322:1377–1380. http://dx.doi.org/10.1126/science.1164266 PMID:19039135
30. Meyer JA, Wang J, Hogan LE et al. (2013). Relapse-specific mutations in NT5C2 in childhood acute lymphoblastic leukemia. Nat Genet, 45:290–294. http://dx.doi.org/ 10.1038/ng.2558 PMID:23377183
31. Diaz LA Jr, Williams RT, Wu J et al. (2012). The molecular evolution of acquired resis- tance to targeted EGFR blockade in colo- rectal cancers. Nature, 486:537–540. ht t p: //dx.do i.o r g /10.10 3 8 /nature11219 PMID:22722843
32. Baca SC, Prandi D, Lawrence MS et al. (2013). Punctuated evolution of prostate cancer genomes. Cell, 153:666–677. http:// d x .d o i.o r g /10.1016 / j.c e ll. 2 013 .0 3 .0 21 PMID:23622249
33. Gould SJ, Eldredge N (1993). Punctuated equilibrium comes of age. Nature, 366:223–227. http://dx.doi.org/10.1038/36 6223a0 PMID:8232582
34. Greaves M (2013). Does everyone de- velop covert cancer? Nat Rev Cancer, (in press).
35. Alexandrov LB, Nik-Zainal S, Wedge DC et al.; Australian Pancreatic Cancer Genome Initiative; ICGC Breast Cancer Consortium; ICGC MMML-Seq Consortium; ICGC PedBrain (2013). Signatures of mutation- al processes in human cancer. Nature, 500:415–421. http://dx.doi.org/10.1038/na ture12477 PMID:23945592
36. Poon SL, Pang S-T, McPherson JR et al. (2013). Genome-wide mutational signa- tures of aristolochic acid and its applica- tion as a screening tool. Sci Transl Med, 5:ra101. http://dx.doi.org/10.1126/scitransl med.3006086 PMID:23926199
37. Pfeifer GP (2010). Environmental expo- sures and mutational patterns of cancer genomes. Genome Med, 2:54. http://dx.doi. org/10.1186/gm175 PMID:20707934
344
38. Hottes AK, Freddolino PL, Khare A et al. (2013). Bacterial adaptation through loss of function. PLoS Genet, 9:e1003617. http:// dx.doi.org/10.1371/journal.pgen.1003617 PMID:23874220
39. Breen MS, Kemena C, Vlasov PK et al. (2012). Epistasis as the primary factor in molecular evolution. Nature, 490:535–538. ht t p: //d x.d o i.o r g /10.10 3 8 /nature11510 PMID:23064225
40. Friedl P, Alexander S (2011). Cancer inva- sion and the microenvironment: plasticity and reciprocity. Cell, 147:992–1009. http:// d x .d o i .o r g /10 .1016 / j .c e l l . 2 011.11. 016 PMID:22118458
41. Kreso A, O’Brien CA, van Galen P et al. (2013). Variable clonal repopulation dy- namics influence chemotherapy response in colorectal cancer. Science, 339:543– 548. http://dx.doi.org/10.1126/science.122 7670 PMID:23239622
42. Biddle A, Liang X, Gammon L et al. (2011). Cancer stem cells in squamous cell carci- noma switch between two distinct pheno- types that are preferentially migratory or proliferative. Cancer Res, 71:5317–5326. h t t p: //d x .d o i .o r g /10 .115 8 / 0 0 0 8 - 5 47 2 . CAN-11-1059 PMID:21685475
43. Costanzo M, Baryshnikova A, Bellay J et al. (2010). The genetic landscape of a cell. Science, 327:425– 431. http://dx. doi.org/10.1126/science.1180823 PMID: 20093466
44. Vogelstein B, Papadopoulos N, Velculescu VE et al. (2013). Cancer genome land- scapes. Science, 339:1546–1558. http:// d x . d o i . o r g / 10 .112 6 / s c i e n c e .12 3 512 2 PMID:23539594
45. Wilson TR, Fridlyand J, Yan Y et al. (2012). Widespread potential for growth- factor-driven resistance to anticancer kinase inhibitors. Nature, 487:505–509. ht t p: //dx.doi.o rg /10.10 3 8 /nature1124 9 PMID:22763448
46. Frank NY, Schatton T, Frank MH (2010). The therapeutic promise of the cancer stem cell concept. J Clin Invest, 120: 41–50. http://dx.doi.org/10.1172/JCI41004 PMID:20051635
47. Lewis K (2007). Persister cells, dormancy and infectious disease. Nat Rev Microbiol, 5:48–56. http://dx.doi.org/10.1038/nrmicro 1557 PMID:17143318
48. Dick JE (2008). Stem cell concepts renew cancer research. Blood, 112:47 93–4807. http://dx.doi.org/10.1182/blood- 2008-08-077941 PMID:19064739
49. Greaves M (2013). Cancer stem cells as ‘units of selection’. Evol Appl, 6:102–108. http://dx.doi.org/10.1111/eva.12017 PMID: 23396760
50. O’Brien CA, Kreso A, Jamieson CHM (2010). Cancer stem cells and self-renew- al. Clin Cancer Res, 16:3113–3120. http:// dx.doi.org/10.1158/1078 - 0432.CCR- 09 - 2824 PMID:20530701
51. Ford CB, Shah RR, Maeda MK et al. (2013). Mycobacterium tuberculosis mu- tation rate estimates from different line- ages predict substantial differences in the emergence of drug-resistant tuberculo- sis. Nat Genet, 45:784–790. http://dx.doi. org/10.1038/ng.2656 PMID:23749189
52. Bissell MJ, Hines WC (2011). Why don’t we get more cancer? A proposed role of the microenvironment in restraining can- cer progression. Nat Med, 17:320–329. http://dx.doi.org/10.1038/nm.2328 PMID: 21383745
53. Nakajima Y, Meyer EJ, Kroesen A et al. (2013). Epithelial junctions maintain tis- sue architecture by directing planar spin- dle orientation. Nature, 500:359–362. ht tp: //dx.doi.org /10.103 8 /nature123 3 5 PMID:23873041
54. Domazet-Loso T, Tautz D (2010). Phylostratigraphic tracking of cancer genes suggests a link to the emergence of multicellularity in metazoa. BMC Biol, 8:66. http://dx.doi.org/10.1186/1741-7007- 8-66 PMID:20492640
55. Goldberg DE, Siliciano RF, Jacobs WR Jr (2012). Outwitting evolution: fighting drug-resistant TB, malaria, and HIV. Cell, 148:1271–1283. http://dx.doi.org/10.1016/j. cell.2012.02.021 PMID:22424234
56. Pienta KJ, McGregor N, Axelrod R, Axelrod DE (2008). Ecological therapy for cancer: defining tumors using an ecosys- tem paradigm suggests new opportunities for novel cancer treatments. Transl Oncol, 1:158–164. PMID:19043526
57. Gillies RJ, Verduzco D, Gatenby RA (2012). Evolutionary dynamics of carcinogenesis and why targeted therapy does not work. Nat Rev Cancer, 12:487–493. http://dx.doi. org/10.1038/nrc3298 PMID:22695393
58. Silva AS, Kam Y, Khin ZP et al. (2012). Evolutionary approaches to prolong pro- gression-free survival in breast cancer. Cancer Res, 72:6362–6370. http://dx.doi. org /10.115 8 / 0 0 0 8 - 5 472.CA N -12-2 23 5 PMID:23066036
59. Gatenby RA, Brown J, Vincent T (2009). Lessons from applied ecology: cancer control using an evolutionary double bind. Cancer Res, 69:7499–7502. http://dx.doi. org /10.115 8 / 0 0 0 8 - 5 472.CA N - 0 9 -13 5 4 PMID:19752088
60. Gatenby RA, Silva AS, Gillies RJ, Frieden BR (2009). Adaptive therapy. Cancer Res, 69:4894– 4903. http://dx.doi. org /10.115 8/0 0 0 8 - 5 472.CA N - 0 8 - 3 65 8 PMID:19487300
61. Redmond KM, Wilson TR, Johnston PG, Longley DB (2008). Resistance mecha- nisms to cancer chemotherapy. Front Biosci, 13:5138–5154. http://dx.doi.org/10. 2741/3070 PMID:18508576
62. Ishikawa F, Yoshida S, Saito Y et al. (2007). Chemotherapy-resistant human AML stem cells home to and engraft with- in the bone-marrow endosteal region. Nat Biotechnol, 25:1315–1321. http://dx.doi. org/10.1038/nbt1350 PMID:17952057
63. Saito Y, Uchida N, Tanaka S et al. (2010). Induction of cell cycle entry eliminates human leukemia stem cells in a mouse model of AML. Nat Biotechnol, 28:275– 280. http://dx.doi.org/10.1038/nbt.1607 PMID:20160717
64. Maley CC, Galipeau PC, Finley JC et al. (2006). Genetic clonal diversity predicts progression to esophageal adenocarcino- ma. Nat Genet, 38:468– 473. http://dx.doi. org/10.1038/ng1768 PMID:16565718
65. Landau DA, Carter SL, Stojanov P et al. (2013). Evolution and impact of subclonal mutations in chronic lymphocytic leuke- mia. Cell, 152:714–726. http://dx.doi.org/ 10.1016/j.cell.2013.01.019 PMID:23415222
66. Mroz EA, Tward AD, Pickering CR et al. (2013). High intratumor genetic hetero- geneity is related to worse outcome in patients with head and neck squamous cell carcinoma. Cancer, 119:3034–3042. h t t p: //d x . d o i . o r g /10 .10 0 2 /c n c r. 2 815 0 PMID:23696076
67. Greaves M (2011). Cancer stem cells renew their impact. Nat Med, 17:1046– 1048. http://dx.doi.org/10.1038/nm.2458 PMID:21900918
68. Lapouge G, Beck B, Nassar D et al. (2012). Skin squamous cell carcinoma propagating cells increase with tumour progression and invasiveness. EMBO J, 31:4563– 4575. http://dx.doi.org/10.1038/ emboj.2012.312 PMID:23188079
69. Yuan Y, Failmezger H, Rueda OM et al. (2012). Quantitative image analysis of cel- lular heterogeneity in breast tumors com- plements genomic profiling. Sci Transl Med, 4:ra143. http://dx.doi.org/10.1126/ scitranslmed.3004330 PMID:23100629
70. Anderson ARA, Weaver AM, Cummings PT, Quaranta V (2006). Tumor morphology and phenotypic evolution driven by selective pressure from the microenvironment. Cell, 127:905–915. http://dx.doi.org/10.1016/j.cell. 2006.09.042 PMID:17129778
71. Doll R, Peto R (1981). The Causes of Cancer. Oxford: Oxford University Press.
72. Rothwell PM, Fowkes FGR, Belch JFF et al. (2011). Effect of daily aspirin on long- term risk of death due to cancer: analysis of individual patient data from randomised trials. Lancet, 377:31–41. http://dx.doi. o r g /10 .1016 / S 014 0 - 6 7 3 6 (10) 6 2110 -1 PMID:21144578
73. Al-Lazikani B, Banerji U, Workman P (2012). Combinatorial drug therapy for can- cer in the post-genomic era. Nat Biotechnol, 30:679–692. http://dx.doi.org/10.1038/nbt. 2284 PMID:22781697
74. Sullivan RJ, Lorusso PM, Flaherty KT (2013). The intersection of immune- directed and molecularly targeted therapy in advanced melanoma: where we have been, are, and will be. Clin Cancer Res, 19:5283–5291. http://dx.doi. o r g /10.115 8 /1078 - 0 4 32 .C C R-13 -2151 PMID:24089441
An evolutionary foundation for cancer control 345
75. Ashworth A, Lord CJ, Reis-Filho JS (2011). Genetic interactions in cancer progression and treatment. Cell, 145:30–38. http:// d x .d o i.o r g /10.1016 / j.c e l l. 2 011. 0 3 . 0 2 0 PMID:21458666
76. Yap TA, Gerlinger M, Futreal PA et al. (2012). Intratumor heterogeneity: seeing the wood for the trees. Sci Transl Med, 4:27ps10. http://dx.doi.org/10.1126/scitran slmed.3003854 PMID:22461637
77. Druker BJ (2008). Translation of the Philadelphia chromosome into therapy for CML. Blood, 112:4808– 4817. http://dx.doi. o r g / 1 0 .118 2 / b l o o d - 2 0 0 8 - 0 7- 0 7 7 9 5 8 PMID:19064740
78. Gupta PB, Onder TT, Jiang G et al. (2009). Identification of selective inhibi- tors of cancer stem cells by high-through- put screening. Cell, 138:645–659. http:// dx.doi.o rg /10.1016 / j.c ell. 20 0 9.0 6.0 3 4 PMID:19682730
79. Abrahamsson AE, Geron I, Gotlib J et al. (2009). Glycogen synthase kinase 3beta missplicing contributes to leukemia stem cell generation. Proc Natl Acad Sci U S A, 106:3925–3929. http://dx.doi.org/10.1073/ pnas.0900189106 PMID:19237556
This page intentionally left blank
Different tumour types may be characterized with ref- erence to epidemiology, etiology, pathology, genetics, and prevention. Current knowledge in these fields is the culmination of decades, if not centuries, of re- search, but progress has not been uniform. Notably, remarkable advances have accrued through a par- ticular approach during a relatively short period. This edition of World Cancer Report corresponds to a new dimension in characterization of cancer as a genetic disease. Investigation of single genes – epitomized by oncogenes and tumour suppressor genes – has been eclipsed by sequencing of the whole genome, transcriptome, epigenome, or comparable entity for each of the major tumour types, typically involving multi-institutional collaborations based on hundreds of specimens. The benefits – definition of susceptibility, improved means of diagnosis, and development of targeted therapies – vary markedly between tumour types. It is clear that cancer is not a single disease but a multiplicity of different diseases. Identification of exogenous causes, screening methods, high-risk groups, means of diagnosis, and effective therapy vary across tumour types; in most cases the spectrum extends from comprehensive understanding to, as yet, no meaningful impact. Accordingly, knowledge about cancer must be specified with respect to each tumour type.
Cancer by organ site
5
348
A guide to the epidemiology data in World Cancer Report
Incidence Incidence is defined as the number of new cases of a disease arising in a given period in a specified popu- lation. It can be expressed as an absolute number of cases per year or as a rate per 100 000 persons (or some other denominator) per year. The rate provides an approximation of the average risk of developing a cancer.
Mortality Mortality is defined as the number of deaths due to a specific under- lying cause occurring in a given period in a specified population. It can be expressed as an absolute number of deaths per year or as a rate per 100 000 persons (or some other denominator) per year.
Age standardization In this Report, all incidence and mortality rates are age-standard- ized. An age-standardized rate (ASR) is a summary measure of the rate that a population would have if it had a standard age structure. Standardization is necessary when comparing several populations (or the same population at different time points); age has a powerful influence on the risk of cancer, and populations differ with respect to their age distribution. Here, the ASR uses the World Standard Population (of Segi [1], as modi- fied by Doll et al. [2]). The calcu- lated incidence or mortality rate is then called the age-standardized incidence or mortality rate (World) and is conventionally expressed per 100 000 persons.
Graphics Maps, bar charts, pie charts, and time trend figures are shown in the global cancer burden overview (in Section 1) and in site-specific chapters (in Section 5). In the chapters in Section 5, maps, pie charts, and time trend
figures are provided for both inci- dence and mortality where important differences are in evidence (e.g. for breast cancer and colorectal cancer). For cancer sites where survival is rel- atively low and has changed relatively little, the figures focus on incidence (e.g. for lung cancer and pancreatic cancer). The figures are generally dis- played for men and women separate- ly, although for certain sites, a choice has been made to select patterns and trends in either men or women where they are reasonably comparable (e.g. for leukaemia) or where certain sex- specific features can be highlighted (e.g. for thyroid cancer).
Data sources The graphics are built on three data sources. First, the maps and the pie charts are based on national inci- dence and mortality estimates from GLOBOCAN 2012 [3]. This provides incidence, mortality, and prevalence estimates for 27 site-specific can- cers and for all sites combined in 184 countries worldwide. The un- derlying principle in the estimation process is a reliance on the best available data on cancer incidence and/or mortality within a country to build up the global picture. The results are more accurate or less accurate for different countries, de- pending on the extent and accuracy of locally available data. Second, the incidence time trends and bar charts are based on observed data series from regional or national population-based cancer registries, extracted from Cancer Incidence in Five Continents (CI5) Volumes I to X [4,5] and completed with more recent data available from corre- sponding cancer registry websites. Third, the mortality data for the time trends and bar charts originate from the WHO Mortality Database [6] (or the Centers for Disease Control and Prevention [7] for the USA) and, ex- cept for China, are available only at the national level.
Maps
The maps provide a global picture of the burden of a cancer site, by sex.
Pie charts
When both incidence and mortality pie charts are presented, the area of the mortality pie is proportional to the number of deaths relative to the number of new cases.
Bar charts
Bar charts of incidence and mor- tality by cancer, and of incidence and mortality by registry/country for all sites combined, are shown in Chapter 1.1.
A guide to the epidemiology data in World Cancer Report 349
5 O
R G
A N
S IT
E
Time trends of 0.5 was used for seven specific cancer sites (oesophagus, liver, pancreas, kidney, testis, thyroid, and nervous system) due to the small number of cases. As annual data for Uganda were not available, natural cubic spline interpolation was used to predict the values for the missing years. Care should be taken when comparing incidence graphs (time trends or bar charts) as the scales used on the vertical axes may differ.
Notes for bar charts Average annual mortality rates for China are for 1996–2000. Average annual mortality rates for Australia are for 2000–2004.
Notes for time trends and bar charts Countries represented by regional registries are marked with *; other- wise, registries are national.
Countries represented by one or more than one regional regis- try: China (Hong Kong Special Ad- ministrative Region and Shanghai), Colombia (Cali), India (Chennai and Mumbai), Japan (Miyagi, Nagasaki, and Osaka Prefectures), Spain (Granada, Murcia, Navarra, and Tarragona), Uganda (Kampala), and USA (the Surveillance, Epidemiology
and End Results [SEER] 9 registries: states of Connecticut, Hawaii, Iowa, New Mexico, and Utah and metro- politan areas of San Francisco– Oakland, California; Detroit, Michi- gan; Seattle–Puget Sound, Wash- ington; and Atlanta, Georgia).
References 1. Segi M (1960). Cancer Mortality for
Selected Sites in 24 Countries (1950– 57). Department of Public Health, Tohoku University of Medicine, Sendai, Japan.
2. Doll R, Payne P, Waterhouse JAH, eds (1966). Cancer Incidence in Five Continents, Vol. I. Geneva: International Union Against Cancer.
3. Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
4. Ferlay J, Parkin DM, Curado MP et al. (2010). Cancer Incidence in Five Continents, Volumes I to IX: IARC CancerBase No. 9 [Internet]. Lyon: IARC. Available at http://ci5.iarc.fr.
5. Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X. Available at http://ci5. iarc.fr.
6. WHO Mortality Database. Available at http://www.who.int/healthinfo/statistics/ mortality _rawdata/en/index.html.
7. National Center for Health Statistics, U.S. Centers for Disease Control and Prevention. Available at http://www.cdc. gov/nchs/.
To highlight recent temporal devel- opments, time trends are described from about 1975 onwards. For all cancer sites, data were selected that cover the same set of 11 coun- tries for incidence and 9 countries for mortality (no mortality data were available for Uganda or India, and mortality data for China were avail- able only for specific cancer sites). For the USA, incidence and mortality data are shown separately for Black and White populations. A logarithmic scale is used for the age-standard- ized rates to show the variety of pat- terns between populations.
Locally weighted regression curves were fitted to provide smoothed lines through the scat- ter plot of ASR by calendar year. A bandwidth of 0.3 was used, i.e. 30% of the data were used in smoothing each point. A bandwidth
350
Summary • Lung cancer is the most com-
mon cancer in men and the third most common in women.
• Tobacco smoking, including second-hand smoke, is the pre- dominant cause of lung cancer worldwide. Other causes of lung cancer include radon, occupa- tional exposure to polycyclic aromatic hydrocarbons, certain metals, asbestos, and crystalline silica, as well as exposure cir- cumstances relevant to certain categories of work, and expo- sure to outdoor air pollution, and specifically to particulate matter and diesel engine exhaust, and to indoor air pollution, including second-hand tobacco smoke and emissions from household combustion of coal.
• Historically, small cell lung car- cinoma has been distinguished from non-small cell lung carcino- ma, which includes the histologi- cal types of adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. Further subtyping is increasingly done on a molecular basis.
• Adenocarcinomas have driver mutations, most commonly involv- ing EGFR and KRAS mutations
5.1 5 ORGAN SITE
Lung cancer Elisabeth Brambilla William D. Travis
Paul Brennan (reviewer) Curtis C. Harris (reviewer) José Rogelio Pérez Padilla (reviewer)
as well as ALK fusions; squa- mous cell carcinomas have molecular alterations involv- ing SOX2, TP63, FGFR1, and DDR2, and most small cell carci- nomas express neuroendocrine markers and have TP53 and RB1 mutations. Lung cancers from smokers exhibit up to 10 times as many mutations as tumours from never-smokers.
• Screening for lung cancer is un- der development.
• Lung cancer is one of the most aggressive human cancers, with a 5-year overall survival of 10–15%.
There are four major histological types of lung cancer: adenocarci- noma, squamous cell carcinoma, small cell carcinoma, and large cell carcinoma. These tumours are defined primarily by morphology; however, in the past 10 years these tumours have started to be classified according to immunohistochemical and genetic characteristics as well.
Etiology The most common cause of lung cancer is cigarette smoking; tobacco smoke is the most intensively inves- tigated of all known carcinogens [1].
Local forms of tobacco smoking, such as bidis in India (Fig. 5.1.1), are also important. Risk is often related to the product of smoking rate and duration of smoking (“pack-years”), but more precise specification is possible. The lung cancer incidence rate is proportional to the fourth pow- er of age in never-smokers, and the excess in smokers is proportional to the fourth power of smoking dura- tion multiplied by the number of ciga- rettes smoked per day [2]. Recent analysis based on European and Canadian cases indicated that ad- enocarcinoma was the most preva- lent lung cancer subtype in never- smokers and in women; squamous cell carcinoma predominated in male smokers. Smoking exerted a steeper risk gradient on squamous cell car- cinoma and small cell lung cancer than on adenocarcinoma [3]. Diverse indicators of injury to respiratory tis- sue likely related to tumorigenesis are evident before the onset of clini- cal disease (see “Earliest molecular evidence of tobacco-induced injury in the airway”). Smoking cessation efforts have resulted in a growing percentage of lung cancers among nonsmokers, particularly those with adenocarcinoma histology.
Although smoking is estimated to account for about 90% of lung cancer cases in high-income coun- tries, a wide range of agents are
Chapter 5.1 • Lung cancer 351
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
Epidemiology
Lung cancer • Lung cancer remains the most
frequent cancer worldwide. There were more than 1.8 million new cases (13% of total cancer incidence) and almost 1.6 mil- lion deaths (20% of total cancer mortality), as estimated in 2012. More than one third of all newly diagnosed cases occurred in China.
• Lung cancer is the leading cause of cancer death in men in 87 countries and in women in 26 countries. Age-standardized rates vary 80-fold internationally, are highest in North America, Europe, and East Asia, and tend to still be relatively low in many African countries and some Asian countries.
• Due to a high and rather stable case fatality rate, patterns and trends for mortality rates are sim- ilar to those for incidence rates, irrespective of level of resource within a given country.
• Recent trends in lung cancer re- flect the evolution of the smok- ing epidemic. In men, incidence rates have peaked in a number of highly developed countries at a late stage of the tobacco epi- demic, while rates continue to rise among women.
• Only in a few countries (Australia and the USA), where the tobacco epidemic is most advanced and smoking prevalence has been declining for several decades, are there recent downward in- cidence trends among women.
Map 5.1.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for lung cancer in men, 2012.
Map 5.1.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for lung cancer in women, 2012.
recognized to cause lung cancer, and for others the evidence is less than definitive. A review of IARC Group 1 carcinogens encountered at highest concentrations in an occupational context predominantly involved cau- sation of lung cancer as covered in, although not restricted to, two vol- umes of the IARC Monographs. One
volume considered arsenic, met- als, fibres, and dusts [4]. The other considered mainly organic com- pounds [5], not only specific chemi- cals but also certain categories of employment, including work as a painter and work in the rubber in- dustry. In an analysis of the burden of occupational lung cancer in the
United Kingdom, asbestos exposure contributed by far the largest number of cases [6]. Radon is also identified as a risk factor for lung cancer [7].
An editorial, prompted by publi- cation of a meta-analysis based on 17 cohorts in nine European studies, specified that air pollution may be recognized as a cause of lung cancer.
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
352
Chart 5.1.1. Estimated global number of new cases and deaths with proportions by major world regions, for lung cancer in both sexes combined, 2012
Chart 5.1.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for lung cancer in men, circa 1975–2012.
Chart 5.1.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for lung cancer in women, circa 1975–2012.
Chapter 5.1 • Lung cancer 353
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
Fig. 5.1.1. Older men smoking tobacco in India. Bidis, made of coarse and uncured tobacco, account for about 60% of smoked tobacco products in India, whereas ciga- rettes account for 20%.
The data included a statistically sig- nificant association between risk of lung cancer and particulate matter with a diameter less than 10 µm (PM10) (hazard ratio, 1.22; confi- dence interval, 1.03–1.45) [8]. IARC recently concluded that long-term exposure to outdoor air pollution, and specifically particulate matter in outdoor air, causes lung cancer [9]. Such findings involve outdoor air pollution (see “Biomarkers of air pollution”). Categories of air pollu- tion recognized to cause lung cancer also include exposure to second- hand tobacco smoke and to emis- sions from household combustion of coal [10,11].
Using defined terminology, the World Cancer Research Fund re- ported convincing evidence for causation of lung cancer by arsenic in drinking-water and by β-carotene supplements, as well as probable ev- idence of decreased risk consequent upon consumption of fruits and of foods containing carotenoids [12]. Rare familial cases occur, and com- pared with nonfamilial cases these patients present with earlier patho- logical stage and more cancers with adenocarcinoma subtype [13].
Pathology and genetics Historically, small cell lung carcino- ma has been distinguished from non- small cell lung carcinoma (NSCLC), which includes the histological types of adenocarcinoma, squamous cell carcinoma, and large cell carcino- ma. However, in the past decade, the distinction between adenocarci- noma and squamous cell carcinoma has been increasingly recognized because of major differences in ge- netics and also in responses to spe- cific therapies [14]. Therefore, lung cancers are increasingly classified according to molecular subtypes, predicated on particular genetic alterations that drive and maintain lung tumorigenesis. Such driver mutations, and the associated con- stitutively active mutant signalling proteins, are critical to tumour cell survival, leading to the development of novel targeted therapies. These
driver mutations are best established for adenocarcinoma, and more re- cently a list of oncogenic drivers for squamous cell carcinoma has been discerned, allowing the identifica- tion of potential targeted therapies. Nevertheless, for the majority of lung cancers, there is as yet no clear first- line chemotherapy targeting a driver signature.
A genetic basis of susceptibility to lung cancer generally, as distinct from susceptibility to particular tu- mour types, has been extensively investigated. Large, collaborative genome-wide association studies have identified three separate loci that are associated with lung cancer (5p15, 6p21, and 15q25) and that include genes regulating nicotinic acetylcholine receptors and telo- merase production. However, much about genetic risk remains to be discovered, and rarer gene vari- ants, such as those of the CHEK2 gene, likely account for most of the remaining genetic susceptibility [15]. Studies of single-nucleotide poly- morphisms, based on carcinogen metabolism and DNA repair mediat- ing susceptibility to tobacco smoke- induced carcinogenesis, have pro- vided limited insight.
Lung adenocarcinoma Adenocarcinomas represent 40% of all lung cancers in the USA [16]. Most lung adenocarcinomas are diagnosed in the periphery of the lung (Fig. 5.1.2). Major changes to the 2004 WHO classification are recommended in light of the 2011 International Association for the Study of Lung Cancer, American Thoracic Society, and European Respiratory Society classification of lung adenocarcinoma [17]. The term “bronchioloalveolar carcinoma” is no longer used.
In small biopsies or cytology specimens, a tumour is classified as adenocarcinoma if it shows clear glandular morphology such as acinar, papillary, lepidic, or solid with mucin patterns. In tumours that lack any clear adenocarcinoma or squa- mous cell carcinoma morphology, the tumour requires further evalu- ation with immunohistochemistry through a limited work-up using a single adenocarcinoma marker (i.e. TTF-1) and squamous marker (i.e. p63 or p40). If a tumour is positive for TTF-1 but negative for a squa- mous marker, it is classified as “NSCLC, favour adenocarcinoma” [17]. However, if no adenocarcino- ma or squamous differentiation is
354
Earliest molecular evidence of tobacco-induced injury in the airway Avrum Spira
Cigarette smoke creates a molecular “field of injury” in all airway epithelial cells exposed to the toxins associ- ated with that exposure (Fig. B5.1.1). It has been known from early au- topsy observations of cellular atypia throughout the airways of smokers that the cellular injury produced by smoking involves the whole respi- ratory tract. Several studies have shown that bronchial airway epithe- lial cells of current and former smok- ers with and without lung cancer dis- play allelic loss, TP53 mutations, and changes in promoter methylation and in telomerase activity of non- cancerous bronchial epithelial cells [1]. More recently, multiple groups have demonstrated that genome- wide gene expression profiling of bronchial airway epithelial cells collected by brushing at the time of bronchoscopy reflects the physi- ological response to and damage
from exposure to cigarette smoke. The smoking-related changes in the airway transcriptome may be regu- lated, in part, by changes in DNA methylation and microRNA expres- sion in response to the toxic expo- sure. While most of these genomic alterations (including those involved in xenobiotic metabolism and oxida- tive stress) reverse within months of smoking cessation, a subset of these transcriptomic changes are irreversible among former smokers (including those who have stopped smoking for > 10 years) and may be associated with the increased risk of smoking-related lung disease in former smokers.
Based on the heterogeneity in the bronchial airway genomic re- sponse to smoking, gene expres- sion biomarkers in the cytologically normal bronchial airway epithelium have been developed for the early
detection of lung cancer in current or former smokers being evaluated for clinical suspicion of disease [2]. These gene expression alterations in the airway have also been shown to reflect activation of oncogenic signalling pathways among high- risk smokers that can be reversed with chemopreventive agents that target these pathways. Furthermore, methylation changes in the promoter region of cancer-related genes mea- sured in sputum have been shown to be associated with lung cancer risk among smokers. This molecular field of injury has recently been extended to chronic obstructive pulmonary disease, where gene expression alterations in the bronchial airway have been found among smokers with obstructive airway disease.
These observations suggest that the entire respiratory tree is af- fected by cigarette smoke, and that easily obtainable upper airway cells might provide insight into the types and degree of epithelial cell injury that have occurred in an individual smoker. Recent studies have indi- cated that the gene expression re- sponses to smoking in the bronchial airway also occur in the extratho- racic airway epithelium that lines the mouth and nose [3]. This has led to the promise of genomic biomarkers of tobacco exposure and disease risk that can be developed in these non-invasive biosamples and can be applied to large-scale population- based studies.
References 1. Steiling K et al. (2008). Cancer Prev
Res (Phila), 1:396–403. http://dx.doi. org/10.1158/1940-6207.CAPR-08-0174 PMID:19138985
2. Spira A et al. (2007). Nat Med, 13:361– 366. http://dx.doi.org/10.1038/nm1556 PMID:17334370
3. Zhang X et al. (2010). Physiol Genomics, 41:1–8. http://dx.doi.org/10.1152/physiol genomics.00167.2009 PMID:19952278
Fig. B5.1.1. The molecular field of injury that is induced among airway epithelial cells throughout the respiratory tract. miRNA, microRNA; mRNA, messenger RNA; SNPs, single-nucleotide polymorphisms.
Chapter 5.1 • Lung cancer 355
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
identified by morphology or immuno- histochemical staining, the tumour is classified as “NSCLC - not other- wise specified”. This is necessary to preserve as much tissue as possible for molecular testing.
Adenocarcinoma in situ is an ad- enocarcinoma of pure lepidic growth without invasion that measures 3 cm or less. Minimally invasive adeno- carcinoma is a lepidic-predominant adenocarcinoma measuring 3 cm or less that has 5 mm or less of an in- vasive component [17]. Invasive ad- enocarcinomas are now classified according to the predominant sub- type. This classification is based on comprehensive histological sub- typing by estimating different his- tological patterns in a semi-quanti- tative manner in 5–10% increments (Fig. 5.1.3).
Multiple studies have demon- strated consistent correlations of survival with the predominant sub- types, with very favourable (adeno- carcinoma in situ, minimally invasive adenocarcinoma, lepidic), interme- diate (acinar, papillary), and poor (solid, micropapillary) prognostic categories. As expected, so far all cases reported as adenocarcinoma in situ and minimally invasive ad- enocarcinoma have demonstrated 5-year disease-free survival of 100% [14].
The majority of lung adenocar- cinomas from never-smokers har- bour a mutation in either EGFR or
HER2 (ERBB2) or a fusion involv- ing ALK or ROS1 [18]. Driver muta- tions in adenocarcinoma occur in the EGFR, HER2 (ERBB2), KRAS, ALK, BRAF, PIK3CA, and ROS1 genes (Fig. 5.1.4A). Mutations in KRAS and EGFR are mutually ex- clusive, as are most of the kinase domain mutations listed above, ex- cept for PIK3CA [19,20]. Several of these genes may have increased gene copy number or amplification, mostly of the mutant allele. EGFR (20%), HER2 (2%), and MET (1%) may be amplified preferentially on the mutant allele. MET ampli- fication occurs more frequently in
resistant disease and is associated with EGFR tyrosine kinase inhibi- tion. All of these mutations offer the possibility of targeted therapy, although only three (EGFR muta- tion, ALK and ROS1 fusion) are subject to targeted treatment with currently approved drugs: gefitinib or erlotinib for EGFR mutation and crizotinib for ALK and ROS1 fusion. Agents targeting the others are in development in phase 2 or phase 3 (Table 5.1.1).
Whole-exome and whole-ge- nome sequencing of DNA from 183 pairs of lung adenocarcinoma tu- mours and matched normal tissues
Fig. 5.1.2. Gross appearance of lung ad- enocarcinoma. The tumour is situated in the periphery of the lung, involving the sub- pleural tissue with a tan grey cut surface.
Fig. 5.1.3. Adenocarcinoma histology. (A) Lepidic pattern: atypical pneumocytes line the alveolar walls with no invasive component. (B) Acinar pattern: cytologically malig- nant tumour cells form round to oval-shaped glands. (C) Papillary pattern: malignant glandular cells grow along the surface of fibrovascular cores. (D) Micropapillary pattern: these tumour cells form papillae lacking fibrovascular cores. (E) Solid pattern: these malignant cells are growing in sheets with hyperchromatic nuclei. (F) Invasive mucinous adenocarcinoma: malignant cells grow in acinar and lepidic patterns. The tumour cells have abundant apical mucin and small, basally oriented nuclei.
356
identified, among many other find- ings, statistically recurrent somatic mutations in the splicing factor gene U2AF1 and truncating mutations af- fecting RBM10 and ARID1A [21].
Adenocarcinomas from never- smokers have a higher frequency of EGFR mutations (38% vs 14%) and ALK rearrangements (12% vs 2%) compared with tumours from former or current smokers [22]. Such a com- parison in respect of NSCLC and ad- jacent normal tissue from 17 patients identified 3726 point mutations and more than 90 insertions or deletions in the coding sequence, with an av- erage mutation frequency more than 10 times higher in smokers than in never-smokers (Fig. 5.1.5). Novel alterations were identified in genes involved in chromatin modification and DNA repair pathways, along with DACH1, CFTR, RELN, ABCB5, and HGF [23].
Squamous cell lung carcinoma Squamous cell carcinoma accounts for approximately 20% of all lung
cancers in the USA [16]. At diagno- sis, most squamous cell carcinomas are located in the central portion of the lung, but in recent years, the fre- quency of tumours located peripher- ally appears to have been increasing.
In resection specimens, tu- mours of this type are subclassi- fied as squamous cell carcinoma, basaloid, clear cell, and small cell variants. Squamous differentiation is identified morphologically as intercellular bridging, squamous pearl formation, and individual cell keratinization (Fig. 5.1.6). In small biopsies or cytology specimens, in poorly differentiated tumours that lack clear squamous morphology, if the tumour is positive for p63 or p40 but negative for TTF-1, the tumour is classified as “NSCLC, favour squa- mous cell carcinoma”.
Genetic alterations in squamous cell carcinoma have been revealed as part of the Cancer Genome Atlas; the data involve genomic and epige- nomic alterations in 178 squamous cell lung carcinomas. Most of the previously identified driver mutations were confirmed and several others discovered [24,25]. Although squa- mous cell carcinomas have been regarded as tumours without spe- cific molecular abnormalities, new discoveries allow most squamous cell lung cancers to be characterized with reference to available targeted therapies (Fig. 5.1.4B). If mutations in PIK3CA (20–30%), FGFR1 (20%), and DDR2 (4%) are considered,
Fig. 5.1.4. (A) Mutation spectrum in adenocarcinoma. Mutations in TP53 and STK11/LKB1 are common occurrences, not included in the pie chart due to high overlap with other mutations. (B) Mutation spectrum in squamous cell carcinoma. Mutations in TP53 are a common occurrence, not included in the pie chart due to high overlap with other mutations.
Table 5.1.1. Molecular targeted therapies for lung cancer
Gene or genetic alteration Histological type Therapy
Targeted therapiesa
EGFR mutation Advanced adenocarcinoma EGFR tyrosine kinase inhibitor Erlotinib, gefitinib
ALK fusion Adenocarcinoma Crizotinib
ROS1 fusion Advanced adenocarcinoma Crizotinib
Drugs in development
HER2 Adenocarcinoma Afatinib (BIBW 2992)
PI3KCA NSCLC Trastuzumab, PI3K inhibitor
BRAF NSCLC Sorafenib?
MET NSCLC Phase 2: rilotumumab (AMG 102), MetMAb
RAS Adenocarcinoma PI3K + MEK inhibitors
NSCLC, non-small cell lung carcinoma; PI3K, phosphatidylinositol 3-kinase. a Approved by the United States Food and Drug Administration.
Chapter 5.1 • Lung cancer 357
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
about one half of squamous cell lung carcinomas have genetic alterations that are potential molecular targets. The most frequent mutations in tumours from patients were in TP53, CDKN2A (P16), PTEN, PIK3CA, KEAP1, and MLL2 [24,25]. Conspicuous regions of somatic copy number alteration, includ- ing amplification, were evident, as had been previously detected for SOX2, PI3KCA, and FGFR1 [26], together with deletion at CDKN2A [24]. Other regions of marked gene copy alteration were discovered on chromosomal segments contain- ing NFE2L2, MYC, CDK6, MDM2, FOXP1, PTEN, and NF1 [24].
Interestingly, many of the so- matic alterations identified in squa- mous cell carcinoma have been recognized as drivers of initiation and tumour progression. Two path- ways affected by genetic alteration
were the oxidative stress response pathway and the squamous differ- entiation pathway. The oxidative stress response pathway, involving NFE2L2 and KEAP1 or CUL3, was evident in 30% of the cases, and the alterations described were mutually exclusive. The squamous differ- entiation pathway was affected by gene alterations in 44% of cases; the affected genes included SOX2 and TP63 (which were mutually exclusive in relation to Notch1 or Notch2 ALSCL4 loss-of-function mutations) and focal deletions of FOXP1. Messenger RNA (mRNA) expression profiling indicated altered expression, enhancing the recogni- tion of these alterations as oncogen- ic driver events. Therapeutic agents for DDR2 mutation (dasatinib) and FGFR1 amplification (FGFR1 ty- rosine kinase inhibitor) are already available, and United States Food and Drug Administration (FDA) ap- proval is currently anticipated for a target for PIK3CA.
Small cell lung carcinoma Small cell lung cancer accounts for 15% of all lung cancers in the USA
Fig. 5.1.5. Graphical summary of the distinction, including extent of genomic mutation, evident when lung cancer and adjacent normal tissues from smokers and never-smokers are compared. (For details, see [23].)
Fig. 5.1.6. Squamous cell carcinoma. The tumour consists of nests of malignant cells with abundant eosinophilic cytoplasm. Abundant extracellular keratin is present.
358
Biomarkers of air pollution Paolo Vineis Air pollution can induce cancer (par- ticularly of the lung) through several mechanisms, including damage to DNA. Measurement of biomarkers related to air pollution can improve the investigation of the health ef- fects, by facilitating improved expo- sure assessment and increased un- derstanding of mechanisms, thereby providing biological plausibility.
Demetriou et al. [1] have re- viewed biomarkers for which there is consistent evidence in the literature to support the results of epidemio- logical studies on the effects of am- bient air pollution. Epidemiological evidence from the selected studies has been assessed using a set of criteria that evaluate their credibility. The criteria take into account the to- tal number of subjects investigated, the degree of replication of findings across studies, and potential protec- tion from bias and/or confounding. Overall, 524 papers were evaluated, which described the relationships between ambient air pollution and biomarkers of dose and early re- sponse. Studies were ranked from A (strong) to C (weak). The biomarkers that fulfilled the criteria to achieve A or B scores were 1-hydroxypyrene, DNA adducts, chromosomal aberra- tions, micronuclei, oxidative damage
Fig. B5.1.2. Biological markers of exposure and effects of air pollution.
to nucleobases, and methylation changes. These biomarkers cover the whole spectrum of disease onset and progression, from external expo- sure to tumour formation, and some have also been suggested as risk predictors of future cancer, reinforc- ing causal reasoning (Fig. B5.1.2).
Overall, existing biomarkers of dose and effect appear to reinforce the causal nature of the associa- tion between air pollution and lung cancer, although the markers in the review were not all specific to lung carcinogenesis. DNA adducts, chromosomal aberrations, micro- nuclei, and oxidized nucleobase markers have been suggested to be predictive for the risk of cancer. 1-Hydroxypyrene is a marker of in- ternal dose, DNA adducts and oxi- dized nucleobases are markers of the biologically effective dose, and micronuclei, chromosomal aberra- tions, and DNA methylation are good markers of early biological effect. The available evidence is stronger for oxidized nucleobase markers, and the mechanisms supported by these biomarkers are likely to be central to the biological process of air pollution-induced lung cancer.
However, certain aspects of biomarkers used in epidemiological
studies need to be clarified. These include their reliability, including inter-laboratory as well as inter- technique variation.
Although chronic inflammation is probably relevant to particle- induced lung carcinogenesis, the overall evidence is still relatively scant. Exposure to air pollutants has been associated with acute inflammation in the airways and to elevated levels of systemic markers of inflammation, such as C-reactive protein and fibrinogen. Recent stud- ies found that medium-term expo- sure to traffic-related air pollution may induce an increased inflamma- tory/endothelial response, especial- ly among people with diabetes [2]. So far the inflammatory response has mainly been associated with risk of cardiovascular diseases rather than of cancer [3].
References 1. Demetriou CA et al. (2012). Occup
Environ Med, 69:619–627. http://dx.doi. org/10.1136/oemed-2011-100566 PMID: 22773658
2. Alexeeff SE et al. (2011). Environ Health Perspect, 119:481–486. http://dx.doi.org/ 10.1289/ehp.1002560 PMID:21349799
3. Frampton MW (2006). Clin Occup Environ Med, 5:797–815. PMID:17110293
Chapter 5.1 • Lung cancer 359
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
[16]. Most cases of this tumour type are diagnosed on the basis of a peri- hilar mass. Tumour cells are small in size with a round to fusiform shape, scant cytoplasm, finely granular nuclear chromatin, and absent or inconspicuous nucleoli (Fig. 5.1.7) [27]. Necrosis is usually extensive and mitotic rates are high, averag- ing 80 mitoses per 2 mm2 area [27]. Small cell lung cancer is reliably di- agnosed in small biopsies and cytol- ogy specimens.
Small cell lung cancer typically occurs in heavy smokers and is char- acterized by the most aggressive growth among lung malignancies. No single molecularly targeted drug has yet shown any clinical activity in small cell lung cancer. Until recently, the relevant molecular genetics was poorly characterized except for the prevalence of inactivating muta- tions in TP53 and RB1. However, the results of global lung cancer genome research have now been published, including the results of single-nucleotide polymorphism ar- ray analysis, transcriptome sequenc- ing, and full-genome sequencing of a total of 63 cases (Table 5.1.2) [28].
In addition to invariable p53/Rb in- activation, mutations were found in CREBBP, EP300, and MLL, which all encode histone modifiers. Also detected were mutations in PTEN, SLIT2, and EPHA7, as well as focal amplification of the FGFR1 tyrosine kinase gene, thus identifying further oncogenic driver events nominally susceptible to targeted therapy that
may be developed. This is an exam- ple of how integrative computational genome analysis can provide func- tionally tractable information in the context of a highly mutated cancer genome, due to the genotoxic effects of tobacco smoke.
Large cell lung carcinoma Large cell carcinoma accounts for 3% of all lung cancers in the USA [16]. Large cell carcinoma consists of sheets and nests of large polygo- nal cells with vesicular nuclei and prominent nucleoli and is diagnosed on the basis of exclusion to the ex- tent that the presence of squamous cell or glandular differentiation can be excluded by light microscopy [27]. The diagnosis of large cell car- cinoma cannot be made on small bi- opsies or cytology specimens and requires a resection specimen.
Prospects Prevention The primary approach to preven- tion is smoking cessation, which has been proven effective in both men and women. This subject is not discussed here but is sepa- rately addressed in this Report (see Chapter 4.1). Encouragement of smoking cessation by women has yet to make a major impact in Europe, where lung cancer mortality
Fig. 5.1.7. Small cell carcinoma. The tumour consists of a cellular sheet of small tumour cells with scant cytoplasm, finely granular nuclear chromatin, and frequent mitoses, and nucleoli are absent.
Table 5.1.2. Genetic alterations in small cell lung carcinoma
Significantly altered (one or more alleles) genes (drivers) Mode Percentage
TP53 Inactivation 100
RB1 Inactivation 100
CREBBP Mutation 18
EP300 Mutation 18
MLL Mutation 10
PTEN Mutation 10
SLIT2 Mutation 10
EPHA7 Mutation 5
FGFR1 Amplification 4
MYCL Amplification 16
E2F2 Amplification 5
CCN2 Amplification 5
360
rates are still increasing in women. Prevention of occupational cancer and regulatory measures to ad- dress air pollution are discussed in Chapter 4.5. In a region of China, decreased risk of lung cancer was associated with improvement of household stoves by adding a chim- ney to reduce exposure to smoky coal emissions [29].
Screening Among options to reduce mortality from lung cancer is the introduc- tion of population-based screening. Several small randomized trials of low-dose computed tomography (CT) screening are under way in Europe. From the USA, a decisive outcome was reported for the National Lung Screening Trial [30] conducted from 2002 to 2004 and involving 53 454 individuals screened either with low-dose CT or with chest radiography. The in- cidence of lung cancer in the two groups was identical. The relative reduction in mortality from lung can- cer through low-dose CT screening was 20% (P = 0.004). The rate of death from any cause was reduced in the low-dose CT group by 6.7% (P = 0.02) compared with the radi- ography group, indicating that the screening was essentially reduc- ing deaths from lung cancer. The United States Preventive Services Task Force has published a rec- ommendation that CT screening
be available to healthy individuals aged 55 –79 years with at least a 30 pack-year history of smoking and who have smoked within the past 15 years – estimated at about 10 million people.
Limitations relevant to data from screening trials specifically in Europe are the healthy volunteer effect, which may bias results, the possibility that scanners currently used are technologically more ad- vanced than those used during the trial, and the fact that the United States trial mentioned above was conducted in high-profile medical institutions, recognized for their expertise in radiology diagnosis and treatment of cancer. When the cause of death in the two arms of the trial was reduced to “no death by lung cancer”, the reduction in overall mortality was 3.2%, which was not statistically significant. Obviously, the cost-effectiveness of low-dose CT screening must also be consid- ered in the context of competing interventions, particularly smoking cessation.
Other strategies for early detec- tion of lung cancer, including par- ticular molecular markers in blood, sputum, and urine, are being stud- ied in specimens obtained as part of the American College of Radiology Imaging Network and related activ- ity, but are not yet finalized. There is unfortunately no proven biomarker for lung cancer detection [31].
Prognosis Lung cancer is one of the most aggressive human cancers, with a 5-year survival of 10–15% [16]. Patients with clinical stage IV lung cancer have a 5-year overall sur- vival of 2%, and even for the lowest stage IA, 5-year overall survival is only 50% [32]. Lung cancer accounts for 13% of all cancers and 20% of all cancer deaths. In the USA, lung cancer is the second most common cancer in both men and women, ac- counting for 14% of cancer cases in each sex, and is the most common cause of cancer mortality in both men and women, accounting for 29% and 26% of cancer deaths, re- spectively [33]. Poor survival of lung cancer patients is due, at least in part, to 80% of patients being diag- nosed with metastatic disease and more than half of patients having dis- tant metastases [16]. Advanced lung cancer has also been resistant to traditional chemotherapy. However, recent advances have led to exciting progress in therapies that are de- pendent on histology and genetics. The level of scrutiny is exemplified by trials of adjuvant chemotherapy designed to differentiate not only between mutations in codons 12 and 13 of KRAS, but also between different amino acid substitutions as determined by particular mutations at codon 12 [34].
References 1. IARC (2012). Personal habits and indoor
combustions. IARC Monogr Eval Carcinog Risks Hum, 100E:1–575. PMID:23193840
2. Peto J (2012). That the effects of smok- ing should be measured in pack-years: misconceptions 4. Br J Cancer, 107:406– 407. http://dx.doi.org/10.1038/bjc.2012.97 PMID:22828655
3. Pesch B, Kendzia B, Gustavsson P et al. (2012). Cigarette smoking and lung can- cer – relative risk estimates for the major histological types from a pooled analy- sis of case-control studies. Int J Cancer, 131:1210–1219. http://dx.doi.org/10.1002/ ijc.27339 PMID:22052329
4. Straif K, Benbrahim-Tallaa L, Baan R et al.; WHO International Agency for Research on Cancer Monograph Working Group (2009). A review of human carcinogens – Part C: met- als, arsenic, dusts, and fibres. Lancet Oncol, 10:453–454. http://dx.doi.org/10.1016/S14 70-2045(09)70134-2 PMID:19418618
5. Baan R, Grosse Y, Straif K et al.; WHO International Agency for Research on Cancer Monograph Working Group (2009). A review of human carcinogens – Part F: chemical agents and related occupations. Lancet Oncol, 10:1143–1144. http://dx.doi. org /10.1016/S1470 -20 45(0 9)7035 8 - 4 PMID:19998521
6. Brown T, Darnton A, Fortunato L, Rushton L; British Occupational Cancer Burden Study Group (2012). Occupational cancer in Britain. Respiratory cancer sites: larynx, lung and mesothelioma. Br J Cancer, 107 Suppl 1:S56–S70. http://dx.doi.org/10.1038/ bjc.2012.119 PMID:22710680
7. IARC (2012). Radiation. IARC Monogr Eval Carcinog Risks Hum, 100D:1– 437. PMID:23189752
8. Yorifuji T, Kashima S (2013). Air pollution: another cause of lung cancer. Lancet Oncol, 14:788–789. http://dx.doi.org/10.1016/S14 70-2045(13)70302-4 PMID:23849839
Chapter 5.1 • Lung cancer 361
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
9. Loomis D, Grosse Y, Lauby-Secretan B et al. (2013). The carcinogenic- ity of outdoor air pollution. Lancet Oncol, 14:1262–1263. http://dx.doi.org/10.1016/ S1470-2045(13)70487-X
10. Secretan B, Straif K, Baan R et al.; WHO International Agency for Research on Cancer Monograph Working Group (2009). A review of human carcinogens – Part E: tobacco, areca nut, alcohol, coal smoke, and salted fish. Lancet Oncol, 10:1033– 1034. http://dx.doi.org/10.1016/S1470-204 5(09)70326-2 PMID:19891056
11. Barone-Adesi F, Chapman RS, Silverman DT et al. (2012). Risk of lung cancer associ- ated with domestic use of coal in Xuanwei, China: retrospective cohort study. BMJ, 345:e5414. http://dx.doi.org/10.1136/bmj. e5414 PMID:22936785
12. World Cancer Research Fund/American Institute for Cancer Research (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research.
13. Haraguchi S, Koizumi K, Mikami I et al. (2012). Clinicopathological characteris- tics and prognosis of non-small cell lung cancer patients associated with a fam- ily history of lung cancer. Int J Med Sci, 9:68–73. http://dx.doi.org/10.7150/ijms.9.68 PMID:22211092
14. Travis WD, Brambilla E, Riely GJ (2013). New pathologic classification of lung can- cer: relevance for clinical practice and clini- cal trials. J Clin Oncol, 31:992–1001. http:// dx.doi.org/10.1200/JCO.2012.46.9270 PMID:23401443
15. Brennan P, Hainaut P, Boffetta P (2011). Genetics of lung-cancer susceptibility. Lancet Oncol, 12:399– 408. http://dx.doi. o rg /10.1016 /S1470 -2 0 4 5 (10)7012 6 -1 PMID:20951091
16. Howlader N, Noone AM, Krapcho M et al. (2012). Lung cancer. In: SEER Cancer Statistics Review, 1975-2009 (Vintage 2009 Populations). Bethesda, MD: National Cancer Institute. Available at http://seer. cancer.gov/csr/1975_2009_pops09/.
17. Travis WD, Brambilla E, Noguchi M et al. (2011). International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society in- ternational multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol, 6:244–285. http://dx.doi.org/10.1097/JTO.0b 013e318206a221 PMID:21252716
18. Pao W, Hutchinson KE (2012). Chipping away at the lung cancer genome. Nat Med, 18:349–351. http://dx.doi.org/10.1038/nm. 2697 PMID:22395697
19. Pao W, Girard N (2011). New driver mu- tations in non-small-cell lung cancer. Lancet Oncol, 12:175–180. http://dx.doi. org /10.1016 /S1470 -20 4 5(10)70 0 87- 5 PMID:21277552
20. Ding L, Getz G, Wheeler DA et al. (2008). Somatic mutations affect key pathways in lung adenocarcinoma. Nature, 455:1069– 1075. http://dx.doi.org/10.1038/nature07423 PMID:18948947
21. Imielinski M, Berger AH, Hammerman PS et al. (2012). Mapping the hallmarks of lung adenocarcinoma with massively parallel sequencing. Cell, 150:1107–1120. http://dx.doi.org/10.1016/j.cell.2012.08.029 PMID:22980975
22. Paik PK, Johnson ML, D’Angelo SP et al. (2012). Driver mutations determine survival in smokers and never-smokers with stage IIIB/IV lung adenocarcinomas. Cancer, 118:5840–5847. http://dx.doi.org/10.1002/ cncr.27637 PMID:22605530
23. Govindan R, Ding L, Griffith M et al. (2012). Genomic landscape of non-small cell lung cancer in smokers and never-smokers. Cell, 150:1121–1134. http://dx.doi.org/10.1016/j. cell.2012.08.024 PMID:22980976
24. Hammerman PS, Hayes DN, Wilkerson MD et al.; Cancer Genome Atlas Research Network (2012). Comprehensive genomic characterization of squamous cell lung can- cers. Nature, 489:519–525. http://dx.doi. org/10.1038/nature11404 PMID:22960745
25. Drilon A, Rekhtman N, Ladanyi M, Paik P (2012). Squamous-cell carcinomas of the lung: emerging biology, controver- sies, and the promise of targeted therapy. Lancet Oncol, 13:e418–e426. http://dx.doi. o rg /10.1016 /S1470 -2 0 4 5 (12)702 91-7 PMID:23026827
26. Weiss J, Sos ML, Seidel D et al. (2010). Frequent and focal FGFR1 amplification associates with therapeutically tractable FGFR1 dependency in squamous cell lung cancer. Sci Transl Med, 2:62ra93. http:// dx.doi.org/10.1126/scitranslmed.3001451 PMID:21160078
27. Travis WD, Brambilla E, Müller-Hermelink HK, Harris CC, eds (2004). Pathology and Genetics of Tumours of the Lung, Pleura, Thymus and Heart. Lyon: IARC.
28. Peifer M, Fernández-Cuesta L, Sos ML et al. (2012). Integrative genome analy- ses identify key somatic driver mutations of small-cell lung cancer. Nat Genet, 44:1104–1110. http://dx.doi.org/10.1038/ ng.2396 PMID:22941188
29. Lee KM, Chapman RS, Shen M et al. (2010). Differential effects of smoking on lung cancer mortality before and after household stove improvement in Xuanwei, China. Br J Cancer, 103:727–729. http:// dx.doi.org/10.1038/sj.bjc.6605791 PMID: 20648014
30. Aberle DR, Adams AM, Berg CD et al.; National Lung Screening Trial Research Team (2011). Reduced lung-cancer mortal- ity with low-dose computed tomographic screening. N Engl J Med, 365:395– 409. http://dx.doi.org/10.1056/NEJMoa1102873 PMID:21714641
31. Hassanein M, Callison JC, Callaway-Lane C et al. (2012). The state of molecular biomarkers for the early detection of lung cancer. Cancer Prev Res (Phila), 5:992– 1006. http://dx.doi.org/10.1158/1940-6207. CAPR-11-0441 PMID:22689914
32. Goldstraw P, Crowley J, Chansky K et al.; International Association for the Study of Lung Cancer International Staging Committee; Participating Institutions (2007). The IASLC Lung Cancer Staging Project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edi- tion of the TNM Classification of Malignant Tumours. J Thorac Oncol, 2:706–714. http:// dx.doi.org/10.1097/JTO.0b013e31812f3c1a PMID:17762336
33. Siegel R, Naishadham D, Jemal A (2012). Cancer statistics, 2012. CA Cancer J Clin, 62:10–29. http://dx.doi.org/10.3322/caac. 20138 PMID:22237781
34. Shepherd FA, Domerg C, Hainaut P et al. (2013). Pooled analysis of the prognos- tic and predictive effects of KRAS muta- tion status and KRAS mutation subtype in early-stage resected non-small-cell lung cancer in four trials of adjuvant che- motherapy. J Clin Oncol, 31:2173–2181. http://dx.doi.org/10.1200/JCO.2012.48.1390 PMID:23630215
362
Summary • Breast cancer is the most com-
mon cancer in women world- wide. This chapter is concerned with female breast cancer.
• Well-characterized breast cancer risk factors include age, family his- tory, reproductive factors, mam- mographic density, and atyp- ia in a prior benign breast biopsy. Agents that cause breast cancer include alcohol consumption, use of combined estrogen–pro- gestogen contraceptives and menopausal therapy, and expo- sure to X- and γ-radiation.
• A small proportion of breast cancers are due to inherited mutations in high-penetrance breast cancer susceptibility genes (BRCA1 and BRCA2). Several lower-penetrance genes have also been associated with breast cancer, and there are many loci that are linked to an increased risk.
• Recent molecular and genetic studies have emphasized that breast cancer is a highly het- erogeneous group of diseases that differ in their prognosis and response to treatment.
• Improved understanding of the molecular pathways and genetic
5.2 5 ORGAN SITE
Breast cancer Stuart J. Schnitt Sunil R. Lakhani
Benjamin O. Anderson (reviewer) Beela Sarah Mathew (reviewer) Thangarajan Rajkumar (reviewer)
alterations that underlie the dif- ferent breast cancer subtypes is leading to a more targeted and personalized approach to breast cancer treatment.
Etiology The etiology of breast cancer is mul- tifactorial, involving endocrine and reproductive factors including nul- liparity, first birth after age 30, and hormonal history; environmental fac- tors such as consumption of alcoholic beverages, use of certain contracep- tives and menopausal (hormone re- placement) therapy, and exposure to ionizing radiation; and lifestyle factors such as high-calorie diets and lack of exercise (see “Biological mecha- nisms mediating reduced breast can- cer risk through physical activity”). The annual incidence in industrial- ized countries where such lifestyle factors have existed for some time is 70–90 new cases per 100 000 wom- en. Countries where industrialization is a more recent phenomenon have a rising incidence and higher mortality.
On the basis of a case–control study in Germany, population attrib- utable risk (95% confidence inter- val) for non-modifiable breast cancer risk factors (age at menarche, age at menopause, parity, benign breast disease, and family history of breast cancer) was 37.2% (27.1–47.2%) for
all invasive tumours considered. Of the modifiable risk factors assessed, use of hormone therapy and physi- cal inactivity had the highest impact, with population attributable risks of 19.4% (15.9–23.2%) and 12.8% (5.5–20.8%), respectively, for in- vasive tumours; all such findings varied depending on the estrogen and progesterone receptor status of tumours [1].
A small proportion of breast can- cer is due to a familial predisposition, and two high-risk, high-penetrance genes have been identified: BRCA1 and BRCA2. Mutations in these genes greatly increase the risk of developing breast cancer. Several lower-penetrance genes have also been identified, and there are many loci within the genome that are linked to an increased risk but for which the specific genes have yet to be iden- tified (see Pathology and Genetics sections below).
Hence, for most patients with breast cancer, multiple factors – in- cluding personal and family history as well as reproductive and lifestyle factors – are implicated in tumour development.
Pathology Breast cancer is not a single disease and is heterogeneous both clinically and morphologically. The current WHO Classification of Tumours of
Chapter 5.2 • Breast cancer 363
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .2
Epidemiology
Breast cancer • Breast cancer is by far the most
frequently diagnosed cancer and cause of cancer death among women. There were an estimated 1.7 million new cases (25% of all cancers in women) and 0.5 mil- lion cancer deaths (15% of all cancer deaths in women) in 2012.
• Breast cancer is the most com- mon cancer diagnosis in women in 140 countries and the most fre- quent cause of cancer mortality in 101 countries.
• Age-standardized incidence rates are highest in western Europe and lowest in East Asia. Incidence rates tend to be elevated in coun- tries attaining the highest levels of human development. There is a greater than 2-fold difference between countries categorized as having low versus very high levels of development.
• About 43% of the estimated new cases and 34% of the cancer deaths occurred in Europe and North America.
• Mortality rates vary approxi- mately 2–5-fold worldwide; the case fatality rate is lower in coun- tries with higher levels of human development.
• Whereas incidence has been generally increasing in most ar- eas of the world, it has peaked and declined over the past dec- ade in a number of highly devel- oped countries.
• Mortality rates have been declin- ing in a number of highly devel- oped countries since the late 1980s and early 1990s, a result of a combination of improved detection and earlier diagno- sis (through population-based screening) and more effective treatment regimens.
Map 5.2.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for breast cancer in women, 2012.
Map 5.2.2. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for breast cancer in women, 2012.
the Breast (4th edition) [2] recogniz- es more than 20 different subtypes.
Most breast cancers arise from epithelial cells (carcinomas); these
tumours are subdivided into in situ and invasive lesions. In situ carcino- mas are preinvasive lesions in which the malignant epithelial cells are still
confined to the ductal/lobular tree of the breast and the basement mem- brane surrounding the ducts and lob- ules is intact to the extent that it has
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
364
Chart 5.2.1. Estimated global number of new cases and deaths with proportions by major world regions, for breast cancer in women, 2012.
Chart 5.2.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for breast cancer in women, circa 1975–2012.
Chart 5.2.3. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for breast cancer in women, circa 1975–2012.
Chapter 5.2 • Breast cancer 365
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .2
not been penetrated by cancer cells. Preinvasive lesions are further sub- divided into ductal carcinoma in situ (DCIS) and lobular carcinoma in situ (LCIS). Despite the “ductal” versus “lobular” terminology, the fact that almost all carcinomas in the breast arise from the structural and func- tional unit of the breast called the terminal duct lobular unit has been well established for some time [3]. The distinction between DCIS and LCIS is the result not of the micro- anatomical site of origin (ducts vs lobules) but rather of a difference in the architectural and cytological features of the cells. DCIS and LCIS also differ in their distribution within the breast, in the associated respec- tive risks of bilateral disease, and in their natural history.
Subsets of DCIS and LCIS have also been recognized. DCIS is sub- classified according to nuclear grade and architectural features; high-grade DCIS has a greater risk of recurrence and progression to invasive cancer than do lower grades of DCIS [4]. LCIS is often a multifocal and multi- centric disease and also carries a risk of bilateral involvement [5]. In its clas- sic form, LCIS presents a small risk of progression to invasive cancer over a period of 20–25 years. Recently, a higher-grade variant, pleomorphic LCIS, has been recognized, but its natural history is unclear at present (Fig. 5.2.1) [6].
Invasive carcinoma character- ized as “no special type”, also known as ductal carcinoma no special type or invasive ductal carcinoma, makes up the largest subset of invasive breast cancer. This designation identifies a heterogeneous group comprising tumours that are not easily categorized by specific mor- phological features that characterize the “special subtypes”. Hence, it is a default diagnosis for all those tu- mours (approximately 70%) that can- not be assigned a “special subtype” designation. The most common of the special subtypes include lobular car- cinoma, tubular carcinoma, mucinous carcinoma, carcinoma with medullary and apocrine features, micropapillary
and papillary carcinomas, and meta- plastic carcinomas (Fig. 5.2.2).
It is now well recognized that tumours of the breast exhibit both intratumour and intertumour het- erogeneity. Histological grade is a measure of how closely a tumour resembles its tissue of origin and is an integral part of a pathology report. The current grading system assess- es three parameters of the tumour: (i) the degree of architectural differ- entiation (tubule formation), (ii) nu- clear pleomorphism (nuclear grade), and (iii) proliferation (mitotic count/ index). Each parameter is given a score from 1 to 3, and the scores for the three parameters are added together. Tumours with a total score of 3–5 are grade 1 (low grade), those with scores of 6 and 7 are grade 2 (intermediate grade), and those with scores of 8 and 9 are grade 3 (high grade). Although this semi- quantitative approach averages the intratumour heterogeneity that exists in many tumours, it remains a pow- erful indicator of patient prognosis (Table 5.2.1).
Histological grade is also strongly associated with histological type and
with patterns of molecular alterations, such as estrogen receptor (ER) and progesterone receptor (PR) expres- sion and human epidermal growth factor receptor 2 (HER2) protein over- expression and gene amplification.
The histopathological analysis of breast cancers also provides information on stage of disease – meaning the extent of disease involvement. This is determined in two ways: from the size of the can- cer and from the assessment of regional lymph node involvement. Both these parameters provide powerful prognostic information for patient management. The overall staging for the patient is recorded with the American Joint Committee on Cancer staging system using the tumour–node –metastasis (TNM) classification. Additional prognostic data, such as lymphovascular per- meation and degree of response to neoadjuvant therapy, can also be derived from the histopathological assessment.
The pathological assessment of tumours as described above is inex- pensive and quick and can provide useful information on which to base
Fig. 5.2.1. Ductal carcinoma in situ of low nuclear grade. (A) Classic lobular carci- noma in situ is diagnosed because the characteristic cells distend and distort the acini. (B) Pleomorphic lobular carcinoma in situ. Nuclear enlargement, variability, and promi- nent nucleoli are associated with abundant cytoplasm. (C) Cribriform pattern. Multiple adjacent ducts are distended by a sieve-like proliferation of monotonous uniform cells. The multiple spaces within the proliferation are rounded and distributed in an organized fashion. (D) This space shows a rim of viable cells with high-grade nuclei. There is comedo-type necrosis and calcification.
A
C
B
D
366
plans for further management. Nonetheless, limitations are recog- nized. Patients with a similar type or grade of breast cancer may have very different response to therapy or long-term outcome. There has been considerable progress over the past three decades in under- standing the biology of breast can- cer and in translating some of the
molecular data into routine clinical practice.
Genetics The assessment of ER and PR sta- tus using immunohistochemistry, and of HER2 status using immuno- histochemistry and in situ hybridiza- tion, is now routine for all primary
breast cancers, at least in the more developed countries. Such assess- ment allows prognostication and also provides predictive data with regard to response to anti-estrogen therapy (e.g. tamoxifen) and anti-HER2 therapy (e.g. trastuzumab) [7]. Such therapy is not appropriate for ap- proximately 15% of tumours, which are designated “triple-negative” to indicate lack of expression of ER, PR, or HER2.
DNA microarray technology for gene expression profiling, which allows thousands of genes to be ex- amined simultaneously, has been used to classify breast cancer, de- velop signatures for “good” versus “poor” prognosis, and identify tu- mours that may or may not respond to particular therapy [8]. This type of analysis has not only confirmed the two large subsets of breast cancer – ER-positive and ER-negative – but also brought to the fore differences within the ER categories. In ER- positive tumours, the luminal B sub- set has a worse prognosis than the luminal A subset, probably driven at least in part by higher prolifera- tive activity. Within the ER-negative group, the triple-negative (ER-, PR-, and HER2-negative) and “basal- like” groups are now recognized; these groups are not synonymous but show considerable overlap. The basal-like group comprises tumours that express proteins found in the contractile myoepithelial cell layer of normal breast (e.g. CK14, CK5/6), and a subset of these tumours have a propensity for early recurrence, especially involving metastasis in the brain and lungs [9]. This type of understanding is already determin- ing management strategies and will lead, through clinical trials, to better targeted therapies.
For more than a decade, such gene expression profiling has been applied to define molecular pheno- types of breast cancer. The luminal A and basal-like subtypes men- tioned above are the two main sub- types of five that have been identi- fied (Fig. 5.2.4). In an evaluation of differences, signature genes of the luminal A subtype were genes
Fig. 5.2.2. (A) Invasive carcinoma no special type, grade 2. (B) Classic invasive lobular carcinoma with uniform, single-file cells compared with (C). (C) Invasive pleomorphic lobular carcinoma with characteristic pleomorphic, atypical nuclei. (D) Tubular carcinoma. There is a haphazard distribution of rounded and angulated tubules with open lumina, lined by only a single layer of epithelial cells separated by abundant reactive, fibroblas- tic stroma. (E) The neoplastic cells lining the teardrop-shaped tubules lack significant atypia. (F) Mixed metaplastic carcinoma with spindle, mesenchymal (chondroid), and squamous differentiation. (G) Neuroendocrine carcinoma of the breast. Alveolar pat- tern with rounded solid nests of spindle cells invading a dense collagenous stroma. (H) Invasive micropapillary carcinoma. Tumour cell clusters with irregular central spaces proliferate within empty stromal spaces. Some clusters have reversed polarity with an “inside-out” morphology.
A
C
E
G
B
D
F
H
Chapter 5.2 • Breast cancer 367
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .2
Biological mechanisms mediating reduced breast cancer risk through physical activity Christine M. Friedenreich
Physical activity has been associ- ated with a 25–30% decrease in breast cancer risk, and research is now focused on understanding the biological mechanisms mediating this association, with the aim of de- termining the optimal type, dose, and timing of activity needed for maxi- mum risk reduction. Several mech- anisms have been hypothesized (Fig. B5.2.1) [1], and intervention trials in postmenopausal, inactive, healthy women have investigated these pathways [2]. These studies have found that exercise decreases endogenous estrogens, adiposity, insulin resistance, leptin, and inflam- mation, which are all associated in-
dependently with increased breast cancer risk. Other emerging hypoth- esized mechanisms include oxida- tive stress and genomic instability.
Exercise reduces the endog- enous estrogen levels in pre- menopausal women by induc- ing anovulatory menstrual cycles or amenorrhea, or through other mechanisms. In postmenopausal women, exercise decreases circulat- ing estrogens directly, and indirectly with exercise-induced weight loss. Obesity, an established risk factor for postmenopausal breast cancer, is decreased with physical activity. Obesity-related effects, notably in- sulin resistance and altered levels
of insulin-like growth factor 1 (IGF-1) and its binding protein IGFBP-3, in addition to altered production of adipokines, specifically leptin and adiponectin, may be important con- tributors to increasing breast cancer risk. Mechanisms independent of changes in body fat are also rel- evant. Exercise acutely enhances insulin sensitivity and glucose up- take by skeletal muscle, which can be sustained through prolonged high- intensity activity. With decreased cir- culating insulin, free IGF-1 levels may also decrease because of lowered growth hormone-mediated hepatic synthesis of IGF-1 and/or increased synthesis of IGFBP-1.
Fig. B5.2.1. Biological mechanisms determining the optimal type, dose, and timing of physical activity needed for maximum reduction in breast cancer risk. IL-6, interleukin-6; SHBG, sex hormone-binding globulin; TNF-α, tumour necrosis factor α.
368
involved in fatty acid metabolism and steroid hormone-mediated signalling pathways, in particular ER signalling, while the corresponding genes for the basal-like subtype were involved in cell proliferation and differentiation, the p21-mediated pathway, and the G1–S cell-cycle checkpoint. A mini-
mal set of 54 genes best discrimi- nated between the two subtypes [10].
Such molecular analyses are also relevant to survival. One study used modelling of messenger RNA (mRNA), copy number alterations, microRNAs, and methylation [11]. For all breast cancers, the stron- gest predictor of good outcome
was acquisition of a gene signa- ture that favoured a high T helper type 1 (Th1)/cytotoxic T lymphocyte response at the expense of Th2- driven immunity.
Next-generation sequencing technology has also been applied to examine the genomes of 100 breast cancers for somatic copy number changes and mutations in the coding exons of protein-coding genes [12]. The number of somatic mutations varied markedly between individual tumours. Strong correla- tions were evident between number of mutations, age at which cancer was diagnosed, and cancer histolog- ical grade, and multiple mutational signatures were observed, includ- ing one, present in about 10% of tu- mours, characterized by numerous mutations of cytosine at TpC dinucle- otides. Driver mutations were identi- fied in several new cancer genes, including AKT2, ARID1B, CASP8, CDKN1B, MAP3K1, MAP3K13, NCOR1, SMARCD1, and TBX3 (Fig. 5.2.5). Among the 100 tu- mours, there were driver mutations in at least 40 cancer genes and 73 different combinations of mutated cancer genes. Overall, one of the most highly contributing genes to de- velopment of breast cancer, TP53, is nonetheless probably only involved in 25% of cancers, although its role
Table 5.2.1. Semi-quantitative method for assessing histological grade
Feature Score
Tubule and gland formation
Majority of tumour (> 75%) 1
Moderate degree (10–75%) 2
Little or none (< 10%) 3
Nuclear pleomorphism
Small, regular, uniform cells 1
Moderate increase in size and variability 2
Marked variation 3
Mitotic count
Dependent on microscope field area 1–3
Final grading
Add scores for tubule and gland formation, nuclear pleomorphism, and mitotic count:
Grade 1 Total score, 3–5
Grade 2 Total score, 6 or 7
Grade 3 Total score, 8 or 9
Chronic inflammation is asso- ciated with sustained proliferation and oxidative stress, which togeth- er promote malignancy. C-reactive protein, a marker of chronic low- grade systemic inflammation, may have a predictive role for breast cancer risk. Exercise may protect against postmenopausal breast can- cer by inducing an anti-inflammatory environment and subsequently pro- tecting against chronic low-grade inflammation. Exercise may lower inflammation via reductions in adi- posity and other pathways.
Oxidative stress plays an impor- tant role in breast cancer carcino- genesis – specifically, the oxida- tion of DNA in carcinogenesis and
tumour promotion. Epidemiological data also suggest associations be- tween oxidative stress and breast cancer. As a part of a favourable biological adaptive response, ex- ercise enhances the capacity of antioxidant and oxidative damage repair enzymes, and subsequently reduces oxidative damage.
Investigators have called for tri- als of exercise and breast cancer incidence, to provide the ultimate level of evidence about the type, dose, and timing of activity needed for breast cancer prevention, but no studies are currently planned be- cause of the considerable expense and time commitment required [3]. Thus, to date, the epidemiological
research performed in healthy popu- lations is restricted to evidence on intermediate end-points for breast cancer. Nonetheless, evidence ex- ists that these pathways are amena- ble to change by exercise.
References 1. Neilson HK et al. (2009). Cancer Epidemiol
Biomarkers Prev, 18:11–27. http://dx.doi. org/10.1158/1055 -9965.EPI- 08 - 0756 PMID:19124476
2. Winzer BM et al. (2011). Cancer Causes Control, 22:811–826. http://dx.doi.org/10. 1007/s10552-011-9761-4 PMID:21461921
3. Ballard-Barbash R et al. (2009). J Natl Cancer Inst, 101:630–643. http://dx.doi. org/10.1093/jnci/djp068 PMID:19401543
Chapter 5.2 • Breast cancer 369
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .2
in subsets of breast cancers, such as triple-negative/basal-like, is much higher. A large number of genes appear to be implicated in a small percentage of tumours, and this will provide additional challenges to achieving the dream of a patient- specific targeted and personalized medical intervention.
A better understanding of famil- ial breast cancer due to germline mutations in BRCA1 and BRCA2 is determining optimal patient man- agement (see “Treatment deter- mined by BRCA1/2 mutation”). The
pathology of these tumours is well detailed now, and it is clear that most BRCA1-associated breast cancers are high-grade, medullary-like, and triple-negative, and have a basal- like phenotype [13]. There are also data indicating that a proportion of ER-positive tumours in patients with BRCA1 mutations are also due to in- activation of BRCA1 and may not be sporadic [14]. The morphological fea- tures of BRCA2-associated tumours are more heterogeneous, although most are high-grade breast cancers. The pathology has been combined with various personal and family his- tory factors to develop algorithms to predict the risk of carrying a BRCA1 mutation, and these have been used in the clinic with varying degrees of success.
Prospects Prevention To date, most breast cancer pre- vention strategies have focused on reducing the development of breast cancer in women considered to be at
moderately or greatly increased risk of the disease based on calculated risks determined from prediction models, or in women with germline mutations in high-penetrance breast cancer susceptibility genes (BRCA1 and BRCA2).
Several clinical trials have eval- uated the effect on the develop- ment of breast cancer of endocrine interventions using selective ER modulators (SERMs) and aromatase inhibitors. The results of the National Surgical Adjuvant Breast and Bowel Project’s Breast Cancer Prevention Trial [15] and Study of Tamoxifen and Raloxifene Trial [16] demonstrated that the SERMs tamoxifen and ralox- ifene are each associated with about a 50% reduction in the development of breast cancer in women consid- ered to be at moderately increased risk of breast cancer based on the Gail risk assessment model. In two other prevention trials, tamoxifen was associated with more modest reduc- tions in the development of breast cancer. However, in both of these trials some of the women received
Fig. 5.2.3. Adenocarcinoma of the breast below the nipple, in cross-section.
Fig. 5.2.4. Identification of breast cancer subtypes. (A) Correlation of breast tumour samples with five recognized subtypes of breast tumours. For this analysis, 526 out of 552 previously identified “intrinsic” genes were cross-mapped and subject to hierarchical clus- tering. (B) Unsupervised hierarchical clustering of 20 breast tumour tissues analysed using the 526 mapped intrinsic genes (the two microarray replicates for each sample are shown). The level of expression of each gene in each sample, relative to the median level of expression of that gene across all the samples, is represented using a colour scale as shown in the key (green, below median; black, equal to median; red, above median). (Left panel) Scaled-down representation of the entire cluster of the 526 intrinsic genes and 20 tissue samples. (Right panel) Experimental dendrogram displaying the clustering of the tumours into three distinct subgroups.
370
hormone replacement therapy and this may have attenuated the che- mopreventive benefit of tamoxifen [17,18]. Despite the demonstrated efficacy of tamoxifen and raloxifene in reducing the development of breast cancer, these SERMs have had lim- ited patient acceptance in the chemo- prevention setting.
The role of aromatase inhibitors in breast cancer prevention is also an area of active research. In one recent study, treatment with exemes- tane was associated with a 65%
relative reduction in the annual inci- dence of invasive breast cancer in postmenopausal women at moder- ately increased risk of breast cancer, with no serious toxicity and minimal effect on health-related quality of life [19]. The role of anastrozole in the chemoprevention setting is currently being evaluated. The reduction of breast cancer development associ- ated with these endocrine interven- tions has been notably limited to ER-positive tumours. These agents have not been effective in reducing
the risk of ER-negative tumours, which account for about 20–30% of breast cancers and are associated with a poorer prognosis [20]. As a result, the potential role in breast cancer prevention of various drugs that target non-endocrine signal- ling pathways is under active inves- tigation. For example, recent data from the Women’s Health Initiative indicated that among postmeno- pausal women with diabetes, treat- ment with metformin was associated with a reduced incidence of breast cancer [21]. Other non-endocrine drugs being studied as possible breast cancer prevention agents in- clude cyclooxygenase 2 inhibitors, retinoids, and receptor tyrosine kinase inhibitors, among others. However, clinical trials will be re- quired to evaluate the role of these agents in breast cancer prevention.
Bilateral prophylactic mastec- tomy is a highly effective strategy to prevent the development of breast cancer in women with BRCA muta- tions. This procedure is associated with a 90% reduction in the devel- opment of breast cancer in these
Fig. 5.2.5. Recently identified breast cancer genes and involvement of the JUN kinase signalling pathway. (A) Representations of the protein-coding sequences and major domains in cancer genes relevant to breast cancer. Somatic mutations evident in tu- mours are shown as circles: truncating (red), essential splice site (blue), missense (green), and in-frame insertion or deletion (yellow). Red lines indicate the positions of large homozygous deletions. aa, amino acids. (B) Pathways regulating the JUN kinases MAP2K7 and MAP2K8, indicating genes with mutations in this series. Genes in green are activated by mutations, whereas genes in purple are inactivated.
Fig. 5.2.6. Pregnancy before age 30 and breastfeeding reduce a woman’s total number of lifetime menstrual cycles, and both factors are associated with lower risk of breast cancer. (For more details, see Chapter 2.5.)
Chapter 5.2 • Breast cancer 371
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .2
women, and an even higher (95%) reduction in breast cancer among those who have also had their ovaries removed [22]. Bilateral prophylactic mastectomy is associ- ated with an even greater (> 99%) reduction in breast cancer devel- opment among women who were considered to be at high risk of the disease and who were cared for in community practices and were therefore not as highly selected as in referral centres [23]. Recent technical improvements in onco- plastic surgery have resulted in better cosmetic outcomes in wom- en who undergo post-mastectomy reconstruction and have, in turn, made the option of prophylactic
mastectomy more appealing than in the past. The influence of modifying dietary and lifestyle factors on the prevention of breast cancer is also under active investigation [24].
Screening Mammography remains the mainstay of population-based breast cancer screening although access to mam- mographic screening varies widely around the world and the procedure is unavailable in most developing coun- tries. Mammography results in the detection of more in situ lesions than would be identified otherwise and of invasive cancers that are smaller, of lower grade, more often node- negative, and more often of special
types (especially tubular carcinomas) than are evident in non-screened populations. Controversies remain about the magnitude of mortality reduction attributable to mammo- graphic screening, the lower and upper age limits of the patients who should be screened, and the most cost-effective screening interval [25]. In addition, mammographic screen- ing can result in overdiagnosis due to the detection of tumours that will never become a threat to life, as well as an excessive number of biopsies, and patient anxiety [25,26].
In recent years, technical ad- vances have improved the sensitivity of mammography for at least a sub- set of women [25]. In particular, in a
Treatment determined by BRCA1/2 mutation Susan M. Domchek
Germline mutations in BRCA1 and BRCA2 are associated with an in- creased lifetime risk of breast and ovarian cancers, as well as other cancers such as male breast cancer, prostate cancer, and pancreatic can- cer. The lifetime risk of breast cancer is 60–70% in women harbouring ei- ther mutation. Interventions such as prophylactic mastectomy and prophylactic oophorectomy have been demonstrated to reduce the risk of breast and ovarian cancers in several observational cohort stud- ies. Risk-reducing oophorectomy has also been associated with im- proved overall survival [1]. In women diagnosed with breast cancer, the knowledge that there is an increased risk of developing a second primary breast cancer can factor into deci- sion-making about local therapy with the option of bilateral mastectomy.
Due to the biological functions of BRCA1 and BRCA2 (particularly their roles in DNA damage repair of double-strand breaks), it has been postulated that there may be dif- ferential responses of BRCA1/2- associated tumours to different types of systemic therapy. In gen- eral, BRCA-associated cancers (both breast and ovarian) appear
to be more sensitive than sporadic cancers to chemotherapy. Emerging data suggest that specific chemo- therapy type may matter clinically. Several small studies have reported high response rates in breast cancer to cisplatin-based chemotherapy given in both the pre-operative and metastatic setting [2]. For exam- ple, in a study of 20 patients with BRCA1-positive metastatic breast cancer, the objective response rate to cisplatin chemotherapy was 80%. There are no data available from randomized studies compar- ing cisplatin with other agents, but the reported response rates and progression-free survival data are encouraging.
Novel agents called poly(ADP- ribose) polymerase (PARP) inhibi- tors are being studied in BRCA1 and BRCA2 mutation carriers with both metastatic breast cancer and recurrent ovarian cancer. These drugs, which remain in clinical tri- als, aim to exploit the fundamental defects in BRCA1/2-associated cancers via the concept of synthet- ic lethality. Synthetic lethality refers to the idea that inhibition or muta- tion of two pathways leads to cell death when inhibition or mutation of
either alone does not. Initial reports have demonstrated high response rates in both BRCA1/2-associated metastatic breast cancer and recur- rent ovarian cancer [3]. Interestingly, although PARP inhibitors do not ap- pear to have activity in unselected sporadic triple-negative breast can- cer, studies have been encouraging in sporadic high-grade serous ovar- ian cancer.
Mechanisms of resistance to both cisplatin and PARP inhibitors are beginning to be elucidated. One very interesting mechanism is that of secondary or “reversion” mutations, which restore full-length protein. Studies are under way to determine the optimal use of both platinum agents and PARP inhibitors in BRCA1/2 mutation carriers.
References 1. Domchek SM et al. (2010). JAMA,
304:967–975. http://dx.doi.org/10.1001/ jama.2010.1237 PMID:20810374
2. Byrski T et al. (2012). Breast Cancer Res, 14:R110. http://dx.doi.org/10.1186/bcr3231 PMID:22817698
3. Tutt A et al. (2010). Lancet, 376:235–244. http://dx.doi.org/10.1016/S0140 - 6736 (10)60892-6 PMID:20609467
372
large clinical trial digital mammogra- phy was shown to be more sensitive than traditional film mammography in younger women (aged < 50 years, or premenopausal) and in women with dense breasts. Breast tomosynthe- sis, a variation of digital mammog- raphy, produces three-dimensional digital images of the breast but has not been demonstrated to be more sensitive than standard digital mam- mography. Both ultrasonography and magnetic resonance imaging may be of value as adjuncts to mam- mography in selected cases. These techniques are associated with high false-positive rates and are not rec- ommended for screening the general
population [25]. Magnetic resonance imaging, however, may have a role in the screening of women with BRCA1 germline mutations. Newer modali- ties that may ultimately find a role in breast cancer screening include molecular imaging and functional im- aging techniques [27]. Alternatives to mammography for breast cancer screening, particularly in developing countries, are being evaluated and include breast awareness and clinical breast examination (see Chapter 4.8).
Targeted agents Selection of breast cancer treatment, particularly systemic therapy, is
currently based primarily on clinical and pathological factors, supple- mented by hormone receptor and HER2 status. The standard options for systemic therapy currently include hormone therapy, cytotoxic therapy, and HER2-targeted therapy, singly and in various combinations. Rapidly evolving knowledge of the molecu- lar events and signalling pathways underlying the development and progression of breast cancer has led to the identification of a growing number of new therapeutic targets and to the development of drugs against these targets. Many clinical trials are currently under way world- wide to evaluate the role of these new targeted therapies in the treat- ment of patients with breast cancer. Newer targeted therapies that have been or are being evaluated, sin- gly and in combination with each other and with traditional cytotoxic agents, include angiogenesis inhibi- tors, tyrosine kinase inhibitors, in- hibitors of mammalian target of rap- amycin (mTOR), poly(ADP-ribose) polymerase 1 (PARP1) inhibitors, insulin-like growth factor 1 recep- tor (IGF-1R) inhibitors, proteasome inhibitors, phosphatidylinositol 3-ki- nase (PI3K) inhibitors, and others [28]. The ultimate goal of this re- search is to be able to tailor breast cancer treatment for individual pa- tients based on the particular mo- lecular features of the tumour.
Fig. 5.2.7. A woman undergoing mammography screening. Her breast is being com- pressed to obtain the optimal mammographic image.
Chapter 5.2 • Breast cancer 373
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .2
References
1. Barnes BB, Steindorf K, Hein R et al. (2011). Population attributable risk of invasive postmenopausal breast cancer and breast cancer subtypes for modifiable and non- modifiable risk factors. Cancer Epidemiol, 35:345–352. http://dx.doi.org/10.1016/j. canep.2010.11.003 PMID:21159569
2. Lakhani SR, Ellis IO, Schnitt SJ et al., eds (2012). WHO Classification of Tumours of the Breast, 4th ed. Lyon: IARC.
3. Wellings SR, Jensen HM, Marcum RG (1975). An atlas of subgross pathology of the human breast with special reference to possible precancerous lesions. J Natl Cancer Inst, 55:231–273. PMID:169369
4. Silverstein MJ, Poller DN, Waisman JR et al. (1995). Prognostic classification of breast ductal carcinoma-in-situ. Lancet, 345:1154–1157. http://dx.doi.org/10.1016/ S0140-6736(95)90982-6 PMID:7723550
5. Beute BJ, Kalisher L, Hutter RV (1991). Lobular carcinoma in situ of the breast: clini- cal, pathologic, and mammographic fea- tures. AJR Am J Roentgenol, 157:257–265. http://dx.doi.org/10.2214/ajr.157.2.1853802 PMID:1853802
6. Eusebi V, Magalhaes F, Azzopardi JG (1992). Pleomorphic lobular carci- noma of the breast: an aggressive tu- mor showing apocrine differentiation. Hum Pathol, 23:655–662. http://dx.doi. org/10.1016/0046-8177(92)90321-S PMID: 1592388
7. Hammond ME, Hayes DF, Dowsett M et al. (2010). American Society of Clinical Oncology/College of American Pathologists guideline recommendations for immunohistochemical testing of estro- gen and progesterone receptors in breast cancer. J Clin Oncol, 28:2784–2795. http:// dx.doi.org/10.1200/JCO.2009.25.6529 PMID:20404251
8. van de Vijver MJ, He YD, van’t Veer LJ et al. (2002). A gene-expression signa- ture as a predictor of survival in breast cancer. N Engl J Med, 347:1999–2009. http://dx.doi.org/10.1056/NEJMoa021967 PMID:12490681
9. Fulford LG, Reis-Filho JS, Ryder K et al. (2007). Basal-like grade III invasive ductal carcinoma of the breast: patterns of me- tastasis and long-term survival. Breast Cancer Res, 9:R4. http://dx.doi.org/ 10.1186/bcr1636 PMID:17217540
10. Sørlie T, Wang Y, Xiao C et al. (2006). Distinct molecular mechanisms underlying clinically relevant subtypes of breast cancer: gene expression analyses across three dif- ferent platforms. BMC Genomics, 7:127. http://dx.doi.org/10.1186/1471-2164-7-127 PMID:16729877
11. Kristensen VN, Vaske CJ, Ursini-Siegel J et al. (2012). Integrated molecular profiles of invasive breast tumors and ductal car- cinoma in situ (DCIS) reveal differential vascular and interleukin signaling. Proc Natl Acad Sci U S A, 109:2802–2807. http://dx.doi.org/10.1073/pnas.1108781108 PMID:21908711
12. Stephens PJ, Tarpey PS, Davies H et al.; Oslo Breast Cancer Consortium (OSBREAC) (2012). The landscape of cancer genes and mutational processes in breast can- cer. Nature, 486:400–404. http://dx.doi. org/10.1038/nature11017 PMID:22722201
13. Lakhani SR, Reis-Filho JS, Fulford L et al.; Breast Cancer Linkage Consortium (2005). Prediction of BRCA1 status in patients with breast cancer using estro- gen receptor and basal phenotype. Clin Cancer Res, 11:5175–5180. http://dx.doi. org/10.1158/1078 - 0432.CCR- 04 -2424 PMID:16033833
14. Tung N, Miron A, Schnitt SJ et al. (2010). Prevalence and predictors of loss of wild type BRCA1 in estrogen receptor positive and negative BRCA1-associated breast cancers. Breast Cancer Res, 12:R95. http://dx.doi.org/10.1186/bcr2776 PMID:21080930
15. Fisher B, Costantino JP, Wickerham DL et al. (1998). Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst, 90:1371–1388. http://dx.doi.org/10.1093/ jnci/90.18.1371 PMID:9747868
16. Vogel VG, Costantino JP, Wickerham DL et al.; National Surgical Adjuvant Breast and Bowel Project (NSABP) (2006). Effects of tamoxifen vs raloxifene on the risk of developing invasive breast cancer and other disease outcomes: the NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 trial. JAMA, 295:2727–2741. http:// dx.doi.org/10.1001/jama.295.23.joc60074 PMID:16754727
17. Cuzick J, Forbes JF, Sestak I et al.; International Breast Cancer Intervention Study I Investigators (2007). Long-term re- sults of tamoxifen prophylaxis for breast cancer – 96-month follow-up of the ran- domized IBIS-I trial. J Natl Cancer Inst, 99:272–282. http://dx.doi.org/10.1093/jnci/ djk049 PMID:17312304
18. Powles TJ, Ashley S, Tidy A et al. (2007). Twenty-year follow-up of the Royal Marsden randomized, double-blinded ta- moxifen breast cancer prevention trial. J Natl Cancer Inst, 99:283–290. http://dx.doi. org/10.1093/jnci/djk050 PMID:17312305
19. Goss PE, Ingle JN, Alés-Martínez JE et al.; NCIC CTG MAP.3 Study Investigators (2011). Exemestane for breast-cancer prevention in postmenopausal women. N Engl J Med, 364:2381–2391. http:// d x.d o i.o r g /10.10 5 6 / N E J M oa110 3 5 07 PMID:21639806
20. Cazzaniga M, Bonanni B (2012). Prevention of ER-negative breast cancer: where do we stand? Eur J Cancer Prev, 21:171–181. http://dx.doi.org/10.1097/CEJ. 0b013e32834c9c26 PMID:21968686
21. Chlebowski RT, McTiernan A, Wactawski- Wende J et al. (2012). Diabetes, metformin, and breast cancer in postmenopausal wom- en. J Clin Oncol, 30:2844–2852. http://dx. doi.org/10.1200/JCO.2011.39.7505 PMID: 22689798
22. Rebbeck TR, Friebel T, Lynch HT et al. (2004). Bilateral prophylactic mastectomy reduces breast cancer risk in BRCA1 and BRCA2 mutation carriers: the PROSE Study Group. J Clin Oncol, 22:1055–1062. http://dx.doi.org/10.1200/JCO.2004.04.188 PMID:14981104
23. Geiger AM, Yu O, Herrinton LJ et al. (2005). A population-based study of bilateral pro- phylactic mastectomy efficacy in women at elevated risk for breast cancer in community practices. Arch Intern Med, 165:516–520. http://dx.doi.org/10.1001/archinte.165.5.516 PMID:15767526
24. Willett WC, Tamimi RM, Hankinson SE et al. (2010). Nongenetic factors in the causation of breast cancer. In: Harris JR, Lippman ME, Morrow M et al., eds. Diseases of the Breast, 4th ed. Philadelphia: Lippincott Williams & Wilkins, pp. 248–290.
25. Warner E (2011). Clinical practice. Breast-cancer screening. N Engl J Med, 365:1025–1032. http://dx.doi.org/10.1056/ NEJMcp1101540 PMID:21916640
26. Lebovic GS, Hollingsworth A, Feig SA (2010). Risk assessment, screening and prevention of breast cancer: a look at cost- effectiveness. Breast, 19:260–267. http:// dx.doi.org/10.1016/j.breast.2010.03.013 PMID:20399656
27. Yang WT (2011). Emerging techniques and molecular imaging in breast cancer. Semin Ultrasound CT MR, 32:288–299. http://dx.doi.org/10.1053/j.sult.2011.03.003 PMID:21782119
28. Perez EA, Spano JP (2012). Current and emerging targeted therapies for metastatic breast cancer. Cancer, 118:3014–3025. http://dx.doi.org/10.1002/cncr.26356 PMID: 22006669
374
Summary • Oesophageal carcinomas are
less common than cancers of the colon and stomach but often prove lethal based on diagno- sis of predominantly late-stage disease.
• The two main types of oesopha- geal carcinoma are squamous cell carcinoma and adenocarcinoma.
• Both types are more common in men than women, but squa- mous cell carcinomas are en- countered in individuals from low-resource regions, whereas adenocarcinomas tend to arise in high-resource populations.
• Both types are successfully managed by mucosal ablation strategies if patients present with early-stage lesions, where- as chemoradiation treatment of- fers limited but slowly improving success in patients presenting with late-stage lesions.
• Screening protocols remain poorly developed for oesopha- geal carcinomas.
Oesophageal cancer is a malig- nant, usually epithelial neoplasm most commonly showing squamous,
5.3 5 ORGAN SITE
Oesophageal cancer Elizabeth A. Montgomery Fred T. Bosman (reviewer)
Paul Brennan (reviewer) Reza Malekzadeh (reviewer)
glandular (adenocarcinoma), or neu- roendocrine differentiation and aris- ing in the oesophagus. Other, rarer types of malignant neoplasms can be encountered (adenoid cystic carci- noma, adenosquamous carcinoma, muco-epidermoid carcinoma, mixed adenoneuroendocrine carcinoma, various sarcomas, and melanoma). Only squamous cell carcinoma and adenocarcinoma are addressed here as they account for most oe- sophageal malignant neoplasms.
Oesophageal squamous cell carcinoma Squamous cell carcinoma of the oe- sophagus (Fig. 5.3.1) is a malignant epithelial neoplasm with squamous differentiation consisting of keratino- cyte cells, often with intercellular bridges or production of keratin.
Etiology Squamous cell carcinoma of the oe- sophagus is characterized by great geographical variation. Pockets of high incidence (> 50 per 100 000 people) occur in the Islamic Republic of Iran (e.g. Golestan Province; > 100 per 100 000), parts of China (e.g. Linxian, Henan Province; > 130 per 100 000), and Zimbabwe. Intermediate incidence is seen in East Africa, southern Brazil, the Caribbean, much of China, parts
of Central Asia, northern India, and southern Europe; incidence is low in North America, northern Europe, and western Africa. There are some data correlating ethnicity with the risk of squamous cell carcinoma, indicating that Turkish or Mongolian people in Central Asia and African Americans in North America are more likely than other people in those regions to be diagnosed with squamous cell carcinoma.
Alcohol consumption and to- bacco smoking and chewing are the strongest risk factors for the devel- opment of oesophageal squamous cell carcinoma, although these associations display marked geo- graphical variation. Cessation of smoking substantially reduces the smoking-associated risk. Intake of hot beverages (hot maté in parts of South America; hot tea, hot coffee, or hot soups elsewhere) is similarly
Fig. 5.3.1. Ulcerated oesophageal squa- mous cell carcinoma.
Chapter 5.3 • Oesophageal cancer 375
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .3
Epidemiology
Oesophageal cancer • Oesophageal cancer is the
eighth most common cancer worldwide, with an estimated 456 000 new cases (3% of all cancers) and 0.4 million can- cer deaths (5% of all cancer deaths) in 2012. About 73% of all new cases occurred in coun- tries at low or medium levels of human development, and 49% of all new cases occurred in China.
• Incidence and mortality rates are elevated in Central and East Asia as well as in eastern Africa. Incidence rates tend to be relatively low in west- ern Africa and in some Latin American countries.
• Incidence varies 15-fold be- tween countries worldwide in men and almost 20-fold in women. Incidence and mortal- ity rates are 2– 4 times as high in men as in women.
• Due to the high fatality rate, mortality rates are close to incidence rates, regardless of sex differences and human de- velopment levels.
• Incidence and mortality trends are variable and reflect the changing prevalence and dis- tribution of the underlying risk factors for oesophageal cancer and its main histological sub- types (adenocarcinoma and squamous cell carcinoma). An increasing tendency to clas- sify cancers located at the gastro-oesophageal junction as adenocarcinoma (rather than gastric cardia cancer) may also have had an impact on the overall trends.
Map 5.3.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for oesophageal cancer in men, 2012.
Map 5.3.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for oesophageal cancer in women, 2012.
associated with risk of oesophageal squamous cell carcinoma, as is in- gestion of caustic substances (such as occurs in suicide attempts and in accidental swallowing of household toxins by children). Certain dietary
habits (low intake of fresh fruits and vegetables, fresh meat or fish, and dairy products, and high intake of barbecued meats and pickled veg- etables, which may result in expo- sure to N-nitroso compounds) are
associated with a high incidence of oesophageal squamous cell carcinoma. Use of opium, poor oral health, and poor nutrition all inter- act to affect outcome in oesopha- geal squamous cell carcinoma in
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
376
Chart 5.3.1. Estimated global number of new cases and deaths with proportions by major world regions, for oesophageal cancer in both sexes combined, 2012.
Chart 5.3.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for oesophageal cancer in men, circa 1975–2012.
Chart 5.3.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for oesophageal cancer in women, circa 1975–2012.
Chapter 5.3 • Oesophageal cancer 377
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .3
high-risk areas. For example, in the Islamic Republic of Iran, opium intake is associated with poor out- come in univariate analysis but not in adjusted models [1]. Other cir- cumstances mediating increased risk include having Plummer– Vinson syndrome (sideropenic dys- phagia, including dysphagia due to oesophageal webs, glossitis, and iron-deficiency anaemia, usually in postmenopausal women), coeliac disease, and achalasia. Finally, ex- posure to ionizing radiation, often from treatment for breast carcino- ma, is a risk factor.
Given that alcohol consumption and tobacco use are the two major known causes of oesophageal squa- mous cell carcinoma, an Australian study assessed respective popula- tion attributable fractions based on 305 cases and 1554 controls. The results were 49% (95% confidence interval [CI], 38–60%) and 32% (95% CI, 25–40%) of oesophageal squamous cell carcinoma cases attributable to smoking and heavy alcohol consumption, respectively. More than 75% of the oesophageal squamous cell carcinoma burden in men could be attributed to smokers with heavy alcohol consumption
[2]. Polymorphisms in acetaldehyde dehydrogenase 2 (ALDH2) are key to individual susceptibility to upper aerodigestive tract cancer; patients with inactive ALDH2 are at high risk of oesophageal squamous cell carci- noma [3]. In Asians, polymorphisms in the ALDH gene family are asso- ciated with altered risk of oesopha- geal squamous cell carcinoma. The effects of these polymorphisms are synergistic with alcohol consumption and smoking. In addition, the poly- morphisms in ALDH associated with oesophageal carcinoma also result in accumulation of acetaldehyde, which causes flushing upon ingestion of alcohol (the “Asian flush”) in about one third of East Asians (Chinese, Japanese, and Koreans) [4]. Genome-wide association studies in Chinese individuals have identified several susceptibility loci, encom- passing 5q11, 6p21, 10q23, 12q24, and 21q22 and including genes re- lated to those encoding ALDH and thereby linked to the association between alcohol consumption and oesophageal squamous cell carci- noma in both Chinese and Japanese people [5,6]. The prevalence of oe- sophageal squamous dysplasia par- allels rates of invasive oesophageal
squamous cell carcinoma; it is typi- cally found in 25% or more of adults older than 35 years in north central China, where the risk of this malig- nancy is among the highest in the world [7].
Pathology and genetics Oesophageal squamous cell car- cinoma is defined as a squamous neoplasm that invades through the epithelial basement membrane into the lamina propria, submucosa, mus- cularis propria, or deeper. Variable amounts of keratinization are seen, manifested by cells showing brightly eosinophilic opaque cyto- plasm (Fig. 5.3.3A,B). Invasion into the lamina propria begins with the proliferation of rete-like projections of neoplastic squamous epithelium. Both horizontal and vertical spread are observed. Tumours can pene- trate vertically through the oesopha- geal wall and invade the intramural lymphatic channels and veins.
The development of oesophageal squamous cell carcinoma is under- stood to be a multistep process that progresses from normal squamous epithelium to include intraepithelial neoplasia (dysplasia; Fig. 5.3.3C) and culminates in the growth of in- vasive carcinoma. Mutation in the TP53 gene is an early event, some- times detectable in intraepithelial neoplasia. Amplification of cyclin D1 occurs in 20–30% of these tumours. Inactivation of CDKN2A, either by homozygous deletion or by de novo methylation, appears to be associ- ated with advanced disease.
The scope of molecular analy- ses applicable to all tumour types is indicated from studies identifying alterations in genes, proteins, and microRNAs in oesophageal squa- mous cell carcinoma, of which 10 or more studies in each category are available. Epigenetic abnormalities are evident from DNA methylation, histone deacetylation, chromatin remodelling, gene imprinting, and non-coding RNA regulation studies [8]. High-throughput genotyping of metastatic tumours has identified phosphatidylinositol 3-kinase (PI3K) and BRAF mutations among the
Fig. 5.3.2. A man makes tea in Rawalpindi, Pakistan. Intake of hot beverages like tea is associated with increased risk of oesophageal squamous cell carcinoma.
378
somatic mutations evident [9]. In a separate study, inactivating NOTCH1 mutations were identified in 21% of oesophageal squamous cell carcino- mas from the USA but only rarely in such tumours from China [10].
Familial predisposition to oe- sophageal cancer has been studied predominantly in association with focal non-epidermolytic palmoplan- tar keratoderma (also known as ty- losis), a rare condition in industrial- ized countries (low-risk populations). This autosomal dominant inherited disorder of the palmar and plantar skin surfaces is associated with oe- sophageal cancer. The causative lo- cus, the tylosis oesophageal cancer (TOC) gene, maps to 17q25, with missense mutations (c.557T → C [p.Ile186Thr] and c.566C → T [p.Pro189Leu]) in RHBDF2, which encodes the inactive rhomboid protease RHBDF2 (also known as iRhom2) [11].
Prospects The overall prognosis of oesophage- al squamous cell carcinoma remains poor based on diagnosis of predomi- nantly late-stage disease. Screening programmes to detect precursor and early-stage lesions have been im- plemented in some high-risk popula- tions. Radiofrequency ablation ther- apy and other forms of endoscopic treatments appear promising for eradication of early lesions [12,13].
Oesophageal adenocarcinoma Adenocarcinoma of the oesophagus is a malignant epithelial neoplasm with glandular differentiation arising predominantly from Barrett mucosa and typically encountered in the lower third of the oesophagus.
Etiology Oesophageal adenocarcinoma is characterized by a high male-to- female ratio (from 4 to 7), with a higher incidence among Caucasians and among subgroups with a high socioeconomic status [14,15]. The incidence and prevalence of oe- sophageal adenocarcinoma was reported as rising during the last decades of the 20th century in a number of high-resource countries [14–16]. For example, in the USA, incidence rates for adenocarcinoma have been rising markedly, whereas rates for squamous cell carcinoma are declining steadily. From 1999 to 2008, rates among White men increased substantially (1.8% per year), as among White women (2.1% per year) and Hispanic men (2.8% per year). No significant changes were observed for men or women of other racial/ethnic groups. While rates of oesophageal adenocarci- noma have been increasing among Caucasians in some populations, it is extremely difficult to quantify the rate of increase from analysis of time trends in registry data, due
partly to improvements in recording and capturing histology information (decreasing numbers of tumours without histological confirmation) but also because of changes in the assignment of cancers at the gastro-oesophageal junction, from the gastric cardia (or gastric, other/ non-specified) to the oesophagus.
Obesity is a strong risk factor for oesophageal adenocarcinoma, as is gastro-oesophageal reflux disease. The trend of increasing incidence rates of oesophageal adenocarci- noma parallels the growing obesity epidemic and the prevalence of gastro-oesophageal reflux disease [15]. However, in a 2012 study in the USA, incidence rates of oesopha- geal adenocarcinoma were 7-fold higher among White men than among Black men, despite data in- dicating that the prevalence of obe- sity is higher among Black men and women than among White men and women [15].
The countries with the high- est incidence of this tumour type are the United Kingdom, Australia, the Netherlands, and the USA. Relatively lower incidence rates are reported from eastern Europe and Scandinavia. In Latin America, Asia, and Africa, oesophageal ad- enocarcinoma remains uncommon, but it is possibly also underreported (particularly in Latin America).
Although familial association has been reported for patients with
Fig. 5.3.3. Invasive squamous cell carcinoma and precursor lesions. (A) Nests of carcinoma cells with brightly eosinophilic cytoplasm in keeping with keratinization embedded in desmoplastic stroma from a mucosal biopsy. (B) This squamous cell carcinoma shows a basaloid pattern such that the cells are small and darkly stained owing to their high nuclear-to-cytoplasmic ratio. The clue that this is a squamous cell carcinoma is the overt keratinization at the right side of the field. (C) High-grade squamous dysplasia (intraepithelial neoplasia) of the oesophagus. This lesion has not invaded the basement membrane (Tis) but shows marked cytological alterations. In the oesophagus, such lesions are typically not associated with human papillomavirus.
A B C
Chapter 5.3 • Oesophageal cancer 379
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .3
Risk of malignant progression in Barrett oesophagus patients Liam J. Murray
Barrett oesophagus (BO) is a pre- cursor of oesophageal adenocar- cinoma. The magnitude of cancer risk in BO is crucial in determining whether the benefits of screening, surveillance, or treatment of BO out- weigh any risks and whether such strategies are cost-effective. The published incidence of malignancy in BO varies widely, from 0 to 35.5 cases per 1000 person-years of risk (pyr) [1]. To date, most studies have been small, and publication bias in favour of studies showing high risks has been demonstrated [2]. Other design issues, such as the inclusion of unrepresentative cohorts of BO patients (e.g. patients from tertiary referral centres), inclusion of those at high risk of progression (e.g. with high-grade dysplasia at baseline), incomplete follow-up of patients, or failure to exclude early incident (most likely prevalent) cancers, have resulted in overestimation of the risks of malignant progression in BO. Differing criteria for diagno- sis of BO, for example inclusion of patients without specialized intes- tinal metaplasia, and differing defi- nitions of malignant progression, for example inclusion/exclusion of incident high-grade dysplasia, have also contributed to the variation in observed rates. The combined risk of oesophageal adenocarcinoma or
high-grade dysplasia may provide a better estimate of malignant pro- gression than risk of oesophageal adenocarcinoma alone, although it will overestimate risk as not all patients with high-grade dysplasia would progress to cancer in the ab- sence of therapeutic intervention [3].
The best estimates of risk of ma- lignant progression in BO are pro- vided by population-based studies with long-term follow-up for a high proportion of patients (excluding those with high-grade dysplasia at baseline) and discounting events occurring soon after diagnosis. Several such studies have recently been published [4–8]. Even in these studies, differing BO definitions, dif- ferent proportions of patients with low-grade dysplasia at baseline, and differing approaches to, and com- pleteness of, follow-up have resulted in variable estimates of malignant progression. Among patients with- out low-grade dysplasia at baseline, the incidence of oesophageal ade- nocarcinoma/high-grade dysplasia ranged from 2.2 per 1000 pyr [4] to 5.2 per 1000 pyr [6,8]; however, the higher estimates were not based on follow-up of all diagnosed cases and may have overestimated risk. Risk of oesophageal adenocarcinoma/ high-grade dysplasia was 2–6-fold higher in patients with low-grade
dysplasia at baseline than in those without low-grade dysplasia.
Using currently available data, it is difficult to reach a definite con- clusion about the risk of malignant progression in BO, but it is lower than previously thought and, in pa- tients without low-grade dysplasia at baseline, appears to be lower than 5 cases of oesophageal adenocar- cinoma/high-grade dysplasia per 1000 pyr.
References 1. Yousef F et al. (2008). Am J Epidemiol,
168:237–249. http://dx.doi.org/10.1093/ aje/kwn121 PMID:18550563
2. Shaheen NJ et al. (2000). Gastroenterology, 119:333–338. http://dx.doi.org/10.1053/ gast.2000.9302 PMID:10930368
3. Schnell TG et al. (2001). Gastroenterology, 120:1607–1619. http://dx.doi.org/10.1053/ gast.2001.25065 PMID:11375943
4. Hvid-Jensen F et al. (2011). N Engl J Med, 365:1375–1383. http://dx.doi.org/10.1056/ NEJMoa1103042 PMID:21995385
5. Bhat S et al. (2011). J Natl Cancer Inst, 103:1049–1057. http://dx.doi.org/10.1093/ jnci/djr203 PMID:21680910
6. de Jonge PJ et al. (2010). Gut, 59: 1030–1036. http://dx.doi.org/10.1136/gut. 2009.176701 PMID:20639249
7. Jung KW et al. (2011). Am J Gastroenterol, 106:1447–1455. http://dx.doi.org/10.1038/ ajg.2011.130 PMID:21483461
8. Rugge M et al. (2012). Ann Surg, 256:788–794. http://dx.doi.org/10.1097/ SLA.0b013e3182737a7e PMID:23095623
oesophageal adenocarcinoma, pop- ulation-based studies show limited influence of familial risk, consistent with the rapid changes in incidence rates observed in many popula- tions over a very short time period. The epidemiological features of ade- nocarcinoma of the oesophagus and of the oesophago-gastric junction match those of patients with known intestinal metaplasia in the distal oe- sophagus, i.e. Barrett oesophagus [15]. Barrett oesophagus has been identified as the single most impor- tant precursor lesion and risk factor for oesophageal adenocarcinoma,
irrespective of the length of the segment with intestinal metaplasia. Intestinal metaplasia of the oesopha- gus develops when the squamous oesophageal epithelium is replaced by columnar epithelium during the process of healing after repetitive injury, typically associated with gastro-oesophageal reflux disease. Intestinal metaplasia can be detect- ed in the majority of patients with oesophageal adenocarcinoma [17].
A meta-analysis of studies in pa- tients with Barrett oesophagus dem- onstrated an incidence of progres- sion to adenocarcinoma of about
6.1 per 1000 person-years, which reduced to 3.9 per 1000 person- years when only high-quality studies were considered [18], correspond- ing to a lifetime risk of about 10% in these patients. However, these results may be overestimations that reflect various types of bias in the studies. More recently, an Irish study reported a rate of 3.8 per 1000 person-years [19], whereas a Danish study showed a considerably lower progression rate of only 0.12% per year [20], although obesity is far less prevalent in the Danish population than in the United States or Irish
380
population (see “Risk of malignant progression in Barrett oesophagus patients”).
Chronic gastro-oesophageal re- flux disease is the usual source of mucosal injury and also provides an abnormal environment during the healing process that predis- poses to intestinal metaplasia and adenocarcinoma. In a large Swedish study, individuals with long-stand- ing and severe gastro-oesophage- al reflux disease symptoms had a 40-fold increased risk of oesopha- geal adenocarcinoma [21]. Both experimental and clinical data in- dicate that combined oesophageal exposure to both gastric acid and duodenal contents (bile acids and pancreatic enzymes) appears to be more carcinogenic than exposure to only gastric juice or duodenal contents. An association between alcohol consumption and oesopha- geal adenocarcinoma has not been well established. A reduced risk of oesophageal adenocarcinoma is associated with Helicobacter pylori infection, and particularly H. pylori strains with the CagA protein. A protective effect has been suggested after the use of non-steroidal anti- inflammatory drugs, but not all stud- ies have supported such data.
Pathology and genetics Adenocarcinoma in Barrett oesoph- agus develops through a progres- sive sequence of morphologically
identifiable premalignant lesions termed “dysplasia”. Dysplasia is diagnosed by the presence and de- gree of cytological and architectural atypia [22]. Low-grade dysplasia shows pits with relatively preserved architecture, or minimal distortion, containing cells with atypical nuclei limited to the basal portion of the cell cytoplasm. High-grade dyspla- sia is diagnosed by the presence of marked cytological abnormalities and/or significant architectural com- plexity of the glands (Fig. 5.3.4A). Cytological abnormalities include nuclear pleomorphism and loss of polarity, irregularity of nuclear contour, and increased nuclear- to-cytoplasmic ratio. Architectural
abnormalities of high-grade chang- es include crypt budding, branch- ing, marked crowding, and, rarely, a cribriform growth pattern.
Intramucosal adenocarcinoma (invasion into the lamina propria; T1a) show single cells or small clus- ters of compact back-to-back glands within the lamina propria, a cribri- form or solid pattern of growth with expansion and distortion of the ad- jacent crypts, and a highly distorted/ irregular glandular proliferation not explained by the presence of pre- existing glands. The presence of necrosis and/or desmoplasia is evi- dence in favour of adenocarcinomas as well, although these features are rarely present in carcinomas limited to the mucosa.
Oesophageal adenocarcinomas are typically papillary and/or tubu- lar. Most cases are of the intestinal type in the Laurén classification (Fig. 5.3.4B), whereas some are of the diffuse type. Multiple ge- netic alterations are involved in the development and progression of Barrett oesophagus to oesophage- al adenocarcinoma, encompassing tumour suppressor genes, onco- genes, growth factor receptors, or enzymes that play important roles in diverse cellular functions such as cell-cycle control, apoptosis, cell signalling, cell adhesion and genetic stability, signal transduction, and DNA repair. Gains in the region of
Fig. 5.3.4. Oesophageal adenocarcinoma and precursor lesions. (A) High-grade colum- nar epithelial dysplasia in Barrett oesophagus. Note the markedly hyperchromatic nuclei at the surface of the sample (mucosal biopsy). This process in glandular mucosa is a precursor to adenocarcinoma. (B) Oesophageal adenocarcinoma. This gland-forming malignant neoplasm has the appearance of an intestinal-type adenocarcinoma, in the Laurén classification for gastric carcinomas.
A B
Fig. 5.3.5. Schematic of protein alterations in DOCK2 and ELMO1 detected by whole- exome sequencing of oesophageal adenocarcinoma. Coding alterations in oesophageal adenocarcinoma are coloured either black (missense) or red (splice site or nonsense); silent mutations are depicted in grey. Conserved domain mapping is from UniProt. DHR, Dlg homologous region, ELMO, engulfment and cell motility; PH, Pleckstrin homology; SH3, SRC homology 3.
Chapter 5.3 • Oesophageal cancer 381
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .3
8q (region of c-myc) and 20q, and losses at 3p (FHIT), 4q, 5q (APC) and 18q (SMAD4, DCC) have all been reported, with an increasing number of chromosomal alterations in the sequence from metaplasia to intraepithelial neoplasia to carci- noma. Gene silencing by promoter hypermethylation has been demon- strated frequently for the CDH1 gene (which encodes E-cadherin) and the APC, p16, MGMT, and HPP1 genes. Disorders of ARID are also noted.
Exome and whole-genome se- quencing of 149 and 15 oesopha- geal adenocarcinomas has been recently reported [23]. A mutational signature was defined by a high prevalence of A → C transversions
at AA dinucleotides. Of 26 signifi- cantly mutated genes, TP53, CDKN2, SMAD4, ARID4, and PIK3CA had been implicated previously. The new significantly mutated genes included chromatin modifying factors and candidate contributors SPG20, TLR4, ELMO1, and DOCK2 (Fig. 5.3.5). Functional analysis of the relevant mutations in ELMO1 identified in- creased cellular invasion.
Prospects Novel endoscopic techniques for ear- ly detection and treatment of oesoph- ageal adenocarcinoma offer prom- ise. Ablation therapy for high-grade dysplasia (the strongest risk factor for adenocarcinoma) is probably
cost-effective [24]. Ideal strategies for population-based screening in high-risk patients remain poorly es- tablished and inconsistent.
A recent review [25] determined that endoscopic screening is inva- sive, costly, and error-prone owing to sampling bias and the subjec- tive diagnosis of dysplasia. Non- endoscopic cell sampling methods (Fig. 5.3.6) are less invasive and more cost-effective than endoscopy, but the sensitivity and specificity of cytological assessment of atypia have been disappointing. The use of biomarkers to analyse samples collected using pan-oesophageal cell collection devices may improve diagnostic accuracy.
Fig. 5.3.6. Oesophageal cell collection using non-endoscopic devices. (A) The deflated balloon is inserted into the stomach. (B) Once deemed to be in the stomach, the balloon is inflated through the attached catheter and pulled up the oesophagus. Mucosal cells adhere to the balloon, which is then deflated and removed. A similar procedure based on the use of a cytosponge may also be used.
A B
382
References 1. Aghcheli K, Marjani HA, Nasrollahzadeh D
et al. (2011). Prognostic factors for esoph- ageal squamous cell carcinoma – a popu- lation-based study in Golestan Province, Iran, a high incidence area. PLoS One, 6:e22152. http://dx.doi.org/10.1371/journal. pone.0022152 PMID:21811567
2. Pandeya N, Olsen CM, Whiteman DC (2013). Sex differences in the proportion of esophageal squamous cell carcinoma cases attributable to tobacco smoking and alcohol consumption. Cancer Epidemiol, 37:579–584. http://dx.doi.org/10.1016/j. canep.2013.05.011 PMID:23830137
3. Morita M, Kumashiro R, Kubo N et al. (2010). Alcohol drinking, cigarette smoking, and the development of squamous cell carcinoma of the esophagus: epidemiology, clinical findings, and prevention. Int J Clin Oncol, 15:126–134. http://dx.doi.org/10.1007/s101 47-010-0056-7 PMID:20224884
4. Brooks PJ, Enoch MA, Goldman D et al. (2009). The alcohol flushing response: an unrecognized risk factor for esophageal cancer from alcohol consumption [Review]. PLoS Med, 6:e50. http://dx.doi.org/10.1371/ journal.pmed.1000050 PMID:19320537
5. Wu C, Kraft P, Zhai K et al. (2012). Genome- wide association analyses of esophageal squamous cell carcinoma in Chinese iden- tify multiple susceptibility loci and gene- environment interactions. Nat Genet, 44:1090–1097. http://dx.doi.org/10.1038/ ng.2411 PMID:22960999
6. Wu C, Hu Z, He Z et al. (2011). Genome- wide association study identifies three new susceptibility loci for esophageal squa- mous-cell carcinoma in Chinese popula- tions. Nat Genet, 43:679–684. http://dx.doi. org/10.1038/ng.849 PMID:21642993
7. Taylor PR, Abnet CC, Dawsey SM (2013). Squamous dysplasia – the precursor le- sion for esophageal squamous cell carci- noma. Cancer Epidemiol Biomarkers Prev, 22:540–552. http://dx.doi.org/10.1158/10 55-9965.EPI-12-1347 PMID:23549398
8. Chen J, Kwong DL, Cao T et al. (2013). Esophageal squamous cell carcinoma (ESCC): advance in genomics and molecu- lar genetics. Dis Esophagus, http://dx.doi. org/10.1111/dote.12088 PMID:23796192
9. Maeng CH, Lee J, van Hummelen P et al. (2012). High-throughput genotyping in metastatic esophageal squamous cell carcinoma identifies phosphoinositide- 3-kinase and BRAF mutations. PLoS One, 7:e41655. http://dx.doi.org/10.1371/journal. pone.0041655 PMID:22870241
10. Agrawal N, Jiao Y, Bettegowda C et al. (2012). Comparative genomic analysis of esophageal adenocarcinoma and squa- mous cell carcinoma. Cancer Discov, 2:899–905. http://dx.doi.org/10.1158/2159- 8290.CD-12-0189 PMID:22877736
11. Blaydon DC, Etheridge SL, Risk JM et al. (2012). RHBDF2 mutations are associated with tylosis, a familial esophageal cancer syndrome. Am J Hum Genet, 90:340–346. http://dx.doi.org/10.1016/j.ajhg.2011.12.008 PMID:22265016
12. Dubecz A, Gall I, Solymosi N et al. (2012). Temporal trends in long-term survival and cure rates in esophageal cancer: a SEER database analysis. J Thorac Oncol, 7:443– 447. http://dx.doi.org/10.1097/JTO. 0b013e3182397751 PMID:22173700
13. Bergman JJ, Zhang YM, He S et al. (2011). Outcomes from a prospective trial of en- doscopic radiofrequency ablation of early squamous cell neoplasia of the esopha- gus. Gastrointest Endosc, 74:1181–1190. http://dx.doi.org/10.1016/j.gie.2011.05.024 PMID:21839994
14. Engel LS, Chow WH, Vaughan TL et al. (2003). Population attributable risks of esophageal and gastric cancers. J Natl Cancer Inst, 95:1404–1413. http://dx.doi. org/10.1093/jnci/djg047 PMID:13130116
15. Simard EP, Ward EM, Siegel R, Jemal A (2012). Cancers with increasing incidence trends in the United States: 1999 through 2008. CA Cancer J Clin, 62:118–128. ht tp://dx.doi.org /10.3 32 2 /c aac. 20141 PMID:22281605
16. Dikken JL, Lemmens VE, Wouters MW et al. (2012). Increased incidence and survival for oesophageal cancer but not for gastric car- dia cancer in the Netherlands. Eur J Cancer, 48:1624–1632. http://dx.doi.org/10.1016/j. ejca.2012.01.009 PMID:22317953
17. Chandrasoma P, Wijetunge S, DeMeester S et al. (2012). Columnar-lined esopha- gus without intestinal metaplasia has no proven risk of adenocarcinoma. Am J Surg Pathol, 36:1–7. http://dx.doi.org/10.1097/ PAS.0b013e31822a5a2c PMID:21959311
18. Yousef F, Cardwell C, Cantwell MM et al. (2008). The incidence of esophageal can- cer and high-grade dysplasia in Barrett’s esophagus: a systematic review and meta- analysis. Am J Epidemiol, 168:237–249. ht tp: //dx.doi.org /10.10 93 /aje / k wn121 PMID:18550563
19. Bhat S, Coleman HG, Yousef F et al. (2011). Risk of malignant progression in Barrett’s esophagus patients: results from a large population-based study. J Natl Cancer Inst, 103:1049–1057. http://dx.doi.org/10.1093/ jnci/djr203 PMID:21680910
20. Hvid-Jensen F, Pedersen L, Drewes AM et al. (2011). Incidence of adenocarcinoma among patients with Barrett’s esophagus. N Engl J Med, 365:1375–1383. http:// dx.do i.o r g /10.10 5 6 / N E J M oa110 3 0 42 PMID:21995385
21. Lagergren J, Bergström R, Lindgren A, Nyrén O (1999). Symptomatic gastroesoph- ageal reflux as a risk factor for esopha- geal adenocarcinoma. N Engl J Med, 340:825–831. http://dx.doi.org/10.1056/NE JM199903183401101 PMID:10080844
22. Montgomery E, Bronner MP, Goldblum JR et al. (2001). Reproducibility of the diag- nosis of dysplasia in Barrett esophagus: a reaffirmation. Hum Pathol, 32:368–378. http://dx.doi.org/10.1053/hupa.2001.23510 PMID:11331953
23. Dulak AM, Stojanov P, Peng S et al. (2013). Exome and whole-genome sequencing of esophageal adenocarcinoma identi- fies recurrent driver events and muta- tional complexity. Nat Genet, 45:478–486. http://dx.doi.org/10.1038/ng.2591 PMID: 23525077
24. Hur C, Choi SE, Rubenstein JH et al. (2012). The cost effectiveness of radiofre- quency ablation for Barrett’s esophagus. Gastroenterology, 143:567–575. http://dx. doi.org /10.10 5 3/ j.gastro. 2012.0 5.010 PMID:22626608
25. Lao-Sirieix P, Fitzgerald RC (2012). Screening for oesophageal cancer. Nat Rev Clin Oncol, 9:278–287. http://dx. doi.org/10.1038/nrclinonc.2012.35 PMID: 22430857
Chapter 5.4 • Stomach cancer 383
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .4
Summary • Most stomach cancers (90%)
are sporadic. In almost all coun- tries, a steady decline in gastric cancer mortality rates has oc- curred in the past few decades.
• Helicobacter pylori is the major environmental cause of gastric cancer development. Other fac- tors contributing to risk include dietary composition – particu- larly intake of pickled vegeta- bles – and smoking.
• Molecular pathogenesis of gas- tric cancer is complex. Genes related to receptor tyrosine ki- nase (RTK)/RAS signalling are frequently amplified, in par- ticular FGFR2, KRAS, ERBB2, EGFR, and MET. The product of the HER2 oncogene, when either overexpressed or ampli- fied, is a prognostic factor.
• Hereditary gastric cancer con- tributes to 1–3% of the bur- den of stomach cancer. Two syndromes have been identi- fied: hereditary diffuse gastric cancer and gastric adenocar- cinoma and proximal polyposis of the stomach.
• Hereditary diffuse gastric can- cer is caused by germline alter- ations of the E-cadherin (CDH1)
5.4 5 ORGAN SITE
Stomach cancer Fátima Carneiro Yung-Jue Bang (reviewer)
Takanori Hattori (reviewer) Elizabeth A. Montgomery (reviewer)
gene; prophylactic gastrectomy is recommended for asymp- tomatic carriers of pathogenic CDH1 mutations.
Most stomach cancers are gas- tric carcinomas, which are malig- nant epithelial neoplasms. Non- epithelial tumours of the stomach predominantly include lymphomas and mesenchymal tumours. Gastric carcinomas represent a biological- ly and genetically heterogeneous group of tumours with multifactorial etiologies, both environmental and genetic. They are characterized by broad morphological heterogene- ity with respect to patterns of archi- tecture and growth, cell differentia- tion, histogenesis, and molecular pathogenesis.
Most cases of gastric cancer are sporadic, and familial clustering is observed in about 10% of cases. Hereditary gastric cancer accounts for a very low percentage of cases (1–3%), and two hereditary syn- dromes have been characterized: hereditary diffuse gastric cancer and gastric adenocarcinoma and proximal polyposis of the stomach. Furthermore, gastric cancer can develop in the setting of other he- reditary cancer syndromes.
Etiology Gastric carcinogenesis is a multi- step and multifactorial process that, in many cases, appears to involve a progression from normal mucosa through chronic gastritis (chronic in- flammation of the gastric mucosa), atrophic gastritis (with loss of gastric glands), and intestinal metaplasia (substitution of gastric epithelium by intestinal epithelium) to dyspla- sia (intraepithelial neoplasia) and carcinoma, a sequence of events that may last several years. This sequence has been designated as the Correa cascade of multistep gastric carcinogenesis [1]. However, the Correa model does not explain all carcinogenic steps of gastric can- cer. Actually, a proportion of gastric adenocarcinomas arise in non- intestinalized mucosa and retain a gastric phenotype (and gastric dif- ferentiation is also observed in gas- tric dysplasia, the ultimate precursor lesion of gastric adenocarcinoma). Also, in the recently identified syn- drome gastric adenocarcinoma and proximal polyposis of the stomach (discussed below), gastric dyspla- sia and gastric adenocarcinoma develop in fundic gland polyps of the proximal stomach and are re- lated not to intestinal metaplasia but instead to foveolar hyperplasia. Together, this evidence shows that gastric adenocarcinoma can arise
384
Epidemiology
Stomach cancer • Stomach cancer is the fifth most
common cancer worldwide, with an estimated 952 000 new cases (7% of total cancer in- cidence) and 723 000 deaths (9% of total cancer mortality) in 2012. Almost three quarters of the new cases occurred in Asia, and more than two fifths occurred in China.
• There is a 10-fold international variation in stomach cancer inci- dence; rates in men are approxi- mately double those observed in women.
• The highest age-standardized incidence rates are in East Asia and central and eastern Europe. Incidence rates tend to be rela- tively low in Africa and in North America.
• The case fatality rate is lower in countries with high levels of human development (overall mortality-to-incidence ratio, 0.65) than in countries at low or medium levels of human devel- opment (0.83).
• Over the past 50 years, inci- dence and mortality rates of the non-cardia type of gastric cancer have been uniformly de- creasing in almost all countries; rates of gastric cardia cancer have, however, been stable or increasing in the past two to three decades.
Map 5.4.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for stomach cancer in men, 2012.
Map 5.4.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for stomach cancer in women, 2012.
from the gastric epithelium. These findings challenge the classic pro- posed histogenetic pathway from chronic atrophic gastritis through intestinal metaplasia to adenocar- cinoma (with glandular structure), which is frequently designated as intestinal carcinoma (Laurén clas- sification), a misnomer for these
cases. Another type of gastric car- cinoma, so-called diffuse carcinoma (Laurén classification), probably de- velops de novo from gastric mucosa. Except for the diffuse carcinomas developing in a hereditary setting (discussed below), no precursor le- sions have been well characterized for this type of gastric cancer.
Among environmental factors con- tributing to increased risk of gastric cancer, Helicobacter pylori infection plays a major role. Almost all non- cardia gastric cancers develop from a background of H. pylori-infected mu- cosa [2]. H. pylori is a Gram-negative bacterium that colonizes the gastric mucosa. In 1994, IARC categorized
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.4 • Stomach cancer 385
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .4
Chart 5.4.1. Estimated global number of new cases and deaths with proportions by major world regions, for stomach cancer in both sexes combined, 2012.
Chart 5.4.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for stomach cancer in men, circa 1975–2012.
Chart 5.4.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for stomach cancer in women, circa 1975–2012.
386
H. pylori as a Group 1 carcinogen for gastric cancer [3] based on results of epidemiological studies that were available at that time; this conclu- sion was later confirmed. H. pylori can be identified in gastric mucosa by routine stains, such as haema- toxylin–eosin and modified Giemsa, and other ancillary methods such as Warthin–Starry staining and im- munohistochemistry (Fig. 5.4.1). The infection is commonly acquired during early childhood and per- sists throughout adult life unless eradicated.
Factors associated with coloni- zation and pathogenicity of H. pylori include virulence factors – cagA in the cag pathogenicity island and the vacuolating cytotoxin vacA [4] – as well as outer membrane proteins of the bacterium. Strains producing the cagA protein that induce a greater de- gree of inflammation are associated with gastric precancerous lesions and a greater risk of developing cancer of the distal stomach [4]. Although the risk of gastric cancer in some coun- tries of Europe and North America has been related to vacA genotype, such relationships have not been observed in countries in East Asia, suggesting that consequences of variation in vacuolating activity are dependent on
geographical region. Corpus- predominant gastritis with multifocal gastric atrophy and hypochlorhydria or achlorhydria is seen in approximately 1% of subjects infected with H. pylori. As a direct consequence of elevation in gastric pH, there is a change in gas- tric flora, with colonization by anaero- bic bacteria responsible for the forma- tion of carcinogenic nitrosamines.
Certain dietary habits are as- sociated with an increased risk of gastric cancer [5]. These include high intakes of salt-preserved and/ or smoked foods and low intakes of fresh fruit and vegetables. An analy- sis of 60 relevant studies suggested a potential 50% higher risk of gas- tric cancer associated with intake of pickled vegetables, and perhaps
Fig. 5.4.1. Helicobacter pylori detected in the lumen of gastric glands and adherent to the apical pole of epithelial cells: (A) haematoxylin–eosin (inset: modified Giemsa); (B) Warthin–Starry; (C) immunohistochemistry.
Fig. 5.4.2. Food in Seoul, Republic of Korea, typically includes pickled vegetables, the consumption of which is associated with increased risk of gastric cancer.
Chapter 5.4 • Stomach cancer 387
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .4
stronger associations in the Republic of Korea and China (Fig. 5.4.2) [6]. Intake of all meat, specifically in- cluding red meat and processed meat, is also associated with an increased risk of gastric cancer in the distal stomach. The so-called “Mediterranean diet” was shown to be associated with a significant re- duction in the risk of gastric cancer incidence; this diet is characterized as involving high consumption of fruit, vegetables, cereals, legumes, nuts and seeds, and seafood, with olive oil as the main fat source, mod- erate alcohol consumption (particu- larly red wine), a low to moderate consumption of dairy products, and a relatively low consumption of red and processed meat.
Tobacco smoking causes stom- ach cancer; the epidemiological as- sociation is not explicable by bias or confounding factors. Smoking also potentiates the carcinogenic ef- fect of infection with cagA-positive H. pylori.
Polymorphisms of the interleukin- 1β (IL1B) gene, which contributes to initiation and amplification of inflam- matory response, and of the inter- leukin-1 receptor antagonist (IL1RN) gene, which modulates inflamma- tion, are associated with individual or familial susceptibility to carcino- genesis associated with H. pylori. Among individuals with alleles that predispose to inflammation, infection with H. pylori may cause increased production of gastric interleukin-1β, leading to severe and sustained in- flammation that increases the risk of developing gastric cancer [7].
Pathology Two major types of gastric carci- noma were described by Laurén in 1965 [8]: the intestinal and diffuse types, which display different clin- icopathological profiles and molec- ular pathogenesis, and often occur in distinct epidemiologic settings. Currently, five major types of gas- tric cancer are recognized by WHO: tubular (Fig. 5.4.3A), papillary, mu- cinous, poorly cohesive (with or with- out signet ring cells) (Fig. 5.4.3B),
and mixed [9]. Tubular and papillary carcinomas roughly correspond to the intestinal type in the Laurén classification, and poorly cohesive carcinomas (encompassing cases constituted, partially or totally, by signet ring cells) correspond to the diffuse type. Furthermore, rare vari- ants account for about 10% of gastric carcinomas.
The gastric antrum (distal stom- ach) is the most common site of gas- tric carcinoma. Carcinomas of the oesophago-gastric junction have been most commonly reported in populations in North America and Europe, associated with gastro- oesophageal reflux disease and other characteristics similar to ad- enocarcinoma arising in Barrett oesophagus, and are unrelated to H. pylori infection. However, in parts of Asia, for example in China, but also for a subset of tumours diag- nosed in North America and Europe, neoplasms of the proximal stomach arise in a setting of chronic atrophic gastritis with H. pylori infection and are similar to distal gastric cancer. Adenocarcinomas located entirely below the oesophago-gastric junc- tion are considered gastric in origin, and for these tumours the use of the ambiguous and often misleading term “carcinoma of the gastric car- dia” is discouraged in favour of “car- cinoma of the proximal stomach”.
Early gastric carcinoma is an in- vasive carcinoma limited to the mu- cosa or submucosa, regardless of nodal status. The term “early” does not imply a stage in the genesis of the cancer but means that these are gastric cancers that can often be cured. However, if untreated, 63% of early gastric carcinomas progress to advanced tumours within 5 years. Countries with a high incidence of gastric cancer and in which asymp- tomatic patients are screened have a high incidence of early gastric carcinomas, ranging from 30% to 50% in Japan and the Republic of Korea, with lower figures for west- ern Europe and North America (16–24%). Lymph-node metastasis occurs in 10–20% of all early gastric carcinomas and correlates with the depth of submucosal invasion and increasing tumour diameter.
Advanced gastric carcinoma invades the muscularis propria of the stomach and beyond. Most pa- tients with advanced carcinoma have lymph-node metastatic disease for which only palliative surgery can be considered. Lymphatic and vascular invasion, often seen in advanced cases, indicate a poor prognosis. Gastric carcinomas can spread by direct extension to adjacent organs, metastasis, or peritoneal dissemina- tion. Carcinomas of the intestinal type preferentially metastasize haema- togenously to the liver, whereas
Fig. 5.4.3. Main histological types of gastric carcinoma: (A) intestinal carcinoma (Laurén classification), tubular type (WHO classification); (B) diffuse carcinoma (Laurén classification), poorly cohesive type, with signet ring cells (WHO classification).
388
Gastric cancer prevention Il Ju Choi
Gastric cancer prevention consists of primary prevention (removing possible causes) and secondary prevention (early detection). Primary prevention is a feasible approach considering the marked decrease in gastric cancer incidence during the past century in industrialized countries. This decrease suggests the major effect of environmental factors on stomach cancer develop- ment. Dietary intervention, includ- ing increased intake of fresh fruit and vegetables and reduced intake of salt and processed or smoked meat, is a main primary prevention strategy.
Helicobacter pylori eradication is another promising strategy. Despite its importance, only two studies have evaluated gastric cancer develop- ment as a primary outcome. The first study (of 1630 healthy individuals from a high-risk region of China, in 2004) found that gastric cancer in- cidence over 7.5 years was similar in H. pylori eradication and placebo groups. [1]. From these results, H. pylori eradication in the general population is currently not consid- ered to have definite supporting data. Eradication of H. pylori, how- ever, may decrease stomach cancer development in the subgroup with no precancerous changes such as atrophy or intestinal metaplasia. The second study (of 544 randomized patients who underwent endoscopic
resection for early gastric cancer in Japan, in 2008) showed at a 3-year follow-up that risk of metachronous gastric carcinoma was reduced to one third in the eradication group [2]. This is the first randomized study showing that H. pylori eradication is effective for stomach cancer preven- tion; however, it has the limitation of not being blinded and of having a short follow-up duration.
Currently, a study recruiting family members of gastric cancer patients, a high-risk population, is under way in the Republic of Korea (ClinicalTrials.gov: NCT01678027). Furthermore, a study with enough statistical power, with recruitment of a larger number of patients from the general population, is also needed before implementation of this strat- egy in the general population from high-risk regions.
Secondary prevention is an- other important aspect of gastric cancer prevention. In the Republic of Korea, where gastric cancer inci- dence is the highest in the world, the National Cancer Screening Program has been providing biennial gastric cancer screening for people aged 40 years or older since 1999. This is a unique screening programme, which uses endoscopy (which has a much higher sensitivity than radiol- ogy with barium meal) as a primary screening tool. Probably owing to the screening effect, the 5-year survival rate of stomach cancer patients in the Republic of Korea has improved, from 46.6% in 1996–2000 to 67.0% in 2006–2010. Furthermore, endo- scopic screening at 2-year intervals detects stomach cancer mostly as an early gastric cancer confined to the mucosal or submucosal layers (Fig. B5.4.1), and many such can- cers can be cured by endoscopic resection rather than by surgical gastrectomy [3].
References 1. Wong BC et al. (2004). JAMA, 291:187–194.
http://dx.doi.org/10.1001/jama.291.2.187 PMID:14722144
2. Fukase K et al. (2008). Lancet, 372:392–397. http://dx.doi.org/10.1016/S0140-6736(08) 61159-9 PMID:18675689
3. Nam SY et al. (2009). Eur J Gastroenterol Hepatol, 21:855–860. http://dx.doi.org/10. 1097/MEG.0b013e328318ed42 PMID: 19369882
Fig. B5.4.1. Early gastric cancer detect- ed by screening endoscopy.
carcinomas composed of poorly co- hesive cells (the diffuse type) pref- erentially metastasize to peritoneal surfaces. Mixed carcinomas exhibit the metastatic patterns of both types. When the carcinoma penetrates the serosa, peritoneal implants generally flourish.
Genetic predisposition and hereditary syndromes First-degree relatives of patients with gastric cancer are almost 3 times as
likely as the general population to develop gastric cancer themselves. This may be partly attributable to H. pylori infection being common in families and to the potential role of IL-1 gene polymorphisms. Susceptibility to carcinogens may play a role as well. For example, polymorphisms of genes encod- ing for glutathione S-transferase enzymes, known to metabolize tobacco-related carcinogens and N-acetyltransferase 1, increase the risk of gastric cancer development.
There is evidence also of familial clustering: about 10% of stomach cancers show evidence of a famil- ial component, and approximately 1–3% of gastric cancers are a re- sult of an inherited predisposition, of which the major type is hereditary diffuse gastric cancer [10]. Genome- wide association studies have im- plicated the prostate stem cell an- tigen (PSCA) gene and the mucin 1 (MUC1) gene as influencing sus- ceptibility. Approximately 95% of the Japanese population have at least
Chapter 5.4 • Stomach cancer 389
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .4
one of the two risk genotypes, and approximately 56% of the population have both risk genotypes [11].
Hereditary diffuse gastric cancer On the basis of clinical criteria, the International Gastric Cancer Linkage Consortium defined families with he- reditary diffuse gastric cancer syn- drome as meeting one of two criteria: (i) two or more documented cases of diffuse gastric cancer in first- or second-degree relatives, with at least one of them diagnosed before the age of 50 years, or (ii) three or more documented cases of diffuse gastric cancer in first- or second- degree relatives, independent of age of diagnosis [12]. Women in these families have an elevated risk of lobular breast cancer. The criteria for genetic testing were updated in 2010 [13].
Alterations of the CDH1 gene, which encodes E-cadherin, con- stitute the genetic causal event of hereditary diffuse gastric cancer [14]. In clinically defined heredi- tary diffuse gastric cancer, CDH1
mutations are detected in 30–40% of cases. Most (75–80%) are truncat- ing mutations, and the remainder are missense mutations. In addition to point mutations, large germline dele- tions have been found in hereditary diffuse gastric cancer families that tested negative for point mutations.
A model of early development of diffuse gastric cancer in CDH1 mutation carriers has been pro- posed, encompassing precursor (intraepithelial) lesions (in situ carci- noma and pagetoid spread of signet ring cells), early intramucosal car- cinoma, and advanced cancer [15].
Gastric adenocarcinoma and proximal polyposis of the stomach Recently, a new hereditary syndrome has been identified: gastric adeno- carcinoma and proximal polyposis of the stomach, characterized by the autosomal dominant transmission of fundic gland polyposis, including areas of dysplasia or intestinal-type gastric adenocarcinoma, restricted to the proximal stomach, with no evidence of colorectal or duodenal
polyposis or other heritable gastro- intestinal cancer syndromes. The genetic defect behind this syndrome has not yet been elucidated [16].
Gastric cancer in other hereditary cancer syndromes The risk of gastric cancer is also in- creased in dominantly inherited can- cer predisposition syndromes such as familial adenomatous polyposis and Lynch syndrome, and also in Li– Fraumeni syndrome with germline mutation of TP53.
Molecular pathology Gastric carcinoma is the result of accumulated genomic damage af- fecting cellular functions essential for cancer development: self-suf- ficiency in growth signals, escape from anti-growth signals, resistance to apoptosis, sustained replicative potential, angiogenesis induction, and invasive or metastatic poten- tial. These genomic changes arise through three genomic instability pathways: microsatellite instabil- ity, chromosomal instability, and a CpG island methylation phenotype. Furthermore, genetic and epigenetic changes affect oncogenes and tu- mour suppressor genes [17].
Some oncogenes are preferen- tially altered in a specific type of gastric cancer, such as HER2 and KRAS in the intestinal type. Human epidermal growth factor receptor 2 (HER2) overexpression and/or amplification is present in 10–20% of gastric cancers, and it has been suggested that HER2 overexpres- sion and/or amplification may be one of the molecular abnormalities linked to the development of gastric cancer, with a negative impact on progno- sis [18]. Recently, a comprehensive survey of genomic alterations in gastric cancer revealed systematic patterns of molecular exclusivity and genes related to receptor tyrosine ki- nase (RTK)/RAS signalling: FGFR2 (in 9% of tumours), KRAS (9%), EGFR (8%), ERBB2 (7%), and MET (4%) [19]. These genes were frequently amplified in gastric can- cer in a mutually exclusive manner
Fig. 5.4.4. A woman gives her daughter a banana from an electrical refrigerator, in a photograph from 1946. Refrigerators have allowed better preservation of food, thus reducing the need for salt preservation.
390
(Fig. 5.4.5), indicating five distinct gastric cancer patient subgroups. Collectively, these subgroups sug- gest that at least 37% of gastric cancer patients may be potentially treatable by RTK/RAS-directed therapies. A separate study revealed KRAS mutations in 4.2% of gastric tumour samples from the United Kingdom, Japan, and Singapore, demonstrating that KRAS mutation and DNA mismatch repair deficiency have a role in a small subgroup of these tumours [20].
Some oncogenes are altered preferentially in diffuse carcinoma, such as BCL2 and FGFR2 (for- merly K-sam). Other oncogenes are altered in both intestinal-type and diffuse carcinomas, including CTNNB1 (encoding β-catenin), MET, and MYC. Many tumour suppressor genes have been implicated in gas- tric carcinoma development, includ- ing APC and DCC in intestinal-type carcinomas and CDH1 and RB1 in diffuse carcinomas. Other tumour suppressor genes are altered in both
types of gastric carcinoma, such as PTEN and TP53, although these are more common in intestinal-type carcinoma.
Prevention, screening, and targeted agents The two major changes that could be made at a population level to reduce gastric cancer incidence are improvement in diet and reduction in the prevalence of H. pylori. These changes are already taking place in many populations, as a conse- quence of economic development, and may explain the decline ob- served in gastric cancer incidence. Active intervention in a population requires proof that the intervention is effective, and this can only come from randomized trials. Prevention of gastric cancer is a primary public health issue in Japan [21] and the Republic of Korea (see “Gastric cancer prevention”).
Radiology with barium meal is used in mass screening protocols in Japan, followed by endoscopy if
an abnormality is detected. In the Republic of Korea, endoscopy is primarily used as a screening proce- dure. Testing for serum pepsinogen is being evaluated for screening to identify high-risk patients and detect early cancers.
Endoscopy is a sensitive and specific diagnostic test for gastric cancer. Although detection of le- sions associated with early gastric cancer can be improved using chro- mo-endoscopy and narrow-band imaging, a substantial number of such lesions still escape detection. Tumour staging before deciding on a treatment plan involves endoscopic ultrasonography for characteriza- tion of the primary tumour but is less useful for lymph-node staging, whereas computed tomography (CT) is used to detect lymph-node and liver metastases. Positron emis- sion tomography in combination with CT imaging may be superior to either alone for preoperative stag- ing. Laparoscopic staging may be the only way to exclude peritoneal seeding.
There is current interest in the immunohistochemical and in situ hybridization detection of HER2 ex- pression in gastric cancer because there is evidence that these tumours may respond well to therapy with the humanized monoclonal antibody trastuzumab [22]. The European Medicines Agency and similar au- thorities recommend HER2 testing by immunohistochemistry and fluo- rescence in situ hybridization to iden- tify patients who may benefit from targeted therapy with trastuzumab.
In families fulfilling the criteria for hereditary diffuse gastric cancer, screening for CDH1 germline alter- ations is indicated. Asymptomatic carriers of pathogenic changes of the CDH1 gene are candidates for total prophylactic gastrectomy or annual endoscopic surveillance (in selected groups) as risk reduction strategies.
Fig. 5.4.5. Mutually exclusive and co-amplified genomic alterations in gastric cancer. (A) Focal regions exhibiting mutually exclusive patterns of genome amplification. Outermost circular track indicates genomic positions by chromosomes (black lines are cytobands, red lines are centromeres). Blue lines indicate pairs of focal regions (genes) exhibiting significant patterns of mutually exclusive genomic amplification. Genes in- volved in receptor tyrosine kinase (RTK)/RAS signalling are highlighted in red. (B) Focal regions exhibiting patterns of genomic co-amplification. Orange lines indicate pairs of focal regions (genes) exhibiting significant patterns of genomic co-amplification. Genes involved in RTK/RAS signalling are highlighted in red.
Chapter 5.4 • Stomach cancer 391
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .4
References 1. Correa P (1992). Human gastric carcino-
genesis: a multistep and multifactorial process – First American Cancer Society Award Lecture on Cancer Epidemiology and Prevention. Cancer Res, 52:6735–6740. PMID:1458460
2. IARC (2012). Biological agents. IARC Monogr Eval Carcinog Risks Hum, 100B:1– 441. PMID:23189750
3. IARC (1994). Schistosomes, liver flukes and Helicobacter pylori. IARC Monogr Eval Carcinog Risks Hum, 61:1–241. PMID:7715068
4. Basso D, Zambon CF, Letley DP et al. (2008). Clinical relevance of Helicobacter pylori cagA and vacA gene polymorphisms. Gastroenterology, 135:91–99. http://dx. doi.org/10.1053/j.gastro. 20 0 8.03.0 41 PMID:18474244
5. World Cancer Research Fund/American Institute for Cancer Research (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research.
6. Ren JS, Kamangar F, Forman D, Islami F (2012). Pickled food and risk of gastric can- cer – a systematic review and meta-analysis of English and Chinese literature. Cancer Epidemiol Biomarkers Prev, 21:905–915. http://dx.doi.org/10.1158/1055-9965.EPI- 12-0202 PMID:22499775
7. El-Omar EM, Carrington M, Chow W-H et al. (2000). Interleukin-1 polymorphisms associated with increased risk of gastric cancer. Nature, 404:398–402. http://dx.doi. org/10.1038/35006081 PMID:10746728
8. Laurén P (1965). The two histological main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma. An attempt at a histo-clinical classification. Acta Pathol Microbiol Scand, 64:31–49. PMID:14320675
9. Lauwers GY, Carneiro F, Graham DY et al. (2010). Gastric carcinoma. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, eds. WHO Classification of Tumours of the Digestive System, 4th ed. Lyon: IARC, pp. 48–58.
10. Carneiro F, Charlton A, Huntsman DG (2010). Hereditary diffuse gastric cancer. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, eds. WHO Classification of Tumours of the Digestive System, 4th ed. Lyon: IARC, pp. 59–63.
11. Saeki N, Ono H, Sakamoto H, Yoshida T (2013). Genetic factors related to gastric cancer susceptibility identified using a genome-wide association study. Cancer Sci, 104:1–8. http://dx.doi.org/10.1111/cas. 12042 PMID:23057512
12. Caldas C, Carneiro F, Lynch HT et al. (1999). Familial gastric cancer: overview and guidelines for management. J Med Genet, 36:873–880. PMID:10593993
13. Fitzgerald RC, Hardwick R, Huntsman D et al.; International Gastric Cancer Linkage Consortium (2010). Hereditary diffuse gas- tric cancer: updated consensus guidelines for clinical management and directions for future research. J Med Genet, 47:436–444. http://dx.doi.org/10.1136/jmg.2009.074237 PMID:20591882
14. Guilford P, Hopkins J, Harraway J et al. (1998). E-cadherin germline mutations in familial gastric cancer. Nature, 392:402– 405. http://dx.doi.org/10.1038/32918 PMID: 9537325
15. Carneiro F, Huntsman DG, Smyrk TC et al. (2004). Model of the early development of diffuse gastric cancer in E-cadherin mutation carriers and its implications for patient screening. J Pathol, 203:681– 687. http://dx.doi.org/10.1002/path.1564 PMID:15141383
16. Worthley DL, Phillips KD, Wayte N et al. (2012). Gastric adenocarcinoma and proxi- mal polyposis of the stomach (GAPPS): a new autosomal dominant syndrome. Gut, 61:774–779. http://dx.doi.org/10.1136/gut jnl-2011-300348 PMID:21813476
17. Carneiro F, Oliveira C, Leite M, Seruca R (2008). Molecular targets and biologi- cal modifiers in gastric cancer. Semin Diagn Pathol, 25:274–287. http://dx.doi. org/10.1053/j.semdp.2008.07.004 PMID: 19013893
18. Jørgensen JT, Hersom M (2012). HER2 as a prognostic marker in gastric cancer – a systematic analysis of data from the litera- ture. J Cancer, 3:137–144. http://dx.doi.org/ 10.7150/jca.4090 PMID:22481979
19. Deng N, Goh LK, Wang H et al. (2012). A comprehensive survey of genomic alterations in gastric cancer reveals sys- tematic patterns of molecular exclusivity and co-occurrence among distinct thera- peutic targets. Gut, 61:673–684. http://dx. d o i . o r g /10 .113 6 /g u t j n l - 2 011- 3 018 3 9 PMID:22315472
20. van Grieken NC, Aoyma T, Chambers PA et al. (2013). KRAS and BRAF mutations are rare and related to DNA mismatch repair de- ficiency in gastric cancer from the East and the West: results from a large international multicentre study. Br J Cancer, 108:1495– 1501. http://dx.doi.org/10.1038/bjc.2013. 109 PMID:23511561
21. Asaka M (2013). A new approach for elimi- nation of gastric cancer deaths in Japan. Int J Cancer, 132:1272–1276. http://dx.doi. org/10.1002/ijc.27965 PMID:23180638
22. Bang YJ, Van Cutsem E, Feyereislova A et al.; ToGA Trial Investigators (2010). Trastuzumab in combination with che- motherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial. Lancet, 376: 687–697. http://dx.doi.org/10.1016/S0140- 6736(10)61121-X PMID:20728210
392
Summary • Colorectal cancer is one of the
most common cancers in men and women, representing al- most 10% of the global cancer incidence.
• Dietary composition, obesity, and lack of physical activity are established as contributing to risk of colorectal cancer, but the underlying causative biological processes are not defined.
• Most colorectal carcinomas de- velop through an adenoma–car- cinoma sequence, underpinning screening colonoscopy for ad- enomatous polyp removal as a preventive option.
• Three molecular pathways op- erate: (i) the chromosomal in- stability pathway, characterized by inactivating mutations of the APC, TP53, and TGF-ß genes, mutations of KRAS, and activa- tion of telomerase; (ii) the micro- satellite instability pathway, char- acterized by deficient mismatch repair, arising through mutation of mismatch repair genes or through promoter methylation of MLH1, resulting in a hypermu- tated state; (iii) the CpG island methylation pathway, charac- terized by a high level of gene
5.5 5 ORGAN SITE
Colorectal cancer Fred T. Bosman Stanley R. Hamilton (reviewer)
René Lambert (reviewer)
promoter methylation, along with typical morphology of the precursor lesions in the form of sessile serrated adenomas.
• Familial colorectal cancer com- prises different syndromes, in- cluding familial adenomatous polyposis, hereditary non-pol- yposis colon cancer or Lynch syndrome, MUTYH-associated polyposis, and hamartomatous polyp syndromes.
• Inflammatory bowel disease, when long-standing, predispos- es to colorectal cancer.
• KRAS mutation renders a tumour refractory to anti-EGFR agents, and is therefore predictive.
Colorectal cancer is defined as a carcinoma, usually an adenocar- cinoma, in the colon or rectum [1]. The term “colorectal cancer” might suggest a homogeneous disease entity, but this is clearly not the case. The term “colorectal” is cus- tomarily used as a topographical indication, but in reality the thera- peutic approach for rectal cancer differs significantly from that for colon cancer, primarily due to the quite different anatomical setting of the rectum, which is embedded in a tight space with genitourinary
organs nearby. Cancer in the right colon (caecum, ascending colon) is biologically different from that in the left colon (from the splenic flexure down), in terms of both molecular characteristics and response to tar- geted treatment. Colorectal cancer can develop sporadically, in the context of a familial syndrome, or in the context of inflammatory bow- el disease; each of these settings leads to a disease with distinct characteristics. Recent molecular profiling studies suggest that the heterogeneity of colorectal cancer extends beyond these more or less established categories [1– 4].
Etiology An evaluation by the World Cancer Research Fund [2] has determined that the evidence that physical activ- ity protects against colon cancer is convincing; the evidence is convinc- ing that consumption of red meat and processed meat, consumption of alcoholic beverages (for men, and probably for women), body fatness and abdominal fatness, and the factors that lead to greater adult at- tained height, or its consequences, are causes of colorectal cancer. Consumption of garlic, milk, and calcium probably protects against colorectal cancer.
Therefore, with reference to specifying foods that increase risk
Chapter 5.5 • Colorectal cancer 393
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .5
Epidemiology
Colorectal cancer • Colorectal cancer represents al-
most 10% of the global cancer incidence burden in 2012, and is the third most common cancer in men (an estimated 746 000 cas- es) and the second most com- mon in women (614 000 cases). Colorectal cancer is the fourth most common cause of death from cancer worldwide, with an estimated 694 000 deaths.
• More than 65% of new cases occurred in countries with high or very high levels of human development; almost half of the estimated new cases occurred in Europe and the Americas. Colorectal cancer is now the third most common cancer in areas with high human devel- opment, and the highest inci- dence rates are in men in cen- tral Europe (Slovakia, Hungary, the Czech Republic) and the Republic of Korea.
• Incidence varies 10-fold be- tween countries worldwide, in both sexes, and rates tend to be relatively low in many African countries. As with inci- dence, mortality rates are lower in women than in men, except in the Caribbean.
• The scale of the colorectal cancer incidence burden and its temporal development in a given country are key markers of human development transi- tions, and incidence (and mor- tality) is increasing in many countries transitioning towards higher levels of human develop- ment. In contrast, trends appear to be stabilizing or declining in countries that have attained the highest levels of human development.
Map 5.5.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for colorectal cancer in men, 2012.
Map 5.5.2. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for colorectal cancer in men, 2012.
of colorectal cancer, the primary fo- cus is on meat. Despite the strength of the epidemiological evidence, relevant biological processes are
poorly defined. Cooked meat may contain two classes of carcinogens – heterocyclic amines and polycy- clic aromatic hydrocarbons – but
these are unlikely to account for the evident risk [3]. An indication of the recognized complexity may be gained from the experimental
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
394
Chart 5.5.1. Estimated global number of new cases and deaths with proportions by major world regions, for colorectal cancer in both sexes combined, 2012.
Chart 5.5.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for colorectal cancer in men, circa 1975–2012.
Chart 5.5.3. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for colorectal cancer in men, circa 1975–2012.
Chapter 5.5 • Colorectal cancer 395
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .5
observation that dietary fat can alter the gut microbiota of mice indirectly by affecting bile acids and hence contributing to inflam- matory bowel disease (see be- low) [4]. Such a direct association between diet and carcinogenesis is complemented by current un- derstanding that obesity may play a crucial role in mediating car- cinogenesis via inflammation [5]. Obesity-enhanced inflammation is largely orchestrated by increases in adipose tissue macrophages, leading to the secretion of inflam- matory cytokines, including tumour necrosis factor α (TNF-α), mono- cyte chemoattractant protein 1, and interleukin-6 (IL-6), all of which are linked to colorectal cancer [6].
Colorectal cancer is considered primarily as a “lifestyle” disease; its incidence is high in countries with a diet high in calories and animal fat and with a largely sedentary pop- ulation. One consequence of the lack of precise knowledge as to how multiple factors contribute to risk is that (chemo)preventive measures can only be, and are only, proposed in rather general terms. Solid evi- dence exists for a role of non-ste- roidal anti-inflammatory drugs in prevention of the progression from adenoma to carcinoma [7].
Pathogenesis In the development of colorectal cancer, three distinct pathways operate: the chromosomal instabil- ity pathway, the microsatellite insta- bility pathway, and the CpG island methylator pathway [8–10].
Chromosomal instability pathway The chromosomal instability path- way is found in about 85% of spo- radic colorectal cancers and is pro- totypical for the molecular evolution of this malignancy. Chromosomal abnormalities (mostly a combination of allelic losses and gains) include newly discovered chromosomal translocations [11]. This category of colorectal cancers is characterized by mutations in the Wnt pathway genes APC (70%) or CTNNB (30%), a crucial element in their molecular carcinogenesis [12]. KRAS muta- tions, usually in codons 12 and 13, occur in 45% of the cases and inter- fere with anti-epidermal growth fac- tor receptor (EGFR) therapy. TP53 mutations are found in 70% of cas- es. Loss of SMAD4 (which functions downstream of the TGF-β2 recep- tor) occurs in later stages and con- fers poor prognosis. In addition to these genes, recent evidence points towards more than 20 genes that are frequently mutated, including
ARID1A, SOX9, and FAM123B/ W XT [11].
Morphologically, the sequence most likely starts as an aberrant crypt focus, a small focus of mu- cosa in which the regular crypt ar- chitecture is disturbed due to mu- tations in KRAS or APC. Aberrant crypt foci with APC mutations show morphological signs of dysplasia, signifying a precancerous condi- tion. Subsequent activation of telo- merase confers unlimited lifespan to adenoma cells; TP53 mutations are involved in the progression from a low-grade adenoma to a high-grade adenoma, and other genes, includ- ing SMAD, in the progression from a non-invasive adenoma to an invasive carcinoma (Fig. 5.5.1).
Microsatellite instability pathway The microsatellite instability pathway is involved in about 15% of sporadic colorectal cancers. This genetic hall- mark is due to an incompetent DNA mismatch repair system, as a result of promoter methylation of MLH1 (in sporadic microsatellite instability- pathway colorectal cancer) or so- matic mutations of the mismatch re- pair genes (MLH1, MSH2, MSH6, or PMS2). Also, patients with a germ- line mutation of one of the mismatch repair genes in Lynch syndrome fall into this category [10,13]. These
Aberrant crypt focus Low-grade adenoma High-grade adenoma Carcinoma Metastases
- P16
- APC
+ telomerase - TP53 - SMAD4
Normal mucosa
+KRAS
Fig. 5.5.1. The chromosomal instability pathway in the development of colorectal cancer (after Vogelstein [8,9]). Morphologically, the adenoma–carcinoma sequence starts with a slight disturbance of crypt architecture and cytonuclear atypia, which is known as an aberrant crypt focus. Activating (+) KRAS mutations regularly occur in these lesions. Some of these may show features of epithelial dysplasia (as is shown in the second panel) and go on to develop into an adenoma. In this case, often inactivating (-) APC mutations are found. Low-grade adenomas are small and mostly tubular and have limited cytonuclear and architectural features of dysplasia. Once telomerase is activated (+), the lesions often show high-grade features: large, villous architecture and cytonuclear features (loss of polarity, nuclear pleomorphism, mitotic activity) of high-grade dysplasia. Progression towards invasive carcinoma is often accompanied by a TP53 mutation. Progression towards metastasis (haematogenous metastases most frequently in the liver, as shown in the right panel) is accompanied by further molecular events, such as inactivation (-) of SMAD4.
396
cancers show a relatively high fre- quency of BRAF (V600E) gene mutations. The early phases of development correspond to an adenoma–carcinoma sequence. Adenomas may be of the sessile serrated type with dysplasia, some of which progress to invasive carci- noma. Carcinomas tend to be situ- ated in the right colon, often with a mucinous or medullary histology with a host response characterized by a lymphocytic infiltrate. Cancers that arise through this pathway have a better prognosis but respond dif- ferently to standard adjuvant chemo- therapy by being less responsive to 5-fluorouracil. When oxaliplatin is added, the survival benefit for can- cers with microsatellite instability is maintained, whereas the addition of irinotecan does not seem to improve survival in such patients compared with those with microsatellite stable tumours, but this needs further con- firmation [14].
CpG island methylator pathway The CpG island methylator pathway [15,16] overlaps significantly with the microsatellite instability path- way and is characterized by meth- ylation of the promoter of a variety of genes (CpG island methylator phe- notype), including MLH1, with resul- tant microsatellite instability, which contributes to the accumulation of additional genetic abnormalities. Characteristically, an early muta- tion of the BRAF gene (typically the V600E mutation) occurs. The CpG island methylator pathway is char- acterized by morphologically rec- ognizable precursor lesions. These resemble benign hyperplastic polyps but are large (> 1 cm) and flat, occur notably in the right colon, and show irregular crypt architecture but not necessarily features of dysplasia, as in the conventional adenoma–carci- noma sequence (Fig. 5.5.2). These lesions are known as sessile serrat- ed adenomas or polyps and have an increased risk for the development of colorectal cancer, the extent of which remains to be established.
Normal mucosa
Hyperplastic polyp
Sessile serrated adenoma/polyp
SSA/P with low-grade dysplasia
SSA/P with high-grade dysplasia Carcinoma
BRAF mut
CpG island hypermethylation
MLH1 methylation microsatellite instability
Accumulation of additional oncogene and tumour suppressor gene
mutations
Fig. 5.5.2. Schematic of the CpG island methylator pathway of colorectal carcinoma. An important difference from the chromosomal instability and microsatellite instability pathways is the early occurrence of CpG island methylation together with BRAF muta- tions. Later in the pathway, gene aberrations that frequently occur in the other pathways are also found. SSA/P, sessile serrated adenoma/polyp.
Pathology and genetics Colorectal cancer exemplifies stepwise progression [8,9] as it develops initially as a benign pre- cursor lesion (adenoma), which can progress to an invasive le- sion (adenocarcinoma) with the capacity to metastasize (meta- static adenocarcinoma). The le- sion arises from an intestinal clonogenic precursor cell (crypt base stem cell [17]) through the accumulation of multiple genetic abnormalities, most notably in the Wnt pathway in the form of inacti- vating mutations of APC, activat- ing mutations of KRAS, and sub- sequent mutations of TP53 and of genes in the TGF-β pathway, which confer invasive and metastatic ca- pacity (see “Lgr5 stem cells in self- renewal and intestinal cancer”).
As outlined above, colorectal cancer comprises many different subtypes, which differ in morphol- ogy, genetic background, associ- ated conditions, molecular profile, clinical behaviour, and response to therapy. Categorization of colorec- tal cancer as sporadic, associated with inflammatory bowel disease, or having familial tendency and associated with familial colorec- tal cancer syndromes is clinically relevant.
Sporadic colorectal cancer This is the prototypical cancer type in the colorectum. Histologically, these are adenocarcinomas that as a rule develop from a benign adenomatous polyp, which can be tubular, villous, or tubulovillous in architecture. Only a limited proportion (estimated at 10%) of benign adenomas progress to car- cinoma; large adenomas with villous architecture have a high risk of pro- gression [1]. Adenomas are treated by endoscopic polypectomy.
The primary modalities of spread of adenocarcinomas of the colo- rectum are invasion into the bow- el wall and metastasis into the re- gional lymph nodes. In later stages, liver and lung metastases occur. Colorectal cancer is typically treated by surgery (hemicolectomy), with (adjuvant) chemotherapy depending on the disease stage; rectal cancer is often treated with a neoadjuvant chemo/radiotherapy protocol, fol- lowed by surgery. Stage determines prognosis as well as the therapeutic approach (Table 5.5.1), notably as re- gards the need for adjuvant therapy. For advanced stages, targeted therapy is available, targeting EGFR through monoclonal antibodies (ce- tuximab, panitumumab), but only in KRAS wild-type tumours because KRAS mutation renders the tumour insensitive to anti-EGFR therapy [18].
Chapter 5.5 • Colorectal cancer 397
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .5
Lgr5 stem cells in self-renewal and intestinal cancer Hans Clevers
The intestinal epithelium is the most rapidly self-renewing tissue in adult mammals. Lgr5 was originally iden- tified as a Wnt target gene, tran- scribed in colon cancer cells. Two knock-in alleles revealed exclusive expression of Lgr5 in cycling, co- lumnar cells at the crypt base. Using an inducible Cre knock-in allele and the Rosa26-LacZ reporter strain, lineage tracing experiments were performed in adult mice. The Lgr5- positive crypt base columnar cells generated all epithelial lineages throughout life, implying that these cells represent the stem cells of the small intestine and colon. Similar observations were made in hair fol- licles and stomach epithelium.
Single-sorted Lgr5-positive stem cells can initiate ever-expanding crypt–villus organoids in three-di- mensional culture. Tracing experi- ments indicate that the Lgr5-positive stem cell hierarchy is maintained in these organoids. The data suggest that intestinal crypt–villus units are self-organizing structures, which can be built from a single stem cell in the absence of a non-epithelial
cellular niche. The same technol- ogy has now been developed for the Lgr5-positive stomach stem cells.
Intestinal cancer is initiated by Wnt pathway-activating mutations in genes such as APC. As in most can- cers, the cell of origin has remained elusive. Deletion of APC in stem cells, but not in other crypt cells, results in progressively growing neoplasia, identifying the stem cell as the cell of origin of adenomas. Moreover, a stem cell/progenitor cell hierarchy is maintained in early stem cell-derived adenomas, lend- ing support to the “cancer stem cell” concept. Fate mapping of individual crypt stem cells using a multicolour Cre reporter revealed that, as a population, Lgr5 stem cells have a lifelong persistence, yet crypts drift towards clonality within 1–6 months. Lgr5 cell divisions occur symmetri- cally. The cellular dynamics are consistent with a model in which the resident stem cells double their numbers every day and stochasti- cally adopt stem or telomerase ac- tivator fates after cell division. Lgr5 stem cells are interspersed between
terminally differentiated Paneth cells that are known to produce bacteri- cidal products. Paneth cells are CD24-positive and express EGF, TGF-α, Wnt3, and the Notch ligand Dll4, all essential signals for stem cell maintenance in culture. Co- culturing of sorted stem cells with Paneth cells dramatically improves organoid formation. This Paneth cell requirement can be substituted by a pulse of exogenous Wnt. Genetic removal of Paneth cells in vivo re- sults in the concomitant loss of Lgr5 stem cells. In colon crypts, CD24- positive cells residing between Lgr5 stem cells may represent the Paneth cell equivalents. The data indicate that Lgr5 stem cells compete for essential niche signals provided by a specialized daughter cell, the Paneth cell.
References Schepers AG et al. (2012). Science, 337: 730–735. http://dx.doi.org/10.1126/science. 1224676 PMID:22855427
Huch M et al. (2013). Nature, 494:247–250. http://dx.doi.org/10.1038/nature11826 PMID: 23354049
Inflammatory bowel disease The generic term inflammatory bowel disease refers to two specific entities: Crohn disease and ulcerative colitis. Ulcerative colitis is restricted to the colon and/or rectum and is charac- terized by a chronic inflammatory reaction restricted to the mucosa and submucosa. Crohn disease is char- acterized by transmural inflammation and granulomas and can affect any part of the gastrointestinal tract but most commonly affects the terminal ileum. Chronic inflammatory dam- age to the colon and rectum in both types of inflammatory disease con- fers an increased risk of malignancy [19]. Some epidemiological evidence exists for an increased incidence of colorectal cancer in chronic colorectal
schistosomiasis [20]. In the connec- tion between inflammation and car- cinogenesis, the transcription factor NF-κB is a key mediator, inducing expression of the pro-inflammatory mediators COX-2 and TNF-α, both of which play a role in colorectal car- cinogenesis (see “Tumour-elicited inflammation and malignant progres- sion in colorectal cancer”). Genome abnormalities in colorectal cancer in the context of inflammatory bowel disease are similar to those in spo- radic colorectal cancer, but TP53 mutations often occur early. APC mutations are less frequent.
Colorectal cancer associated with inflammatory bowel disease de- velops through a dysplasia–carcino- ma sequence (Fig. 5.5.3). Through
endoscopic surveillance, dysplasia can be detected early on and colec- tomy can then prevent the develop- ment of an invasive carcinoma.
Familial colorectal cancer syndromes Colorectal cancer clusters in fami- lies in about 20% of cases, but a familial colorectal cancer syndrome can be pinpointed in less than half of these. When one or more relatives have been diagnosed with this ma- lignancy, the risk of such cancer in- creases 2– 6-fold with an increased number of affected family members and the age at diagnosis of affected relatives. Although a variety of chro- mosomal loci and, more recently, single-nucleotide polymorphisms
398
Table 5.5.1. Tumour–node–metastasis (TNM) classifications of carcinoma of the colon and rectum
T – Primary tumour
TX: Primary tumour cannot be assessed
T0: No evidence of primary tumour
Tis: Carcinoma in situ: intraepithelial or invasion of lamina propria
T1: Tumour invades submucosa
T2: Tumour invades muscularis propria
T3: Tumour invades subserosa or into non-peritonealized pericolic or perirectal tissues
T4: Tumour perforates visceral peritoneum and/or directly invades other organs or structures
T4a: Tumour perforates visceral peritoneum
T4b: Tumour directly invades other organs or structures
N – Regional lymph nodes
NX: Regional lymph nodes cannot be assessed
N0: No regional lymph-node metastasis
N1: Metastasis in 1–3 regional lymph nodes
N1a: Metastasis in 1 regional lymph node
N1b: Metastasis in 2 or 3 regional lymph nodes
N1c: Tumour deposit(s), i.e. satellites, in the subserosa, or in non-peritonealized pericolic or perirectal soft tissue without regional lymph-node metastasis
N2: Metastasis in 4 or more regional lymph nodes
N2a: Metastasis in 4–6 regional lymph nodes
N2b: Metastasis in 7 or more regional lymph nodes
M – Distant metastasis
M0: No distant metastasis
M1: Distant metastasis
M1a: Metastasis confined to one organ
M1b: Metastases in more than one organ or the peritoneum
Stage grouping
Stage T N M
Stage 0 Tis N0 M0
Stage I T1, T2 N0 M0
Stage II T3, T4 N0 M0
Stage IIA T3 N0 M0
Stage IIB T4a N0 M0
Stage IIC T4b N0 M0
Stage III Any T N1, N2 M0
Stage IIIA T1, T2 T1
N1 N2a
M0 M0
Stage IIIB T3, T4a T2, T3 T1, T2
N1 N2a N2b
M0 M0 M0
Stage IIIC T4a T3, T4a T4b
N2a N2b N1, N2
M0 M0 M0
Stage IVA Any T Any N M1a
Stage IVB Any T Any N M1b
have been identified (Table 5.5.2) that are associated with increased colorectal cancer risk, the less-pen- etrant genes and/or gene–environ- ment interactions responsible for such familial disease clusters have yet to be elucidated [12,13].
Familial adenomatous polyposis syndrome accounts for about 1% of all colorectal cancers and is char- acterized by the development early in life of hundreds to thousands of adenomatous polyps in the colon (Fig. 5.5.4). Mucosal neoplasms higher up in the tubal gut, notably in the stomach, also occur. The syndrome is caused by mutations in the APC gene. The APC protein in the Wnt signalling pathway nor- mally binds to β-catenin (in a nor- mal epithelial cell, integrated in the E-cadherin–catenin cell adhesion complex), triggering its degradation in the proteasome [13]. APC loss results in accumulation of β-catenin in the nucleus, where it acts as a transcription factor, promoting tran- scription of proliferation-stimulating genes, such as MYC and cyclin D1. Distinct genotype–phenotype corre- lations exist [12,13]; some mutations (in codons 1250–1464) are associ- ated with severe polyposis (> 1000 adenomas), while others (before co- don 157, after codon 1595, and in the alternatively spliced region of exon 9) are associated with attenuated adenomatous polyposis coli (usu- ally 10–100 adenomas and a 70% lifetime risk of colorectal cancer).
In Gardner syndrome, numerous gastrointestinal polyps occur in com- bination with other neoplasms, in- cluding osteomas, desmoid tumours, epidermoid cysts, and also dental anomalies. APC mutations respon- sible for Gardner syndrome occur in a region encoding for the β-catenin binding site of the protein.
Mismatch repair deficiency syndrome, also known as heredi- tary non-polyposis colon cancer or Lynch syndrome, is the most com- mon hereditary colorectal cancer predisposition syndrome (about 3% of all such disease), with a lifetime risk of about 80% [13]. The syn- drome is autosomal dominant and
Chapter 5.5 • Colorectal cancer 399
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .5
is caused by mutations in one of the DNA mismatch repair genes (MLH1 on chromosome 3p21, MSH2 on 2p16, MSH6 on 2p16, and PMS2 on 7p22). The mismatch repair protein complex corrects base mismatches or small insertions or deletions that occur during DNA replication. MLH1 and MSH2 mutations are the most frequent (about 80%). MSH2 defi- ciency can also occur through gene silencing by transcriptional read- through of TACSTD1, a gene direct- ly upstream of MSH2 that encodes the Ep-CAM protein. As a result of a germline deletion of the last exons of TACSTD1, its transcription is ex- tended into MSH2, with functional loss of MSH2 activity.
These cancers exhibit micro- satellite instability – increased vari- ability in length of the dinucleotide repeats that occur throughout the genome. These are not necessarily pathogenic but are significant when they occur in key genes in processes such as cell proliferation, apoptosis, or DNA repair. Mismatch repair-de- ficient cancers are characterized by point mutations (hypermutation) and,
less often, by chromosomal rear- rangements and allelic imbalances.
Mismatch repair-deficient can- cers develop through an adenoma– carcinoma sequence but not in as- sociation with polyposis, typically occur in the right colon, and are often histologically of a mucinous or medullary type, with a strik- ing lymphoid host reaction. The prognosis is better than that of mismatch repair-competent cancers, but mismatch repair-deficient can- cers respond less well to adjuvant chemotherapy. These cancers also occur in other organs, including (in decreasing order of frequency) the endometrium, ovary, duodenum, urinary tract, stomach, pancreas, biliary tree, and brain.
The MUTYH-associated polypo- sis (MAP) syndrome has been more recently identified and is estimated to account for about 2% of colorectal cancer. MAP syndrome resembles familial adenomatous polyposis in that multiple adenomatous polyps of the colorectum develop, with a very high lifetime risk of colorectal cancer. The number of adenomas
ranges from very few to hundreds, and therefore MAP resembles at- tenuated familial adenomatous polyposis. Extracolonic manifes- tations occur, including duodenal adenomas or adenocarcinomas and various extra-intestinal neoplasms. MAP is an autosomal recessive disorder caused by biallelic muta- tions in the MUTYH gene, located on chromosome 1p. The gene en- codes a protein with an important function in the DNA base excision repair pathway.
Prognostic and predictive factors Currently used criteria for stratifica- tion of colorectal cancer patients in view of eventual adjuvant chemo- therapy are primarily based on the classic tumour–node–metastasis (TNM) stage parameters tumour extension (T) and lymph-node me- tastasis (N). However, these lack precision (Table 5.5.1), and better parameters are urgently needed.
Microsatellite instable carcino- mas have a better prognosis than those that are microsatellite stable [10]. In addition, patients with micro- satellite instable carcinomas might not benefit from 5-fluorouracil-based chemotherapy regimens, but this re- mains controversial. For microsatel- lite stable cancers, BRAF V600E mutations confer poor prognosis, notably in terms of survival after relapse. KRAS mutation status has no prognostic significance for over- all survival of patients with stage II/ III disease [21]. Recently published tests such as Oncotype DX (based on multiplex reverse transcription polymerase chain reaction) and ColoPrint (microarray-based) might provide additional prognostic value but have not been independently validated [22]. New molecular approaches for prognostic classi- fication are being developed [11, 23–25] but need to be validated in a clinical setting.
KRAS mutation testing has be- come the mainstay for advanced colorectal cancer because KRAS- mutated carcinomas do not respond
Chronic inflammation
Low-grade dysplasia
High-grade dysplasia
Adenocarcinoma
10 years?
Promoter hypermethylation microsatellite instability -TP53
-APC +KRAS
-P27 -P16
+ gain of function mutation - loss of function (through LOH, mutation,
or promoter methylation)
Fig. 5.5.3. The dysplasia–carcinoma sequence in colorectal carcinogenesis in inflam- matory bowel disease. Chronic inflammation with the continuous liberation of oxygen radicals and inflammatory cytokines creates an environment that is both mutagenic and growth-stimulating. TP53 mutations occur early in this pathway, associated with the development of low-grade dysplasia. In addition, hypermethylation of CpG islands leads to promoter silencing of P16 and P27 among others. Accumulation of such events drives progression towards high-grade dysplasia. Additional events, such as KRAS mutations and activation of the Wnt pathway, for example through APC mutation, are involved in the progression towards invasive carcinoma. LOH, loss of heterozygosity.
400
to anti-EGFR therapy (cetuximab, panitumumab). This is due to the fact that KRAS is downstream of EGFR in the MAPK pathway, which renders approaches targeting the upstream EGFR ineffective. Only about 40% of KRAS wild-type colorectal cancers respond to an- ti-EGFR treatment, implying that other genes are involved, including PIK3CA, BRAF, NRAS, pTEN, and others unidentified as yet [21].
Prospects Screening for colorectal cancer is done through the faecal occult blood test, which has a reason- able sensitivity but low specific- ity. There is now evidence that
screening with flexible sigmoid- oscopy can substantially reduce colorectal cancer incidence and mortality. Colonoscopy is increas- ingly advocated, as colonoscopic polypectomy would, in principle, enable eradication of the disease. Molecular tests for disease-specific DNA abnormalities in faeces or in the blood are currently being de- veloped [26].
The heterogeneity of colo- rectal cancer continues to chal- lenge clinicians and pathologists. Why cancers of a similar stage that are histologically indistinguishable behave differently in terms of recur- rence and response to chemother- apy has long been a mystery. The
inclusion of molecular parameters in colorectal cancer classification improves on currently used classi- fication systems [27]. Molecular an- notation of large series of colorectal cancers with detailed follow-up data might now lead to new insights into molecular mechanisms that drive this heterogeneity. The Cancer Genome Atlas data [11], along with even more recent approaches to- wards molecular classification of colorectal cancer [23,24], will in the near future allow the redefinition of colorectal cancer subtypes, with im- pacts on clinical management. The most recent data point towards at least five molecularly defined sub- types [24] – which partly overlap
Table 5.5.2. Genetic syndromes associated with a possible risk of colorectal carcinoma
Syndrome Gene (chromosome) MIM number
Autosomal dominant inheritable colorectal carcinoma
No or few adenomatous polyps
Lynch syndromea,b MLH1 (3p21-p23) MSH2 (2p21) MSH6 (2p21) PMS2 (7p22)
120435
Adenomatous polyps
Familial adenomatous polyposis (FAP)a and attenuated FAP APC (5q21-q22) 175100
Hamartomatous/mixed/hyperplastic polyps
Peutz–Jeghers syndrome LKB1/STK11 (19p13.3) 175200
Juvenile polyposis syndrome SMAD4 (18q21.1) BMPR1A (10q22.3)
174900
Hereditary haemorrhagic telangiectasia syndromec ENG (9q33-q34.1) ACVRL1 (12q11-q14)
187300
Hyperplastic polyposis syndromec MUTYH (1p34.1; autosomal recessive) MBD4 (3q21.3)
Unassigned
Hereditary mixed polyposis syndromec CRAC1 (15q13-q21) 601228
PTEN hamartoma tumour syndrome (Cowden syndrome; Bannayan–Ruvalcaba–Riley syndrome)c
PTEN (10q23) 158350/153480
Birt–Hogg–Dubé syndromec FLCN (17p11.2) 135150
Autosomal recessive inheritable colorectal carcinoma
Adenomatous, serrated adenomas and hyperplastic polyps
MUTYH-associated polyposisb MUTYH (1p34.1) 608456
MIM, Mendelian Inheritance in Man (www.omim.org). a Turcot syndrome is a variant of Lynch syndrome, or familial adenomatous polyposis with brain tumours. b Muir–Torre syndrome is a variant of Lynch syndrome, or MUTYH-associated polyposis with sebaceous gland tumours. c Risk of colorectal carcinoma is not clear.
Chapter 5.5 • Colorectal cancer 401
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .5
Tumour-elicited inflammation and malignant progression in colorectal cancer Michael Karin
with classifications based on well- established classifiers (TNM, micro- satellite instability) but do add an additional layer of complexity – that potentially will allow more refined subclassification. In addition, new therapeutic targets and prognos- tic and predictive biomarkers will emerge.
Fig. 5.5.4. (A) Colon mucosa of a 24-year-old man with familial adenomatous polyp- osis who underwent total colectomy. (B) Mucosa with multiple adenomatous polyps (haematoxylin–eosin stain).
A B
Whereas 2% of colorectal cancer arises in the context of pre-existing inflammatory bowel disease, espe- cially ulcerative colitis, and is known as colitis-associated cancer, most colorectal cancer, including familial and sporadic cases, is found in indi- viduals who do not show any signs of inflammatory bowel disease. Yet expression profiling has revealed the same inflammatory gene signature, which depends on activation of NF- κB and STAT3, in both colitis-asso- ciated cancer and colorectal cancer, findings that pose a question about the origin of the colorectal cancer- elicited inflammatory response.
Early experiments established that activation of NF-κB, which leads to production of the pro-inflamma- tory cytokine interleukin-6 (IL-6), a potent activator of STAT3, plays a critical role in the development of colitis-associated cancer. More recently, the origin and role of in- flammation in the development and progression of sporadic malignancy, most of which is driven by loss of the tumour suppressor APC and activa- tion of the β-catenin signalling path- way, has been investigated. Studies focused on IL-23, a heterodimeric cytokine composed of a unique p19 subunit and a p40 subunit, which it shares with IL-12. IL-23 expression is
strongly elevated in colorectal can- cer relative to adjacent non-tumour tissue, and these findings have been extended to the CPC-Apc mouse model of colorectal tumorigenesis. In CPC-Apc mice, the major source of IL-23 expression in colorectal ad- enomas is tumour-associated mac- rophages. Importantly, ablation of IL-23 p19, either in all cells or only in bone marrow-derived cells, at- tenuates the development and slows down the progression of colorectal cancer in CPC-Apc mice, and simi- lar results were observed upon ab- lation of IL-23 receptor (IL-23R). As IL-23R is not expressed on adenoma epithelial cells, IL-23 must exert its pro-tumorigenic effect via an indi- rect mechanism. Indeed, IL-23 sig- nalling promotes the polarization of IL-17-producing T cells (Th17) and the production of IL-6, both of which contribute to the development and progression of sporadic colorectal tumours in mice.
Importantly, molecular epidemio- logical studies have revealed an IL- 23–Th17 signature that is upregulat- ed in about 10% of human colorectal cancer patients, and have shown that the presence of this signature in stage I/II disease is associated with very poor prognosis and a marked decrease in disease-free survival.
The mechanism responsible for the specific induction of IL-23 in tumour- associated macrophages depends on Toll-like receptor/MyD88 signal- ling, which appears to be activated in response to components of the colonic microflora that permeate the adenomas. Eubacterial 16S RNA has been detected in both murine and human colon adenomas, as well as increased penetrance of bacterial endotoxin into adenomas relative to surrounding non-cancerous tissue.
The development of colorectal adenomas in both mice and humans is associated with loss of protective mucins and junctional adhesion pro- teins, and this is likely to be the pri- mary mechanism that accounts for the selective entry of microbial prod- ucts into the tumour and induction of the tumour-promoting IL-23–Th17 signature. Future studies should fo- cus on the therapeutic value of anti- IL-23 or anti-IL-17 interventions and the genetic or environmental causes of the large variation in the magni- tude of IL-17 production among hu- man colorectal cancer patients.
Reference Grivennikov SI et al. (2012). Nature, 491:254– 258. http://dx.doi.org/10.1038/nature11465 PMID:23034650
402
References 1. Hamilton S, Bosman F, Boffetta P et al.
(2010). Carcinoma of the colon and the rec- tum. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, eds. WHO Classification of Tumours of the Digestive System, 4th ed. Lyon: IARC, pp. 134–146.
2. World Cancer Research Fund/American Institute for Cancer Research (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington DC: American Institute for Cancer Research.
3. Cross AJ, Ferrucci LM, Risch A et al. (2010). A large prospective study of meat consumption and colorectal can- cer risk: an investigation of potential mechanisms underlying this association. Cancer Res, 70:2406–2414. http://dx.doi. org/10.1158/0008 - 5472.CAN - 09 -3929 PMID:20215514
4. Devkota S, Wang Y, Musch MW et al. (2012). Dietary-fat-induced taurocholic acid promotes pathobiont expansion and colitis in Il10-/- mice. Nature, 487:104–108. ht tp://dx.doi.org/10.1038/nature11225 PMID:22722865
5. Vazzana N, Riondino S, Toto V et al. (2012). Obesity-driven inflammation and colorectal cancer. Curr Med Chem, 19:5837–5853. http://dx.doi.org/10.2174/ 092986712804143349 PMID:23033947
6. Guffey CR, Fan D, Singh UP, Murphy EA (2013). Linking obesity to colorectal cancer: recent insights into plausible bio- logical mechanisms. Curr Opin Clin Nutr Metab Care, 16:595–600. http://dx.doi. org/10.1097/MCO.0b013e328362d10b PMID:23743611
7. Thun MJ, Jacobs EJ, Patrono C (2012). The role of aspirin in cancer prevention. Nat Rev Clin Oncol, 9:259–267. http:// dx.doi.org/10.1038/nrclinonc.2011.199 PMID:22473097
8. Fearon ER, Vogelstein B (1990). A genetic model for colorectal tumorigenesis. Cell, 61:759–767. http://dx.doi.org/10.1016/0092- 8674(90)90186-I PMID:2188735
9. Kinzler KW, Vogelstein B (1996). Lessons from hereditary colorectal cancer. Cell, 87:159–170. http://dx.doi.org/10.1016/S00 92-8674(00)81333-1 PMID:8861899
10. Boland CR, Goel A (2010). Microsatellite insta- bility in colorectal cancer. Gastroenterology, 138:2073–2087. http://dx.doi.org/10.1053/j. gastro.2009.12.064 PMID:20420947
11. Muzny DM, Bainbridge MN, Chang K et al.; Cancer Genome Atlas Network (2012). Comprehensive molecular char- acterization of human colon and rectal cancer. Nature, 487:330–337. http://dx.doi. org/10.1038/nature11252 PMID:22810696
12. Fearon ER (2011). Molecular genetics of colo- rectal cancer. Annu Rev Pathol, 6:479–507. http://dx.doi.org/10.1146/annurev-pathol- 011110-130235 PMID:21090969
13. Gala M, Chung DC (2011). Hereditary colon cancer syndromes. Semin Oncol, 38:490– 499. http://dx.doi.org/10.1053/j. seminoncol.2011.05.003 PMID:21810508
14. Sinicrope FA, Sargent DJ (2012). Molecular pathways: microsatellite instability in colo- rectal cancer: prognostic, predictive, and therapeutic implications. Clin Cancer Res, 18:1506–1512. http://dx.doi.org/10.1158/1078- 0432.CCR-11-1469 PMID:22302899
15. Liang JJ, Alrawi S, Tan D (2008). Nomenclature, molecular genetics and clinical significance of the precursor lesions in the serrated polyp pathway of colorectal carcinoma. Int J Clin Exp Pathol, 1:317–324. PMID:18787610
16. Snover DC (2011). Update on the serrated pathway to colorectal carcinoma. Hum Pathol, 42:1–10. http://dx.doi.org/10.1016/j. humpath.2010.06.002 PMID:20869746
17. Vries RG, Huch M, Clevers H (2010). Stem cells and cancer of the stomach and intestine. Mol Oncol, 4:373–384. http://dx.doi.org/10.1016/j. molonc.2010.05.001 PMID:20598659
18. Bohanes P, LaBonte MJ, Winder T, Lenz HJ (2011). Predictive molecular classi- fiers in colorectal cancer. Semin Oncol, 38:576–587. http://dx.doi.org/10.1053/j. seminoncol.2011.05.012 PMID:21810517
19. Xie J, Itzkowitz SH (2008). Cancer in inflammatory bowel disease. World J Gastroenterol, 14:378–389. http://dx.doi. org/10.3748/wjg.14.378 PMID:18200660
20. Qiu D-C, Hubbard AE, Zhong B et al. (2005). A matched, case-control study of the as- sociation between Schistosoma japonicum and liver and colon cancers, in rural China. Ann Trop Med Parasitol, 99:47–52. http:// dx.doi.org/10.1179/136485905X19883 PMID:15701255
21. Roth AD, Tejpar S, Delorenzi M et al. (2010). Prognostic role of KRAS and BRAF in stage II and III resected colon cancer: results of the translational study on the PETACC-3, EORTC 40993, SAKK 60- 00 trial. J Clin Oncol, 28:466–474. http:// dx.doi.org/10.1200/JCO.2009.23.3452 PMID:20008640
22. Kelley RK, Venook AP (2011). Prognostic and predictive markers in stage II colon cancer: is there a role for gene expression profiling? Clin Colorectal Cancer, 10:73–80. http://dx.doi.org/10.1016/j.clcc.2011.03.001 PMID:21859557
23. Sadanandam A, Lyssiotis CA, Homicsko K et al. (2013). A colorectal cancer classifica- tion system that associates cellular pheno- type and responses to therapy. Nat Med, 19:619–625. http://dx.doi.org/10.1038/nm. 3175 PMID:23584089
24. De Sousa E Melo F, Wang X, Jansen M et al. (2013). Poor-prognosis colon cancer is defined by a molecularly distinct subtype and develops from serrated precursor le- sions. Nat Med, 19:614–618. http://dx.doi. org/10.1038/nm.3174 PMID:23584090
25. Merlos-Suárez A, Barriga FM, Jung P et al. (2011). The intestinal stem cell signa- ture identifies colorectal cancer stem cells and predicts disease relapse. Cell Stem Cell, 8:511–524. http://dx.doi.org/10.1016/j. stem.2011.02.020 PMID:21419747
26. Bosch LJ, Carvalho B, Fijneman RJ et al. (2011). Molecular tests for colorectal cancer screening. Clin Colorectal Cancer, 10:8–23. http://dx.doi.org/10.3816/CCC.2011.n.002 PMID:21609931
27. Roth AD, Delorenzi M, Tejpar S et al. (2012). Integrated analysis of molecular and clinical prognostic factors in stage II/III colon cancer. J Natl Cancer Inst, 104:1635–1646. http://dx.doi.org/10.1093/ jnci/djs427 PMID:23104212
Websites American Cancer Society Colorectal Cancer home page: http://www.cancer.org/cancer/ colonandrectumcancer/index
MedlinePlus Colorectal Cancer home page: http://www.nlm.nih.gov/medlineplus/colorec talcancer.html
National Cancer Institute Colon and Rectal Cancer home page: http://www.cancer.gov/ cancertopics/types/colon-and-rectal
SEER Stat Fact Sheets: Colon and Rectum: http://seer.cancer.gov/statfacts/html/colorect. html
Chapter 5.6 • Liver cancer 403
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .6
Summary • The highest incidence rates of
hepatocellular carcinoma are in regions with the highest preva- lence of predisposing condi- tions. In Asia and Africa, chronic hepatitis B virus and hepatitis C virus infections are the major causes, and in China and sub- Saharan Africa, exposures to aflatoxins further increase risk.
• Incidence rates of hepatocellu- lar carcinoma in men are more than twice those in women, re- flecting greater rates of predis- posing chronic liver disease and possibly hormonal influences. In North America and parts of Europe, rates are increasing due to chronic hepatitis C virus infection. Epidemic metabolic syndrome and non-alcoholic fatty liver disease are expected to lead to further increases.
• Universal infant vaccination against hepatitis B virus is rec- ommended by WHO, while anti- viral therapy for chronic hepatitis B and C virus infections will lead to reduced incidence of hepa- tocellular carcinoma in relevant communities. Effective interven- tions to reduce aflatoxin expo- sure are also available.
5.6 5 ORGAN SITE
Liver cancer Neil D. Theise Chien-Jen Chen (reviewer)
Michael C. Kew (reviewer)
• For individuals with chronic liver disease, routine screening for hepatocellular carcinoma is rec- ommended; individuals without chronic liver disease, or living in areas where such screening is not performed, are likely to present with symptoms of ad- vanced disease.
• The Wnt/β-catenin pathway is commonly disrupted in hepato- cellular carcinoma. Mutation of a particular codon in TP53 may be identified with environmental exposure to aflatoxin.
• Cholangiocarcinoma is un- common except in South-East Asia where infections with liver flukes, such as Clonorchis and Opisthorchis species, are en- demic. Cholangiocarcinoma re- mains difficult to treat and is usu- ally lethal.
The most common tumour type occurring in the liver is metastatic disease [1]. Of primary hepatic malignancies, hepatocellular car- cinoma, composed of malignant hepatocytes, represents about 80% of tumours. Cholangiocarcinoma, a gland-forming adenocarcinoma de- riving from the biliary tree, is the next most common [2]. Other, rare prima- ry tumours include mucinous cystic
neoplasm and intraductal papillary biliary neoplasm, the classifications of which have recently been aligned with pancreatic ductal malignancies, tumours of mixed hepatocellular and biliary phenotype, hepatoblastoma in children, and mesenchymal tu- mours such as angiosarcoma [1].
Etiology Hepatocellular carcinoma is gener- ally more common in men than in women [1,3]. Men have higher rates of chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infection and chronic alcohol consumption. Sex hormones may also modulate hepatocarcinogenesis [4]. There are also differences in incidence ac- cording to race and ethnic alignment [3]. For example, within the USA, Native Americans and Americans of Asian and Pacific Island descent have higher rates of hepatocellular carcinoma. Intermediate rates are seen in Hispanic/Latino individu- als, African Americans, and Native Americans. Caucasians have the lowest rates. These variations are probably indicative of different rates of predisposing conditions and exposures; genetic and epige- netic differences are also probably significant.
The global regions of greatest hepatocellular carcinoma preva- lence correspond to those with
404
Epidemiology
Liver cancer • Liver cancer represents 6%
and 9% of the global cancer in- cidence and mortality burden, respectively. With an estimated 746 000 deaths in 2012, liver cancer is the second most com- mon cause of death from cancer worldwide.
• Liver cancer is the fifth most common cancer in men (554 000 new cases, 8% of the total) and the ninth most common in wom- en (228 000 cases, 3% of the to- tal). Almost three quarters of the new cases occur in areas with low and medium human devel- opment; more than half of the global incidence and mortality is in China.
• Given the high fatality of liver cancer (overall mortality-to- incidence ratio, 0.95), the geo- graphical patterns and trends for mortality are very similar to those observed for incidence.
• By far the highest age-stan- dardized incidence rate is seen in Mongolia. There are elevated incidence rates in East and South-East Asia, in Africa, and in Melanesia. Incidence rates tend to be lower in most highly developed regions.
Map 5.6.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for liver cancer in men, 2012.
Map 5.6.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for liver cancer in women, 2012.
high rates of predisposing condi- tions (Maps 5.6.1, 5.6.2; Table 5.6.1) [1,3]. Examples include South and East Asia, where HBV infection is endemic; Egypt, with its high rates of HCV infection from inadequate sterilization during mass vaccina- tion campaigns; and sub-Saharan Africa, South-East Asia, and China, where marked exposure to aflatoxin
occurs [3]. In China, the Republic of Korea, Japan, Zimbabwe, and Egypt, incidence rates are more than 20 per 100 000 people [3]. The lowest rates of hepatocellular carcinoma are those reported for North America, South America, and northern Europe, which typically have incidence rates of less than 10 per 100 000 people [3].
Changing incidence rates of hepatocellular carcinoma reflect the changing distribution of predis- posing conditions [1,3]. Thus, coun- tries in Europe, North and South America, and Oceania, where HCV- infected populations have expand- ed, have experienced increasing hepatocellular carcinoma incidence [3]. Non-alcoholic fatty liver disease
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.6 • Liver cancer 405
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .6
Chart 5.6.1. Estimated global number of new cases and deaths with proportions by major world regions, for liver cancer in both sexes combined, 2012.
Chart 5.6.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for liver cancer in men, circa 1975–2012.
Chart 5.6.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for liver cancer in women, circa 1975–2012.
406
is emerging as a relatively newly understood predisposing condition, and hepatocellular carcinoma inci- dence rates are likely to increase in affected regions [5]. Increasing survival of individuals with chronic liver disease also may play a role in these upward trends [3].
In contrast, in East Asia incidence rates of hepatocellular carcinoma are declining [1,3]. In Japan, as the co- hort of individuals infected by HCV in the first half of the 20th century shrinks, incidence of hepatocellu- lar carcinoma is also declining. In China and Singapore, where HBV infection remains a major predispos- ing condition, diminished aflatoxin exposure is leading to a decline in HBV-associated disease. Vaccination and treatment for HBV infection and new, more successful treatments for HCV infection will almost certainly lead to diminished rates of hepato- cellular carcinoma in places where these are significant risk factors. Indeed, HBV vaccination has been well documented to prevent 70% of hepatocellular carcinoma incidence among vaccinated birth cohorts in a 20-year follow-up study in Taiwan, China [6].
Sporadic cholangiocarcinomas and those associated with HCV, HBV, and primary sclerosing cholangitis are uncommon to rare [1]. However, in regions with endemic liver fluke infestations, such as north-eastern Thailand, incidence is as high as 88 per 100 000 men and 35 per 100 000 women [2].
Hepatocellular carcinoma is most commonly associated with chronic liver disease [1,3,5,7]. Although this
malignancy usually occurs in the setting of cirrhosis, cirrhosis per se is not a premalignant lesion [8]. Rather, progression to cirrhosis and hepatocarcinogenesis take place in parallel over years to decades [1,9]. Cirrhosis is not, moreover, required for development of hepatocellular carcinoma [1,10]. The most impor- tant etiological factors for hepatocar- cinogenesis remain viral infections, usually involving HBV or HCV, and toxic injury, typically initiated by ingestion of aflatoxin or consump- tion of alcohol [1,3]. Where HBV and HCV are endemic, there is a very high incidence of this malignancy. In such chronic infection, viral load is the major driver of the progres- sion of viral hepatitis-related liver disease [10,11]. It should be noted that only about one quarter of pa- tients with chronic HBV and HCV infection will develop hepatocellular carcinoma [2]. Risk calculators that incorporate age, sex, family history of hepatocellular carcinoma, alco- hol consumption, serum alanine aminotransferase (ALT) level, and viral factors including HBV e anti- gen (HBeAg) serostatus, viral load, and genotype have been developed to predict hepatocellular carcinoma caused by chronic HBV infection [12]. Co-infection with both hepatitis viruses further increases risk [13]. HIV infection also appears to be a risk factor for hepatocellular carci- noma [14].
Aflatoxin (specifically, aflatoxin B1), a mycotoxin produced by moulds of the Aspergillus species, contami- nates staple food crops in Africa and Asia [15]. Aflatoxin metabolites
are present in the urine of affected individuals, as are aflatoxin–albu- min adducts in serum [1,15]. These indicators may assist in identifying populations at risk and confirm the important influence of aflatoxin on the development of hepatocellular carcinoma. Risk attributable to afla- toxin acts synergistically with HBV infection and possibly with HCV in- fection [15]. Alcohol is another toxin that may also act synergistically with other causes, particularly HBV and HCV, and possibly cigarette smok- ing [16].
Aflatoxin-specific mutations pro- vide a singular relationship between the impact of a carcinogenic agent, the establishment of a biomarker, and inactivation of the relevant tumour suppressor gene [14]. Aflatoxin B1 is metabolized by members of the cy- tochrome P450 CYP family, specifi- cally including CYP3A4 and CYP3A5. These isoenzymes metabolize afla- toxin B1 to its 8,9-exo-epoxide, which may react with DNA to form an afla- toxin–N7-guanine adduct and with protein to generate aflatoxin–albumin adducts. Aflatoxin–albumin adducts may serve as biomarkers for aflatoxin exposure. In hepatocellular carcino- mas from areas where exposure to aflatoxin is high, aflatoxin–DNA ad- ducts result in a high prevalence of thymine substitutions at the third guanine in AGG sequences, thereby replacing serine for arginine in ex- pressed proteins. This mutation is specifically recorded at codon 249 of p53 (TP53) and has been extensively monitored in hepatocellular carci- noma patients and relevant controls
Table 5.6.1. Risk factors for hepatocellular carcinoma and cholangiocarcinoma
Type of condition Hepatocellular carcinoma Cholangiocarcinoma
Epidemic Chronic HBV infection Chronic HCV infection Aflatoxin exposure
Liver flukes
Common Chronic alcohol consumption Primary sclerosing cholangitis Non-alcoholic fatty liver disease Recurrent pyogenic cholangitis
Rare Hereditary haemochromatosis α-1-Antitrypsin deficiency
Chronic HCV infection Chronic HBV infection Non-alcoholic fatty liver disease
HBV, hepatitis B virus; HCV, hepatitis C virus.
Chapter 5.6 • Liver cancer 407
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .6
from many regions, and particularly in studies from China [17].
Metabolic diseases such as hereditary haemochromatosis and α-1-antitrypsin deficiency markedly increase the risk of hepatocellular carcinoma; Wilson disease probably does so with much less frequency [1]. Of probably greater importance is the metabolic syndrome, which is associated with obesity, diabetes, and non-alcoholic fatty liver dis- ease and will become increasingly important as the global epidemics of obesity and diabetes continue to worsen [5]. Moreover, this syndrome is likely to act synergistically for car- cinogenesis in people with chronic viral hepatitis [18].
Cholangiocarcinoma is usually sporadic beyond regions charac- terized by endemic infections with foodborne liver flukes, such as those of the Clonorchis and Opisthorchis species [1,2]. The feeding, migrating flukes mechanically injure the biliary tree and incite abundant chronic in- flammation that contributes to step- wise progression, via increasingly atypical biliary intraepithelial neo- plasia, to invasive cholangiocarci- noma (Fig. 5.6.1) [1,2]. Inflammatory cholangiopathies, in particular pri- mary sclerosing cholangitis, hepa- tolithiasis, and recurrent pyogenic cholangitis, also can lead to biliary intraepithelial neoplasia and cholan- giocarcinoma [1,19]. Increased rates of cholangiocarcinoma are also seen associated with non-alcoholic fatty
liver disease, alcoholic liver disease, chronic HBV infection, and chron- ic HCV infection, although these cholangiocarcinomas remain un- common [19].
Pathology Hepatocellular carcinoma may arise from malignant transformation of hepatocytes or of hepatobiliary stem/ progenitor cells [1]. In chronic liver disease, with or without cirrhosis, several precursor lesions have been described [20]. Most early lesions are recognized in established cirrhosis since mature hepatocellular carcino- ma usually follows the development of such late-stage disease. However, all such lesions are recognized in pre-cirrhotic stages of scarring and regeneration as well [21]. The tumour– node–metastasis (TNM) classifica- tions of hepatocellular carcinoma are shown in Table 5.6.2.
Cellular atypias known as “large cell change” and “small cell change” are related to hepatocarcinogenesis, particularly in patients with chronic viral hepatitis [21]. Large cell change consists of atypical cells with pre- served nuclear-to-cytoplasmic ratio. Small cell change consists of small, atypical hepatocytes with increased
nuclear-to-cytoplasmic ratio. These cells appear in microscopic, expan- sile clusters in the hepatic lobule.
Dysplastic nodules are usu- ally seen in cirrhosis (Fig. 5.6.3) [1,9,20,21]. These are radiologi- cally or grossly distinctive nodules and are further classified as low- grade or high-grade based on the presence of cellular or architectural atypia [9,20,21]. Other high-grade changes include “nodule-in-nodule” lesions in which an expansile sub- nodule shows cellular and architec- tural features indicative of, but not yet prominent enough for diagnosis of, malignancy [1,9,20,21].
As premalignant lesions prog- ress towards carcinoma, they initially spread around adjacent vascular and portal structures [9,20]. This can give rise to a radiologically sub- tle “vaguely nodular” pattern of early hepatocellular carcinoma, some- times lacking diagnostic features by computed tomography or magnetic resonance imaging (Fig. 5.6.4A) [9,20,21]. With stepwise progres- sion to hepatocellular carcinoma, more highly proliferative and more poorly differentiated subnodules arise, which, with increasingly rapid growth, overtake the pre-existing
Fig. 5.6.1. Biliary dysplasia related to liver fluke infestation. Low-grade biliary intraep- ithelial neoplasia (BilIN-1) associated with an intraluminal liver fluke, Opisthorchis viverrini (haematoxylin–eosin; 10×).
Fig. 5.6.2. A young woman arranges maize in a seed house in Dumka, India. Crops stored in a hot and humid environment are particularly susceptible to contamination by strains of Aspergillus, resulting in dietary exposure to aflatoxins.
408
Treating chronic hepatitis with antiviral drugs to prevent liver cancer Mark R. Thursz
During the past decade there have been major advances in the treat- ment of chronic viral hepatitis. Replication of hepatitis B virus (HBV) can be suppressed either by induc- tion of immune responses with type 1 interferon or with nucleos(t)ide analogue drugs. Interferon-based therapy succeeds in only about 30% of cases, so nucleos(t)ide analogues are the more popular treatment op- tion. Nucleos(t)ide analogues need to be administered indefinitely to maintain viral suppression, to prevent progression to cirrhosis and from
cirrhosis to decompensated cir- rhosis. Furthermore, recent studies show resolution of cirrhosis with long-term treatment. However, the impact of viral suppression on the incidence of hepatocellular carci- noma is more controversial. One randomized placebo-controlled trial demonstrated a reduced incidence of cancer in treated patients. Meta- analysis of subsequent studies com- paring successfully treated patients with historical controls or treatment failures now confirms that treatment reduces but does not eliminate the risk of hepatocellular carcinoma.
Chronic hepatitis C virus (HCV) infection can be cured using a com- bination of pegylated interferon and the nucleoside analogue ribavirin in more than 50% of cases. Combining current treatment with novel pro- tease inhibitors increases the cure rate to about 80%, and there are currently combinations of direct antivirals in trials that, used in the absence of interferon, can cure the infection in more than 95% of cases. As with HBV infection, elimination of HCV infection prevents progres- sion of the disease to cirrhosis and cancer. However, as cirrhosis is an independent risk factor for hepato-
cellular carcinoma, elimination of the infection reduces but does not remove the risk of cancer.
Treatment of viral hepatitis, particularly if it is initiated before the onset of cirrhosis, is an effec- tive method of reducing the bur- den of hepatocellular carcinoma. However, several barriers prevent the widespread deployment of treat- ment. First, chronic viral hepatitis is asymptomatic, and it is essential to screen at-risk populations to identify those who need treatment. Second, although some drugs avail- able for treatment of HBV infection are also used to treat HIV infection, The Global Fund to Fight AIDS, Tuberculosis, and Malaria will not supply these drugs for patients with HBV infection. Third, treatment of HCV infection currently requires a high level of medical training, which is not universally available. Finally, in spite of a World Health Assembly resolution in 2010, treatment pro- grammes for viral hepatitis have not been developed for low-resource settings.
Reference Thursz M, Brown A (2011). Gut, 60:1025–1026. http://dx.doi.org/10.1136/gut.2010.236521 PMID:21508419
Fig. B5.6.1. High-magnification micro- graph of ground-glass hepatocytes, as seen in a chronic hepatitis B virus infec- tion with a high viral load; liver biopsy, haematoxylin–eosin stain.
lesion and begin to push adjacent tissues aside to form a pseudocap- sule [9,20]. These are features of “progressed” hepatocellular carci- noma (Fig. 5.6.4B). Throughout these processes, there is increasing arteri- alization, which slowly overtakes the contribution of portal venous blood until it is eventually eliminated in the progressed lesions, a process ex- ploited for radiological assessment (Fig. 5.6.5A) [9,20]. In non-cirrhotic livers, hepatocellular adenomas are also potential precursors of hepato- cellular carcinoma [14].
Cholangiocarcinoma is a gland- forming adenocarcinoma usually identified in advanced stages when
the initial, premalignant lesion has been overgrown (Fig. 5.6.1) [1,2]. Extrahepatic spread often occurs through lymphatic and perineural invasion [1]. Cholangiocarcinoma may be mass-forming, have multi- focal nodules mimicking metastatic disease, or be a diffusely infiltrative process [1]. Since there is as yet no specific antigenic marker expression to immunohistochemically confirm biliary origin, confirmation of the diagnosis of cholangiocarcinoma requires clinical and radiological exclusion of metastasis from extra- hepatic sites, particularly when there is no underlying hepatic disease that predisposes to biliary neoplasia [19].
Molecular and genetic alterations No single sequence of molecular or genetic alterations contributes to the development of hepatocellular carci- noma [1,22]. The two most common early mutational events lead to acti- vation of β-catenin and inactivation of p53. The former are found in up to 40% of these tumours and are more likely to be unrelated to HBV and to demonstrate genetic instability; the latter are found in up to 60% of hepa- tocellular carcinoma and are strongly associated with aflatoxin B1 exposure. However, neither of these alterations is found in pre-malignant, low-grade or high-grade dysplastic nodules.
Chapter 5.6 • Liver cancer 409
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .6
The Wnt/β-catenin pathway is commonly disrupted in hepatocel- lular carcinoma, mostly as a result of mutations in CTNNB1 or AXIN1, epigenetic silencing of CDH1, or changes in the expression of Frizzle receptors. Mutations have also been identified in genes in the interleu- kin-6 (IL-6)/JAK/STAT pathways [1,22]. The Rb1 pathway may be disturbed by mutation, loss, or silenc- ing of Rb1 or CDKN2A, or by reduced expression of other genes, such as CDKN1A.
Recently, a role has been dem- onstrated for short interfering mi- croRNAs as modulators of these pathways, suppressing HNF4α and resulting in malignant transformation of hepatocytes [23]. This finding links chronic inflammation and hepatic IL-6 directly to malignant transforma- tion without initiating mutations. The elucidation of roles in hepatocarcino- genesis of other microRNAs or oth- er epigenetic mechanisms is likely to be fruitful in the coming years. Other pathways have shown less consistent changes in hepatocellular
carcinoma progression, including the MAPK and PI3K/AKT/mTOR signalling pathways and those mod- erated by growth factors [22].
Whole-genome sequencing of 88 matched hepatocellular carcinomas and normal pairs, 81 of which were HBV-positive, indicated β-catenin to be the most frequently mutated on- cogene (15.9%) and TP53 the most frequently mutated tumour suppres- sor gene (35.2%) [24]. The Wnt/β- catenin and JAK/STAT pathways were altered in 62.5% and 45.5% of cases, respectively, and prevalent pathways suggested a basis for cat- egorization of tumours (Fig. 5.6.6).
IL-6 is implicated in the genesis of cholangiocarcinoma; its secretion by tumour cells enhances growth through autocrine mechanisms and may regulate activity of DNA trans- methylases and hence alter expres- sion of other genes, including EGFR. EGFR expression is a prognostic factor and a risk factor for tumour recurrence. Cholangiocarcinoma displays frequent mutations in cell- cycle regulators (p53, p16) and in
Table 5.6.2. Tumour–node–metastasis (TNM) classifications of hepatocellular carcinoma
T – Primary tumour
TX: Primary tumour cannot be assessed
T0: No evidence of primary tumour
T1: Solitary tumour without vascular invasion
T2: Solitary tumour with vascular invasion or multiple tumours, none > 5 cm in greatest dimension
T3: Multiple tumours, any > 5 cm, or tumour involving a major branch of the portal or hepatic vein(s)
T3a: Multiple tumours, any > 5 cm
T3b: Tumour involving a major branch of the portal or hepatic vein(s)
T4: Tumour(s) with direct invasion of adjacent organs other than the gall bladder or with perforation of visceral peritoneum
N – Regional lymph nodes
NX: Regional lymph nodes cannot be assessed
N0: No regional lymph-node metastasis
N1: Regional lymph-node metastasis
M – Distant metastasis
M0: No distant metastasis
M1: Distant metastasis
Stage grouping
Stage T N M
Stage I T1 N0 M0
Stage II T2 N0 M0
Stage IIIA T3a N0 M0
Stage IIIB T3b N0 M0
Stage IIIC T4 N0 M0
Stage IVA Any T N1 M0
Stage IVB Any T Any N M1
Fig. 5.6.3. Dysplastic nodule in cirrho- sis. The distinctive nodule (arrow) in an explanted liver from a patient with end- stage hepatitis B stands out from the sur- rounding cirrhotic parenchyma in terms of colour and size. Dysplastic nodules, low- grade and high-grade, can be discerned by computed tomography and magnetic resonance imaging based on size, arte- rialization, and iron content. Histological examination of this lesion revealed a low- grade dysplastic nodule, devoid of cellular or architectural atypia; nonetheless, even low-grade dysplastic nodules in chronic liver disease indicate an increased risk of hepatocellular carcinoma development in the liver as a whole.
410
MAPK pathways, in particular KRAS mutations [1,22]. KRAS in the bile of patients with primary sclerosing cholangitis without cholangiocarci- noma suggests that KRAS muta- tions are probably an early event in cholangiocarcinogenesis. Altered expression of ERRB2/HER2 has been described. Also, SOCS3 inhibi- tion by promoter methylation leads to the constitutive activation of the IL-6/ STAT3 pathway. Activation of that pathway, a potent mediator of local inflammatory reactions, provides an avenue for understanding the links between chronic biliary inflammation and carcinogenesis [22].
Detection For individuals with chronic liver dis- ease, routine screening for hepato- cellular carcinoma is recommended; individuals without chronic liver dis- ease, or living in areas where such screening is not performed, are likely to be diagnosed with symptoms of advanced disease [25]. Prospective screening, where possible, takes two forms: assessment of serum markers and repeated radiological examina- tions. The most widely used serum marker is α-fetoprotein, although it is ineffective for early lesions [1,25]. Other markers have been assessed, such as serum PIVKA and des-gam- ma-carboxy prothrombin, but none have reached the level of standard of practice [1].
Radiological examination in the course of late-stage (usu- ally cirrhotic) liver disease, in con- trast, is informative when possible [9,20,25]. Ultrasonography is gener- ally insensitive for early lesions, but computed tomography and magnetic resonance imaging can often identify dysplastic nodules or early hepato- cellular carcinoma [20]. If arterializa- tion is complete, then there will be diagnostic features confirming this malignancy, including arterial en- hancement, “washout” appearance, and pseudocapsule (Fig. 5.6.5B) [9,20]. Low-grade, high-grade, and vaguely nodular hepatocellular car- cinoma, however, do not consis- tently show all these features, and
Fig. 5.6.4. Forms of early hepatocellular carcinoma (HCC). (A) The margins of a “vaguely nodular” HCC (arrows) are difficult to distinguish from the surrounding cirrhotic liver. Such lesions may or may not have completely characteristic imaging studies for defining them as HCC. Even biopsy diagnosis in such lesions may be difficult, relying on subtle changes indicative of microscopic stromal invasion. (B) A small but “progressed” HCC will have characteristic imaging features diagnostic of HCC. The pseudocapsule consists of atrophic hepatic tissue compressed by the rapidly expanding tumour.
Fig. 5.6.5. Pathological–radiological correlation of early stages of human hepatocar- cinogenesis. (A) International consensus on the classification of small nodular lesions in cirrhotic liver. The diagnosis must consider the context of the lesion, especially the presence of cirrhosis, imaging findings, and growth rate. HCC, hepatocellular carci- noma; H-DN, high-grade dysplastic nodule; hyper, hypervascular; hypo, hypovascular; iso, isovascular; IWP, International Working Party; L-DN, low-grade dysplastic nodule; MD, moderately differentiated; WD, well-differentiated. (B) Liver magnetic resonance image showing a 1.9 cm, well-circumscribed hepatocellular carcinoma in a hepatitis B virus-associated cirrhosis. The lesion is dark to background liver before contrast administration (PRE), avidly enhances immediately after contrast in the hepatic arterial phase (HAP), and displays central “washout” appearance with a peripherally enhancing pseudocapsule in the 3-minute delayed phase imaging (DEL).
Chapter 5.6 • Liver cancer 411
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .6
the current practice is to follow up such lesions with greater frequency, awaiting emergence of definitively diagnostic features or rapid increase in size [25]. Diagnosis of early le- sions therefore does not depend on biopsy [20,25].
Cholangiocarcinoma has no specific serum markers that are di- agnostic; however, in patients with conditions putting them at high risk, rising or high levels of serum mark- ers such as CA19-9 and CEA may be helpful [26]. Repeat elevation of these markers after treatment can also be used to detect disease recurrence.
Prevention Prevention of hepatic malignancies relies on removing the predisposing causes [26]. In coming decades, vac- cination for HBV in endemic regions and screening of blood products for HCV should begin to diminish the in- cidence of hepatocellular carcinoma [22,25]. Reducing aflatoxin-produc- ing mould in stored food has already resulted in decreased hepatocellular carcinoma incidence, even in HBV- infected populations [3,15]. Antiviral therapies, particularly when virus is eradicated, should also diminish incidence [25] (see “Treating chronic hepatitis with antiviral drugs to pre- vent liver cancer”). Malignancies associated with non-alcoholic fatty liver disease may prove difficult to prevent [5].
For prevention of cholangiocar- cinoma, cooking of infested fish ef- fectively eliminates the possibility of human infection by tumour-causing flukes [2]. However, such a public health intervention may conflict with long-standing cultural practices that include eating raw fish dishes in areas like Thailand. Thus, although prevention is theoretically possible, this goal remains difficult and may be unlikely to be achieved in relevant populations.
S 1 S 2
Cell proliferation
S 3
Liver metabolism
Immune response,
angiogenesis
AFP
Well/moderately differentiated
CTNNB1, JAK1 (mut)
TERT (HBV)
Poor survival Good survival
MLL4 (HBV)
TP53 (mut)
Recurrence
Poorly/moderately differentiated
20 40
P-value=0.013
60 80 100 120
0 .0
0 .2
0 .4
0 .6
0 .8
1 .0
Months
1 2 3
Month
S 3 : n = 3 2
S 1 : n = 2 0
S 2 : n = 1 8 Su rv
iv al
p ro
ba bi
lit y
A
B
Fig. 5.6.6. Molecular subclassification of hepatocellular carcinoma (HCC). (A) Schematic summary of the gene expression, genetic, and clinical profiles of the three HCC sub- classes. AFP, α-fetoprotein; HBV, hepatitis B virus integration; mut, mutation (somatic single-nucleotide variant). (B) Kaplan–Meier survival plot for the three HCC subclasses.
412
References
1. Bosman F, Carneiro F, Hruban R, Theise ND, eds (2010). WHO Classification of Tumours of the Digestive System, 4th ed. Lyon: IARC.
2. Sripa B, Kaewkes S, Sithithaworn P et al. (2007). Liver fluke induces cholan- giocarcinoma. PLoS Med, 4:e201. http:// dx.doi.org/10.1371/journal.pmed.0040201 PMID:17622191
3. McGlynn KA, London WT (2011). The global epidemiology of hepatocellular carcinoma: present and future. Clin Liver Dis, 15:223– 243, vii–x. http://dx.doi.org/10.1016/j.cld. 2011.03.006 PMID:21689610
4. Hou J, Xu J, Jiang R et al. (2012). Estrogen- sensitive PTPRO expression represses hepatocellular carcinoma progression by control of STAT3. Hepatology, 57:678– 688. http://dx.doi.org/10.1002/hep.25980 PMID:22821478
5. Baffy G, Brunt EM, Caldwell SH (2012). Hepatocellular carcinoma in non-alco- holic fatty liver disease: an emerging menace. J Hepatol, 56:1384–1391. http:// dx.doi.org/10.1016/j.jhep. 2011.10.027 PMID:22326465
6. Chang MH, You SL, Chen CJ et al.; Taiwan Hepatoma Study Group (2009). Decreased incidence of hepatocellular carcinoma in hepatitis B vaccinees: a 20-year follow- up study. J Natl Cancer Inst, 101:1348– 1355. http://dx.doi.org/10.1093/jnci/djp288 PMID:19759364
7. Lata J (2010). Chronic liver diseases as liver tumor precursors. Dig Dis, 28:596– 599. http://dx.doi.org/10.1159/000320057 PMID:21088408
8. Theise ND (1996). Cirrhosis and hepato- cellular neoplasia: more like cousins than like parent and child. Gastroenterology, 111:526–528. http://dx.doi.org/10.1053/gast. 1996.v111.agast961110526 PMID:8690221
9. Hytiroglou P, Park YN, Krinsky G, Theise ND (2007). Hepatocarcinogenesis in hu- mans: pathophysiology, radiographic detec- tion and clinical significance. Gastroenterol Clin North Am, 36:867–887. http://dx.doi. org/10.1016/j.gtc.2007.08.010
10. Chen CJ, Yang HI, Su J et al.; REVEAL-HBV Study Group (2006). Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA, 295:65–73. http://dx.doi.org/10.1001/jama. 295.1.65 PMID:16391218
11. Lee MH, Yang HI, Lu SN et al. (2010). Hepatitis C virus seromarkers and sub- sequent risk of hepatocellular carcinoma: long-term predictors from a community- based cohort study. J Clin Oncol, 28: 4587–4593. http://dx.doi.org/10.1200/JCO. 2010.29.1500 PMID:20855826
12. Yang HI, Sherman M, Su J et al. (2010). Nomograms for risk of hepatocellu- lar carcinoma in patients with chronic hepatitis B virus infection. J Clin Oncol, 28:2437–2444. http://dx.doi.org/10.1200/ JCO.2009.27.4456 PMID:20368541
13. Huang YT, Jen CL, Yang HI et al. (2011). Lifetime risk and sex difference of hepa- tocellular carcinoma among patients with chronic hepatitis B and C. J Clin Oncol, 29:3643–3650. http://dx.doi.org/10.1200/ JCO.2011.36.2335 PMID:21859997
14. Bouvard V, Baan R, Straif K et al.; WHO International Agency for Research on Cancer Monograph Working Group (2009). A review of human carcinogens – part B: biological agents. Lancet Oncol, 10:321–322. http://dx.doi.org/10.1016/S14 70-2045(09)70096-8 PMID:19350698
15. Wogan GN, Kensler TW, Groopman JD (2012). Present and future directions of translational research on aflatoxin and hepatocellular carcinoma. A review. Food Addit Contam Part A, 29:249–257. http:// dx.doi.org/10.1080/19440049.2011.563370 PMID:21623489
16. Grewal P, Viswanathen VA (2012). Liver cancer and alcohol. Clin Liver Dis, 16: 839–850. http://dx.doi.org/10.1016/j.cld. 2012.08.011 PMID:23101985
17. IARC (2012). Chemical agents and related occupations. IARC Monogr Eval Carcinog Risks Hum, 100F:1–599. PMID:23189753
18. Chen CL, Yang HI, Yang WS et al. (2008). Metabolic factors and risk of hepatocel- lular carcinoma by chronic hepatitis B/C infection: a follow-up study in Taiwan. Gastroenterology, 135:111–121. http://dx. doi.org/10.1053/j.gastro. 20 08.03.073 PMID:18505690
19. Braconi C, Patel T (2010). Cholangiocar- cinoma: new insights into disease pathogen- esis and biology. Infect Dis Clin North Am, 24:871–884, vii. http://dx.doi.org/10.1016/j. idc.2010.07.006 PMID:20937455
20. International Consensus Group for Hepatocellular Neoplasia (2009). Pathologic diagnosis of early hepatocellular carcinoma: a report of the International Consensus Group for Hepatocellular Neoplasia. Hepatology, 49:658–664. http://dx.doi.org/ 10.1002/hep.22709 PMID:19177576
21. Park YN (2011). Update on precursor and early lesions of hepatocellular carcino- mas. Arch Pathol Lab Med, 135:704–715. PMID:21631263
22. Nault JC, Zucman-Rossi J (2011). Genetics of hepatobiliary carcinogenesis. Semin Liver Dis, 31:173–187. http://dx. doi.org/10.1055/s-0031-1276646 PMID: 21538283
23. Hatziapostolou M, Polytarchou C, Aggelidou E et al. (2011). An HNF4α-miRNA inflamma- tory feedback circuit regulates hepatocel- lular oncogenesis. Cell, 147:1233–1247. http://dx.doi.org/10.1016/j.cell.2011.10.043 PMID:22153071
24. Kan Z, Zheng H, Liu X et al. (2013). Whole- genome sequencing identifies recurrent mutations in hepatocellular carcinoma. Genome Res, 23:1422–1433. http://dx.doi. org/10.1101/gr.154492.113 PMID:23788652
25. El-Serag HB (2011). Hepatocellular car- cinoma. N Engl J Med, 365:1118–1127. http://dx.doi.org/10.1056/NEJMra1001683 PMID:21992124
26. Razumilava N, Gores GJ (2013). Classification, diagnosis, and management of cholangiocarcinoma. Clin Gastroenterol Hepatol, 11:13–21, e1. http://dx.doi.org/ 10.1016/j.cgh.2012.09.009 PMID:22982100
Chapter 5.7 • Pancreatic cancer 413
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .7
Summary • A majority of pancreatic can-
cers occur in countries with high or very high levels of hu- man development.
• Pancreatic cancer is the seventh most common cause of cancer death worldwide, with a 5-year survival rate of 5%. Infiltrating ductal adenocarcinoma is the most common tumour type (90%).
• The leading identified cause of pancreatic cancer is cigarette smoking. Body fatness is an es- tablished risk factor. New-onset diabetes can be an early sign of the disease.
• Exome sequencing of ductal adenocarcinoma has demon- strated 16 significantly mu- tated genes – those previously recognized (KRAS, CDKN2A, TP53, SMAD4, MLL3, ATM, TGFBR2, ARID1A, and SF3B1) as well as novel genes, includ- ing genes involved in chroma- tin modification (EPC1 and ARID2), DNA damage repair, and other mechanisms (ZIM2, MAP2K4, NALCN, SLC16A4, and MAGEA6).
• Among other pancreatic neo- plasms, serous cystadenomas
5.7 5 ORGAN SITE
Pancreatic cancer Ralph H. Hruban Günter Klöppel (reviewer)
G. Johan Offerhaus (reviewer)
are characterized by VHL mu- tations, solid-pseudopapillary neoplasms by β-catenin mu- tations, intraductal papillary mucinous neoplasms by mu- tations in GNAS, KRAS, and RNF43, and mucinous cystic neoplasms by mutations in RNF43, KRAS, and TP53.
• Inherited mutations in BRCA2, p16/CDKN2A, PRSS1, STK11, PALB2, and ATM and in DNA mismatch repair genes increase risk.
• Personalized therapy is slow- ly becoming a reality with poly(ADP-ribose) polymerase (PARP) inhibitors or mitomycin C for cancers with BRCA2 or PALB2 mutations, and everoli- mus for neuroendocrine tu- mours with mTOR pathway abnormalities.
Pancreatic cancer is not one dis- ease. Several distinct neoplasms with unique clinical and pathologi- cal features may arise in the gland. About 95% of all pancreatic cancers develop in the exocrine pancreas, with about two thirds of these in the head of the pancreas. Infiltrating ductal adenocarcinoma is the most common of these neoplasms (ac- counting for 90% of all pancreatic
cancer), and to a large degree the epidemiology of pancreatic cancer reflects the impact of these tumours. Ductal adenocarcinoma is the most aggressive of pancreatic neoplasms [1]. Most patients are diagnosed with metastatic disease, and only 5% sur- vive to 5 years.
Infiltrating ductal adenocarci- noma is an infiltrating epithelial neoplasm with glandular (ductal) differentiation, usually demonstrat- ing luminal and/or intracellular pro- duction of mucin, and without a pre- dominant component of any other histological type. An abundant des- moplastic stromal response is a typi- cal feature.
Pancreatic neuroendocrine tu- mours are less common (1–2% of all pancreatic cancer) but have a sig- nificantly better survival rate of 65% at 5 years [1]. In contrast, almost all patients with a serous cystad- enoma are cured. Intraductal papil- lary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms are important because they are curable precursor lesions that, if left untreat- ed, can progress to an incurable in- vasive carcinoma.
All of the major tumour types of the pancreas have been sequenced. This revolution in our understand- ing of these neoplasms has opened the door to an understanding of why pancreatic cancer aggregates
414
Epidemiology
Pancreatic cancer • Pancreatic cancer is estimated
to be the 12th most common cancer in men (178 000 cases) and the 11th most common in women (160 000 cases) world- wide in 2012; 68% of the new cases occurred in countries at high or very high levels of human development.
• There were an estimated 330 000 deaths from pancreatic cancer in 2012, and because of its very high fatality, pancreatic cancer is the seventh most common cause of death from cancer worldwide.
• In 2012, one third of the es- timated new cases occurred in Europe. The highest age- standardized incidence rates are found in central and east- ern Europe, North America, Argentina, and Uruguay, and among women in Australia. Relatively low incidence rates are observed in most countries in Africa and East Asia.
• Given the overall mortality-to- incidence ratio of 0.98, the geo- graphical patterns and trends for mortality are very similar to those observed for incidence.
• Trends in incidence and mortal- ity rates in both sexes tend to be rather stable over time.
Map 5.7.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for pancreatic cancer in men, 2012.
Map 5.7.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for pancreatic cancer in women, 2012.
in some families, has helped in the identification of precursor lesions that are likely to progress to invasive cancer, and has created opportuni- ties for personalized therapy.
Etiology Non-modifiable risk factors for ductal adenocarcinoma of the pancreas in- clude older age (most cases occur
after the age of 65 years); race (in the USA, risk for Black populations is ≥ 1.5 that for White populations), and adult attained height. Cigarette smoking causes pancreatic cancer [2]. From meta-analysis, risk is sig- nificantly increased, by 74% and 20% for current and former smok- ers, respectively; pooled analysis indicated a significant elevation of
77% for current smokers and a non- significant increase of 9% for former smokers. Smoking is estimated to cause 20% of pancreatic cancer. Smoking cessation reduces risk, and by 20 years after cessation the risk for former smokers drops to that of never-smokers [3].
World Cancer Research Fund evaluations [4] specify convincing
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.7 • Pancreatic cancer 415
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .7
Chart 5.7.1. Estimated global number of new cases and deaths with proportions by major world regions, for pancreatic cancer in both sexes combined, 2012.
Chart 5.7.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for pancreatic cancer in men, circa 1975–2012.
Chart 5.7.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for pancreatic cancer in women, circa 1975–2012.
416
evidence that body fatness increases risk and probable evidence in rela- tion to abdominal fatness, whereas evidence is limited for red meat con- sumption; coffee drinking is unlikely to affect risk. Other risk factors for pancreatic cancer include diabetes and non-O ABO blood type [1,3]. Chronic pancreatitis is an established risk factor for pancreatic ductal ad- enocarcinoma. However, most pa- tients with this disease do not develop cancer, and most pancreatic cancer patients do not have a history of pan- creatitis [5].
Pathology Infiltrating ductal adenocarci- noma of the pancreas elicits an intense desmoplastic stromal reac- tion (Fig. 5.7.2) [1]. As a result, the bulk of the tumour is composed of collagen, stromal cells, inflamma- tory cells, and blood vessels. This dense desmoplastic stroma has two important clinical implications. First, biopsies may miss the neoplastic cells. Second, the dense fibrosis may hinder the delivery of chemo- therapeutic agents to the neoplastic cells [6]. Perineural invasion is also
common, and many patients with pancreatic cancer experience pain.
Pancreatic intraepithelial neo- plasia lesions are non-invasive mi- croscopic epithelial proliferations in the smaller pancreatic ducts [7]. A growing body of evidence sug- gests that pancreatic intraepithelial neoplasia can be a precursor to infiltrating ductal adenocarcinoma. These lesions harbour many of the same genetic mutations found in in- filtrating ductal adenocarcinomas, and rare clinical cases have been reported of patients with pancreat- ic intraepithelial neoplasia lesions who later developed an invasive pancreatic cancer. Although most of these precursor lesions are too small to be detected using currently available imaging techniques, they are a potential target for future early detection strategies.
Pancreatic ductal adenocarci- nomas comprise a hierarchy of tu- mour cells that have been hypoth- esized by some to develop around a population of cancer initiating cells. Primary recognized markers of these cancer initiating cells are CD44, CD24, and ESA [8]. Immune cell infiltrates occur during tumour progression. Infiltration by immune cells with immunosuppressive ac- tivities allows for an environment that fosters tumour growth and pro- gression [5]. One contributing factor to the failure of systemic cytotoxic and targeted therapies may be the abundant tumour stromal content that is characteristic of pancreatic ductal adenocarcinoma. The stro- ma, sometimes referred to as the tumour microenvironment, occupies the majority of the tumour mass and can be targeted in the development of novel therapy [9].
Genetics The exomes of several ductal adeno- carcinomas have been sequenced, and the most commonly mutated genes include an oncogene, KRAS, and three tumour suppressor genes, TP53, p16/CDKN2A, and SMAD4 (Table 5.7.1) [10]. An understand- ing of the genes targeted in ductal adenocarcinomas not only provides
insight into the fundamental nature of these neoplasms but also has clini- cal implications. For example, loss of Smad4 protein expression as as- sessed by immunolabelling is a sur- rogate marker for SMAD4 gene mu- tations, and can be used to suggest origin in the pancreas for a tumour of unknown origin.
Exome sequencing and copy number analysis have more recently been reported for a prospectively ac- crued clinical cohort (n = 142) of early (stage I and II) sporadic pancreatic ductal adenocarcinoma [11]. There were 16 significantly mutated genes, reaffirming known mutations (KRAS, TP53, CDKN2A, SMAD4, MLL3, ATM, TGFBR2, ARID1A, and SF3B1) and identifying novel mutated genes, including additional genes involved in chromatin modification (EPC1 and ARID2), DNA damage repair, and other mechanisms (ZIM2, MAP2K4, NALCN, SLC16A4, and MAGEA6). Also reported were somatic aberra- tions in genes described as embry- onic regulators of axon guidance, particularly SLIT/ROBO signalling. Loss of SLIT/ROBO signalling may be an alternative mechanism for dys- regulating pathways downstream of their receptors, and in addition could influence the activity of inhibitors that target these upstream components, such as MET inhibitors (Fig. 5.7.3).
Approximately 10% of pancreatic cancer is believed to have a famil- ial basis, and several genes have been identified that, when mutated in the germline, increase the risk of
Fig. 5.7.1. A doctor examines a patient in relation to pancreatic disease. Worldwide pancreatic cancer rates are highest for African American men.
Fig. 5.7.2. Photomicrograph of infiltrating ductal adenocarcinoma of the pancreas. Note the haphazard arrangement of the glands and the intense stromal fibrosis.
Chapter 5.7 • Pancreatic cancer 417
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .7
pancreatic cancer. These include p16/ CDKN2A, BRCA2, PALB2, PRSS1, STK11, and ATM, and DNA mismatch repair genes (Table 5.7.2) [12].
Prospects Although ductal adenocarcinomas of the pancreas have seemed to present an impenetrable barrier to progress, several recent advances provide hope. First, although the desmoplas- tic response may hinder the delivery of chemotherapeutic agents to the neoplastic cells, recent studies in animal models specifically targeting the stroma as a way to improve drug delivery have shown promise [6].
Second, recent studies of the genetic evolution of pancreatic neoplasia suggest that it takes many years for a genetically altered cell in the pancreas to invade and eventu- ally metastasize. This suggests a large window of opportunity for the early detection of curable pancreatic neoplasia. If pancreatic intraepithe- lial neoplasia and cystic precursor lesions with high-grade dysplasia (discussed below) can be identified early, they can be treated before an infiltrating ductal adenocarcinoma develops.
Third, the discovery of some of the genes responsible for the familial
aggregation of pancreatic cancer means that it is now possible to identify a population at risk that would most benefit from early detection. With the exception of PRSS1, these germline alterations also increase the risk of extrapancreatic malignancies. Therefore, mortality can be reduced by screening for one of the extrapan- creatic malignancies (Table 5.7.2). For example, germline p16/CDKN2A mutations increase the risk of both pancreatic cancer and melanoma, and careful skin examinations of car- riers of p16/CDKN2A mutations can aid early detection [12].
Table 5.7.1. The most common genetic alterations in ductal adenocarcinoma
Gene Chromosome Mechanism of alteration Percentage of cancers
Oncogenes
KRAS2 12p Point mutation 95
BRAF 7q Point mutation 5 (especially those with MSI)
AIB1 20q Amplification Up to 60
AKT2 19q Amplification 10–20
MYB 6q Amplification 10
Tumour suppressor genes
p16/CDKN2A 9p Point mutation plus LOH, HD, promoter methylation 95
TP53 17p Point mutation plus LOH 75
SMAD4 18q Point mutation plus LOH, HD 55
USP9X X Unclear 50
FAM190A 4q Internal gene rearrangements Up to 40
EP300 22q Point mutation plus LOH 25
ARID1A 1p Point mutation plus LOH 8
MLL3 7q Point mutation 6
TGFbetaR2 3p HD, biallelic intragenic mutations 4–7 (especially those with MSI)
STK11/LKB1 19p Germline, point mutation plus LOH 4–6
FBXW7 4q Point mutation plus LOH < 5
MKK4 17p Point mutation plus LOH, HD 4
ATM 11q Point mutation plus LOH 3
TGFbetaR1 9q HD 2
ACVR1beta 12q Point mutation plus LOH, HD 2
ACVR2 2q Point mutation plus LOH, biallelic intragenic mutations
Tumours with MSI
HD, homozygous deletion; LOH, loss of heterozygosity; MSI, microsatellite instability.
418
Finally, a small percentage of the genetic changes identified are targetable with existing therapeutic agents. For example, it has been sug- gested that infiltrating ductal adeno- carcinomas with biallelic inactivation of one of the Fanconi anaemia genes, including BRCA2 and PALB2, are more sensitive to poly(ADP-ribose) polymerase (PARP) inhibitors and to mitomycin C [6]. This suggests a paradigm for personalized therapy in which tumours can be biopsied or resected and the optimal therapy guided by genetic analyses of the neoplastic cells.
Other tumour types Cystic neoplasms Intraductal papillary mucinous neoplasms IPMNs are non-invasive, usually pap- illary, grossly visible mucin-producing epithelial neoplasms involving the larger pancreatic ducts (Fig. 5.7.4) [1].
Little is known about the epi- demiology of these neoplasms, al- though patients with an IPMN have an increased risk of developing co- lon cancer. A review of computed tomography scans performed for in- dications unrelated to the pancreas revealed pancreatic cysts in 2.6% of the patients scanned, and this in- creased to a prevalence of 8.7% in individuals aged 80–89 years [13]. In individuals with a strong family history
Table 5.7.2. Familial pancreatic cancer syndromes
Hereditary condition Gene (chromosome) Increase in lifetime risk of pancreatic cancer
Other cancers
Familial breast cancer, BRCA2 BRCA2 (13q) 3.5–10× Breast, ovarian, prostate
Hereditary pancreatitis PRSS1 (7q) 50–80× None
Familial atypical multiple mole melanoma p16/CDKN2A (9p) 20–34× Melanoma
Familial breast cancer, PALB2 PALB2 (16p) Unknown Breast
Peutz–Jeghers STK11 (19p) 100–132× Gastrointestinal tract, breast, and others
Hereditary non-polyposis colorectal cancer
MSH2 (2p) MLH1 (3p) PMS2 (7p) MSH6 (2p)
8–9× Colorectal, endometrial, ureteral, and others
ATM ATM (11q) Unknown Breast?
Fig. 5.7.3. SLIT/ROBO signalling in pancreatic ductal adenocarcinoma. SLIT/ROBO signalling normally enhances β-catenin complex formation with E-cadherin and sup- presses WNT signalling activity. Loss of ROBO1/2 signalling promotes stabilization of β-catenin, which decreases E-cadherin complex formation and cell adhesion and aug- ments WNT signalling activity through increased nuclear translocation of β-catenin. In addition, SLIT/ROBO signalling can downregulate MET signalling activity; loss of ROBO signalling activity promotes MET signalling downstream and may have an impact on therapeutic strategies aimed at inhibiting MET activity at the receptor level. Aberrations in SLIT2 and/or ROBO1/2 affect 23% of patients.
Chapter 5.7 • Pancreatic cancer 419
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .7
of pancreatic cancer, the prevalence of these cysts is as high as 40% [14].
IPMNs arise in the larger pan- creatic ducts and, in contrast to pancreatic intraepithelial neoplasia, most are grossly visible [1]. IPMNs can have varying grades of dys- plasia and can be classified by the predominant direction of differentia- tion into intestinal, pancreatobiliary, oncocytic, and gastric subtypes. If untreated, as many as one third progress to infiltrating adenocarci- noma. IPMNs that involve the main pancreatic duct are more likely to have an associated invasive car- cinoma than are those that involve a smaller branch duct [15]. IPMNs are often multifocal, and multifocality has been established at the genetic level. The exomes of a series of well- characterized IPMNs have recently
been sequenced, and the three genes mutated most frequently in these neoplasms are KRAS, GNAS, and RNF43 (Table 5.7.3) [16,17]. The SMAD4 and TP53 tumour suppres- sor genes are inactivated in some high-grade IPMNs.
IPMNs present a unique oppor- tunity to cure pancreatic neoplasia before an invasive carcinoma devel- ops. They are detectable using exist- ing imaging technologies, and some will progress to invasive cancer if left untreated. However, the risk of over- treating patients is significant. Not all cysts in the pancreas are IPMNs. Some are less aggressive serous cystadenomas. In addition, most IPMNs do not progress to infiltrat- ing carcinoma. The Sendai criteria should be used as a guide for decid- ing which IPMNs to resect and which to observe [15]. These criteria sug- gest the resection of all main-duct IPMNs and the resection of branch- duct IPMNs that cause symptoms, have an associated mural nodule, or are associated with dilatation of the main duct. Looking forward, it is easy to imagine that an evaluation of the genetic changes in cyst fluid samples will help guide the manage- ment of pancreatic cysts.
Mucinous cystic neoplasms Mucinous cystic neoplasms are non- invasive mucin-producing epithelial neoplasms of the pancreas charac- terized by a distinctive “ovarian-type” stroma. Most occur in adult women,
with a female-to-male ratio of 20. The etiology is unknown.
In contrast to IPMNs, the cysts of mucinous cystic neoplasms usually do not communicate with the larger pancreatic ducts [1]. Most arise in the tail of the gland. Mucinous cystic neoplasms can show varying grades of dysplasia, and up to one third prog- ress to infiltrating adenocarcinoma. The exomes of a series of mucinous cystic neoplasms have recently been sequenced, and the three genes mu- tated most frequently in these neo- plasms are KRAS, TP53, and RNF43 (Table 5.7.3) [17].
As is true for IPMNs, mucinous cystic neoplasms offer an opportu- nity to detect and treat early curable pancreatic neoplasia before an incur- able invasive carcinoma develops. In contrast to IPMNs, mucinous cystic neoplasms do not have a significant risk of multifocal disease. Surgical re- section of a non-invasive mucinous cystic neoplasm is therefore almost always curative.
Serous cystadenomas A serous cystadenoma is a usually cystic neoplasm composed of uni- form cuboidal glycogen-rich neoplas- tic cells. There is a slight female pre- dominance. Most cases are sporadic, but there is an association with von Hippel–Lindau syndrome.
The classic gross appearance of a serous cystadenoma is a neoplasm composed of innumerable small, thin- walled cysts with a central star-shaped scar, often calcified [1]. The cysts are lined by cuboidal neoplastic cells with round, uniform nuclei. The VHL gene is inactivated in at least half of these tumours [17].
Since serous cystadenomas are almost always benign, the goal of treatment is not to resect these neo- plasms unless they are very large or symptomatic. The challenge is that serous cystadenomas can mimic the more aggressive IPMNs on imaging. It is hoped that preoperative evaluation of cyst fluid for markers of VHL inacti- vation will help guide the management of these lesions in the future.
Table 5.7.3. Genetic alterations in non-ductal neoplasms of the pancreas
Tumour type Gene (chromosome)
Intraductal papillary mucinous neoplasm GNAS (20q), RNF43 (17q), KRAS (12p), TP53 (17p), SMAD4 (18q)
Mucinous cystic neoplasm RNF43 (17q), KRAS (12p), TP53 (17p)
Solid-pseudopapillary neoplasm CTNNB1 (the β-catenin gene) (3p)
Serous cystic neoplasm VHL (3p)
Pancreatic neuroendocrine tumour MEN1 (11q), ATRX (X), DAXX (6p), TSC2 (16p), PTEN (10q), PIK3CA(3q)
Acinar carcinoma SMAD4 (18q), JAK1 (1p), BRAF (7q), RB1 (13q), TP53 (17p)
Fig. 5.7.4. Photomicrograph of an intra- ductal papillary mucinous neoplasm with high-grade dysplasia. This neoplasm ex- tensively involved the larger pancreatic ducts. Note the luminal mucin and the non-invasive nature of the lesion.
420
Solid-pseudopapillary neoplasms A solid-pseudopapillary neoplasm is a low-grade malignant epithelial neoplasm composed of poorly co- hesive cells. Most occur in women (with a female-to-male ratio of 10), usually in their twenties. The etiology is unknown.
While solid-pseudopapillary neo- plasms can be solid, most undergo cystic degeneration. When they de- generate, the neoplasm has a dis- tinct microscopic appearance with foam cells and the formation of pseudopapillae [1]. Genetically, solid- pseudopapillary neoplasms are char- acterized by mutations in CTNNB1, the β-catenin gene (Table 5.7.3) [17].
β-Catenin gene mutations result in an abnormal nuclear pattern of expression of the β-catenin protein. Immunolabelling for the β-catenin protein can therefore aid in diag- nosis. Most solid-pseudopapillary neoplasms are cured by surgical resection.
Acinar carcinoma Interestingly, although the normal pan- creas largely consists of acinar cells, cancers with acinar differentiation are rare (only 1–2% of the cancers) [1]. Some patients with an acinar cell carcinoma present with the devastat- ing lipase hypersecretion syndrome,
in which lipase is released by the neoplasm into the circulation, causing subcutaneous fat necrosis, polyarthral- gias, and peripheral blood eosinophilia. Acinar cell carcinomas are fully malig- nant neoplasms with a 5-year survival rate of only 25–50% [18].
Pancreatic neuroendocrine tumours Pancreatic neuroendocrine tumours are epithelial neoplasms with signifi- cant neuroendocrine differentiation, as can be demonstrated by the expres- sion of synaptophysin or chromogra- nin. Most occur in individuals aged 30– 60 years. Pancreatic neuroendocrine tumours may be sporadic or arise in patients with a genetic syndrome such as multiple endocrine neoplasia type 1 (MEN1) or von Hippel–Lindau syn- drome. The etiology is unknown.
Pancreatic neuroendocrine tu- mours are richly vascular neoplasms usually composed of nests, trabecu- lae, or sheets of relatively uniform cells with characteristic “salt and pepper” nuclei (Fig. 5.7.6) [1]. They are graded histologically based on the proliferation rate. Grade 1 pancre- atic neuroendocrine tumours have a Ki-67 labelling index of 0–2%, grade 2 an index of 3–20%, and grade 3 (which are given the designation neu- roendocrine carcinoma) an index of > 20%. Grade and stage are important
prognosticators. The exomes of a series of these tumours have been sequenced, and three “mountains” were identified (Table 5.7.3). The MEN1 gene is inactivated in about 45% of pancreatic neuroendocrine tumours, ATRX/DAXX in 43%, and one of the mammalian target of rapa- mycin (mTOR) pathway genes (TSC2, PTEN, or PIK3CA) in 16% [19].
Surgery is the treatment of choice, but most patients have metastatic disease at diagnosis. Somatostatin analogues (including octreotide) have been shown to slow tumour growth, and several significant advances have recently been made in targeted therapy for these tumours. The multiple tyrosine kinase inhibitor sunitinib has shown promise in phase III clinical tri- als, as has the mTOR pathway inhibitor everolimus. Although it has not been shown clinically, targeting the mTOR pathway may be most beneficial to pa- tients with pancreatic neuroendocrine tumours that harbour mutations in an mTOR pathway gene [19,20].
Conclusions A better understanding of the precur- sors to invasive adenocarcinoma of the pancreas may form the basis for improved early detection. Genetic changes can be used to classify neo- plasms and thereby guide therapy, and in a small but growing number of cases genetic mutations that produce a therapeutically targetable change offer a promise of advances towards a future of personalized medicine.
Fig. 5.7.5. Mucinous non-cystic adenocarcinoma of the pancreas. The Whipple resection specimen shows a well-demarcated tumour in the head of the pancreas with a nodular pattern and a gelatinous cut surface. The probe indicates the pancreatic duct, which runs into the minor papilla.
Fig. 5.7.6. Photomicrograph of a pan- creatic neuroendocrine tumour. Note the nested growth pattern, the absence of a desmoplastic stroma, and the “salt and pepper” chromatin pattern.
Chapter 5.7 • Pancreatic cancer 421
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .7
References
1. Hruban RH, Pitman MB, Klimstra DS (2007). Tumors of the Pancreas (AFIP Atlas of Tumor Pathology Series 4, Fascicle 6). Washington, DC: American Registry of Pathology and Armed Forces Institute of Pathology.
2. IARC (2012). Personal habits and indoor combustions. IARC Monogr Eval Carcinog Risks Hum, 100E:1–575. PMID:23193840
3. Lowenfels AB, Maisonneuve P (2006). Epidemiology and risk factors for pancreatic cancer. Best Pract Res Clin Gastroenterol, 20:197–209. http://dx.doi.org/10.1016/j.bpg. 2005.10.001 PMID:16549324
4. World Cancer Research Fund/American Institute for Cancer Research (2007). Food, Nutrition, Physical Activity and the Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research.
5. Zheng L, Xue J, Jaffee EM, Habtezion A (2013). Role of immune cells and im- mune-based therapies in pancreatitis and pancreatic ductal adenocarcinoma. Gastroenterology, 144:1230–1240. http:// dx.doi.org/10.1053/j.gastro.2012.12.042 PMID:23622132
6. Olive KP, Jacobetz MA, Davidson CJ et al. (2009). Inhibition of Hedgehog signaling enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science, 324:1457–1461. http://dx.doi.org/10.1126/ science.1171362 PMID:19460966
7. Feldmann G, Beaty R, Hruban RH, Maitra A (2007). Molecular genetics of pancreatic intraepithelial neoplasia. J Hepatobiliary Pancreat Surg, 14:224–232. http://dx. doi.org/10.1007/s00534-006-1166-5 PMID: 17520196
8. Abel EV, Simeone DM (2013). Biology and clinical applications of pancreatic cancer stem cells. Gastroenterology, 144:1241– 1248. http://dx.doi.org/10.1053/j.gastro. 2013.01.072 PMID:23622133
9. Feig C, Gopinathan A, Neesse A et al. (2012). The pancreas cancer microenvi- ronment. Clin Cancer Res, 18:4266– 4276. http://dx.doi.org/10.1158/1078-0432.CCR- 11-3114 PMID:22896693
10. Jones S, Zhang X, Parsons DW et al. (2008). Core signaling pathways in human pancreatic cancers revealed by global ge- nomic analyses. Science, 321:1801–1806. http://dx.doi.org/10.1126/science.1164368 PMID:18772397
11. Biankin AV, Waddell N, Kassahn KS et al.; Australian Pancreatic Cancer Genome Initiative (2012). Pancreatic cancer ge- nomes reveal aberrations in axon guid- ance pathway genes. Nature, 491:399– 405. http://dx.doi.org/10.1038/nature11547 PMID:23103869
12. Shi C, Hruban RH, Klein AP (2009). Familial pancreatic cancer. Arch Pathol Lab Med, 133:365–374. PMID:19260742
13. Laffan TA, Horton KM, Klein AP et al. (2008). Prevalence of unsuspected pancre- atic cysts on MDCT. AJR Am J Roentgenol, 191:802–807. http://dx.doi.org/10.2214/ AJR.07.3340 PMID:18716113
14. Canto MI, Hruban RH, Fishman EK et al.; American Cancer of the Pancreas Screening (CAPS) Consortium (2012). Frequent detection of pancreatic lesions in asymptomatic high-risk individuals. Gastroenterology, 142:796 – 804. http:// dx.doi.org/10.1053/j.gastro.2012.01.005 PMID:22245846
15. Tanaka M, Fernández-del Castillo C, Adsay V et al.; International Association of Pancreatology (2012). International consensus guidelines 2012 for the man- agement of IPMN and MCN of the pan- creas. Pancreatology, 12:183–197. http:// dx.doi.org/10.1016/j.pan. 2012.0 4.0 0 4 PMID:22687371
16. Wu J, Matthaei H, Maitra A et al. (2011). Recurrent GNAS mutations define an unex- pected pathway for pancreatic cyst devel- opment. Sci Transl Med, 3:92ra66. http:// dx.doi.org/10.1126/scitranslmed.3002543 PMID:21775669
17. Wu J, Jiao Y, Dal Molin M et al. (2011). Whole-exome sequencing of neoplastic cysts of the pancreas reveals recurrent mutations in components of ubiquitin- dependent pathways. Proc Natl Acad Sci U S A, 108:21188–21193. http://dx. doi.org/10.1073/pnas.1118046108 PMID: 22158988
18. Jiao Y, Yonescu R, Offerhaus GJA et al. (2013). Whole exome sequencing of pancreatic neoplasms with acinar dif- ferentiation. J Pathol, [epub ahead of print]. http://dx.doi.org/10.1002/path.4310 PMID:24293293
19. Jiao Y, Shi C, Edil BH et al. (2011). DAXX/ ATRX, MEN1, and mTOR pathway genes are frequently altered in pancreatic neuroen- docrine tumors. Science, 331:1199–1203. http://dx.doi.org/10.1126/science.1200609 PMID:21252315
20. Yao JC, Shah MH, Ito T et al.; RAD001 in Advanced Neuroendocrine Tumors, Third Trial (RADIANT-3) Study Group (2011). Everolimus for advanced pancre- atic neuroendocrine tumors. N Engl J Med, 364:514–523. http://dx.doi.org/10.1056/NE JMoa1009290 PMID:21306238
Websites American Cancer Society Pancreatic Cancer home page: http://www.cancer.org/cancer/ pancreaticcancer/index
International Cancer Genome Consortium: http://www.icgc.org/icgc/cgp/68/304/798
Sol Goldman Pancreatic Cancer Research Center: http://pathology.jhu.edu/pc
The Cancer Genome Atlas: http://cancergenome.nih.gov/
422
Summary
• Tobacco smoking, alone and in combination with alcohol, is the most important cause of head and neck cancer.
• Most head and neck cancers are squamous cell carcinoma.
• Infection by human papillomavi- rus causes cancers of the oro- pharynx and base of tongue.
• Nasopharyngeal carcinomas are common in parts of South- East Asia and North Africa; their etiology involves Epstein–Barr virus, volatile nitrosamines, and genetic factors.
• Genetic alterations in oral and laryngeal cancer include acti- vation of cyclin D1, MYC, RAS, PIK3CA, and EGFR and inac- tivation of tumour suppressor genes such as p16INK4A, TP53, and PTEN.
• Mutations in PI3K pathway genes are proposed as a pre- dictive biomarker.
• Early-stage tumours of the up- per aerodigestive tract can be cured; for late-stage disease, prognosis is poor.
5.8 5 ORGAN SITE
Head and neck cancers Lester D.R. Thompson Paul Brennan (reviewer)
Luis Felipe Ribeiro Pinto (reviewer)
Head and neck cancers are a related group of cancers that involve the oral cavity, pharynx (oropharynx, naso- pharynx, hypopharynx), and larynx (Fig. 5.8.1). Most tumours of the oral cavity, larynx, and hypophar- ynx are squamous cell carcinoma arising from the squamous cells that line these spaces. Other tumours that develop in this area (sinonasal tract, salivary glands) are relatively uncommon. Although related, there are four separate anatomical groups, with distinctly different etiologies but similar metastatic pathways.
Etiology Cancers of the oral cavity and of the larynx and hypopharynx The most significant causes of all head and neck cancers are tobac- co use and alcohol consumption. These exposures account for the development of approximately 80% of such cancers globally, with some variation for different subsites (65% for oral cavity vs 86% for larynx) [1]. Smoking of cigarettes or bidis poses the most important risk for cancer development, related to the product of smoking rate in packs per day and duration of smoking in years (“pack-years”), and risk is higher for longer duration of smoking, smok- ers of black tobacco or high-tar
cigarettes versus blond tobacco, young age at start of smoking, and deep smoke inhalation. Cigar and pipe smoking also pose a risk, al- though slightly less [1,2]. The rela- tive risk is higher for glottic than for supraglottic carcinoma. Use of chewing tobacco or smokeless to- bacco, and combinations with other substances such as paan or betel
Fig. 5.8.1. The major anatomical sites within the head and neck include the na- sal cavity, nasopharynx (A, purple), oral cavity, oropharynx (B, blue), larynx, and hypopharynx (C, green). The pharynx is divided using the junction between the hard and soft palate as the start of the nasopharynx; the inferior surface of the soft palate, uvula, base of tongue, tonsils, and tonsillar pillars within the oropharynx; and the hypopharynx extending from the superior border of the hyoid bone to the inferior border of the cricoid cartilage.
Chapter 5.8 • Head and neck cancers 423
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .8
Epidemiology
Laryngeal cancer • Laryngeal cancer is the 14th
most common cancer among men, but it is relatively rare in women: of the 157 000 esti- mated new cases worldwide in 2012, less than 19 000 occurred in women. The estimated total number of deaths from laryngeal cancer was about 83 000.
• About 53% of the cases occurred in countries at high or very high levels of human development; 49% of the cases occurred in Asia.
• The highest incidence rates in men tend to be observed in east- ern Europe and in certain coun- tries in the Caribbean.
Cancer of the lip, oral cavity, and pharynx • For cancers of the oral cavity and
pharynx, an estimated 529 000 new cases occurred worldwide in 2012, with 292 000 deaths. When the main subsites (lip, oral cavity, nasopharynx, and pharynx) are examined sepa- rately, they do not rank highly, but combined would rank above cervical cancer as the seventh most frequent type of cancer by incidence and the ninth most common cause of cancer death.
• The highest incidence rates are in Papua New Guinea, Bangladesh, Hungary, and Sri Lanka. The 120 000 new cases and 88 000 deaths occurring in India represent almost one quar- ter and one third, respectively, of the total burden from these cancers.
• Incidence trends show a de- cline since the 1990s in popu- lations where the rates were high (India, China, Blacks in the USA, Australia), in both sexes. Conversely, some populations with historically rather low in- cidence rates show increasing trends (e.g. Denmark, Japan).
Map 5.8.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for laryngeal cancer in men, 2012.
Map 5.8.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for pharyngeal cancer in men, 2012.
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
424
Chart 5.8.1. Estimated global number of new cases and deaths with proportions by major world regions, for laryngeal cancer in both sexes combined, 2012.
Chart 5.8.2. Estimated global number of new cases and deaths with proportions by major world regions, for oral and pharyngeal cancer in both sexes combined, 2012.
quid (betel leaf, areca nut, lime, and tobacco) poses an increased risk for oral cancer development, which is highest in India and in Taiwan, China, and especially
affects the floor of the mouth [2]. The risk decreases within 10 years of smoking cessation and is the low- est for groups of never-smokers, such as Seventh-day Adventists and
Mormons. Compared with never- smokers/teetotallers, the relative risk of head and neck cancer is in- creased between 10- and 100-fold in people who drink and smoke
Chapter 5.8 • Head and neck cancers 425
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .8
Chart 5.8.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for laryngeal cancer in men, circa 1975–2012.
Chart 5.8.4. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for oral and pharyngeal cancer in men, circa 1975–2012.
heavily. If there were total absti- nence from drinking and smoking (or quid chewing) worldwide, the risk of oral, pharyngeal, and laryn- geal cancers would be extremely low.
Alcohol consumption shows a strong multiplicative effect with to- bacco, perhaps related to acetal- dehyde, an intermediate metabolite of ethanol and a known carcinogen [3]. There are significant differences between countries in terms of per capita average alcohol consumption and the preferred type of beverage (beer, liquor, or wine). However, even with these differences, the most commonly consumed alco- holic beverage appears to be the one most strongly associated with cancer risk. For increased alco- hol consumption, the relative risk is higher for supraglottic and hy- popharyngeal carcinoma than for
glottic and subglottic carcinoma. There is a greater risk of pharyn- geal and oral cancer related to the number of years of heavy drinking and not the number of drinks per day [1]. There is an increased rela- tive risk of laryngeal carcinoma in patients who are heavy drinkers (> 8 drinks or > 207 ml/day) versus teetotallers or moderate drinkers, even without tobacco use. If indi- viduals are classified as alcohol- ics, there is an even stronger risk of cancer development. Tumours may develop within a background of field cancerization, with additional tumours arising from the same envi- ronmental or epigenetic milieu that gave rise to the first tumour, often in adjacent head and neck sites.
Oral and oropharyngeal carcino- mas are also associated with poor oral hygiene, habitual consumption of khat leaves, smoking marijuana,
and drinking maté, but causality with mouthwash use or with coffee drinking is not established [4].
Industrial exposure by working in construction, metal, textile, or ce- ramic jobs with exposure variously to isopropanol, polycyclic aromatic hydrocarbons, inorganic acid mists containing sulfuric acid and/or mustard gas, and diesel exhausts, and working in the food industry increase the risk of developing la- ryngeal cancer, after controlling for alcohol consumption and tobacco use. Human papillomavirus (HPV) may play a minor causative role in laryngeal and oral carcinomas [5]. Gastro-oesophageal reflux disease is associated with an increased risk of laryngeal carcinoma, but may also act as a promoter when alcohol and tobacco are used.
426
Cancers of the oropharynx and base of tongue Infection with HPV is one of the major leading causes of cancers of the oropharynx, tonsil, and base of tongue. Due to the relatively high HPV infection rate in sexually ac- tive people (> 50%), a significant in- crease in the incidence of this type of cancer has occurred over the past several decades. This increase is correlated with changes in sexual habits, including the practice of oral sex, lifetime and recent number of sexual partners, the practice of pre- marital sex, and earlier ages of initial sexual activity, all of which contrib- ute to the likelihood of oral infection. Biologically active HPV prevalence in oropharyngeal cancers ranges from 28% in Europe to 46% in Asia and 47% in North America, although more recent studies show rates of 65 –70% [6]. HPV16 accounts for nearly all oropharyngeal cancers, while HPV18 is also a contributor to oral cavity cancers. The odds ratio for oropharyngeal oncogenic type- specific HPV16 E6/E7 antibodies is 72.8 (confidence interval, 16–330), which underscores the causal asso- ciation of HPV with oropharyngeal carcinoma, along with integration of HPV DNA into the human genome [7], high HPV viral copy numbers [8], and high-level expression of HPV oncogenes (E6 and E7) in tu- mours. HPV also plays a causative role in selected oral (20.2%, weight- ed prevalence [WP]), sinonasal
(29.6%, WP), laryngeal (23.6%, WP), and nasopharyngeal (31.1%, WP) tumours [5]. Marijuana use is also associated with HPV-positive oro- pharyngeal cancer development [9].
Nasopharyngeal cancer Nasopharyngeal carcinoma exhibits a strong association with Epstein–Barr virus (EBV) infection, occupational exposures to wood dusts and formal- dehyde, cigarette smoking, radiation exposure, consumption of specific preserved or salted foods, malaria infection, and genetic predisposi- tion, as determined by multivariate- adjusted hazard ratios [10].
EBV is consistently implicated in nasopharyngeal carcinoma on- cogenesis. EBV infection seems to be an early initiating event as EBV is found in precursor lesions and shows a clonal episomal form, indicating that the virus entered the tumour cell nucleus before clonal expansion started. This phenomenon results in nearly every tumour cell containing the full length of the EBV genome. Positive serology for the presence of immunoglobulin A antibodies to EBV capsid antigen is present in up to 93% of patients, with an increased titre correlated with high tumour bur- den. This elevated titre may be used as a marker to screen populations in high-risk areas or to detect tu- mour recurrence [11]. EBV-encoded early RNA is expressed in nearly all tumour cells. This type of infection
shows a type II latency pattern, with expression of EBV nuclear antigen 1 (EBNA1) and latent membrane pro- tein 1 (LMP1). LMP1 can induce squamous hyperplasia, inhibit squa- mous differentiation, activate NF-κB, and induce expression of epidermal growth factor receptor (EGFR).
Meta-analysis of 16 case–con- trol studies showed that highest- versus-lowest intake of preserved vegetables was associated with a 2-fold increase in the risk of naso- pharyngeal cancer, while high intake of non-preserved vegetables was associated with a 36% decrease in the risk of this cancer, irrespective of vegetable type or country of study. It is thought that high levels of volatile nitrosamines in preserved food may be a putative carcinogen. Based on case–control studies, consumption of salt-preserved food (specifically as Cantonese-style salted fish), fer- mented foods, or foods with related preservation processes, including quaddid (dried mutton stored in oil), and consumption of rancid butter during weaning and early childhood have an adjusted relative risk of up to 7.5, increasing to 37.7 when con- sumed at least once per week ver- sus less than once per month at age 10 years [12].
Fig. 5.8.3. Tobacco smoking and alco- hol consumption are causes of head and neck cancers, specifically cancers of the oral cavity, pharynx, and larynx. People who both smoke and drink alcohol are at markedly greater risk of developing these cancers than those who either smoke or drink alcohol in the absence of the other habit, although both individually increase the likelihood of developing head and neck cancers.
Fig. 5.8.2. (A) Endoscopic view of a laryngeal cancer. (B) A laryngectomy specimen demonstrates an exophytic supraglottic squamous cell carcinoma.
Chapter 5.8 • Head and neck cancers 427
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .8
Pathology and genetics Cancer of the oral cavity and of the larynx and hypopharynx The squamous cell carcinomas of these two sites are similar one to another. There is in general a pro- gression from epithelial hyperplasia through dysplasia to carcinoma in situ to invasive carcinoma (Fig. 5.8.4), but not all invasive tumours arise from an overlying dysplasia. There is a loss of maturation, architectur- al disorganization, and increased pleomorphism and cell size with dyskeratosis and increased mitoses to a variable degree with each grade of dysplasia. Although the length of time from dysplasia to invasive car- cinoma is quite variable, it is gener- ally measured in years. It is during this interval that disease progression
may be altered by changes in habits (smoking, alcohol consumption) or surgical intervention. Tumours are separated into keratinizing and non- keratinizing, in situ versus invasive, and well, moderately, or poorly dif- ferentiated (Fig. 5.8.5). Variants are uncommon.
The genetic alterations observed in oral and laryngeal cancer are complex and interrelated. These alterations include activation of proto-oncogenes such as cyclin D1 (CDKN2A), MYC, RAS, PIK3CA, and EGFR and inactivation of tumour suppressor genes such as p16INK4A, TP53, and PTEN. TP53 mutations and overexpression are seen in the progression of precursor lesions to invasive carcinomas, although this is reported with greater frequency in developed countries (40–50%) than
in developing countries (5–25%). In contrast, tumours in patients from India and South-East Asia are characterized by the involvement of RAS oncogenes, including mutation, loss of heterozygosity (HRAS), and amplification (KRAS and NRAS). Mutations may also be seen in genes regulating squamous differentiation (such as NOTCH1, IRF6, and TP63) [13].
Cancers of the oropharynx and base of tongue HPV-positive oropharyngeal car- cinomas are distinct from HPV- negative cancers. HPV-positive cancers tend to be smaller primary tumours but with increased nodal involvement (high N stage), show- ing a poorly differentiated and non- keratinizing or basaloid histology
Fig. 5.8.4. The histological progression from normal to invasive squamous cell carcinoma is shown in parallel to genetic and epigenetic events. The accumulation of these genetic changes, and not the exact order, determines the progression to invasive carcinoma.
428
compared with HPV-negative cancers [6]. The tumours can be separated histologically into non- keratinizing, non-keratinizing with maturation, and keratinizing types; the first two types show the stron- gest association with HPV [14]. The tumours form nests with broad pushing borders and limited stromal response, frequent mitoses, and comedonecrosis. The cells form a syncytium, with indistinct borders, with hyperchromatic nuclei lacking prominent nucleoli. These tumours lack any squamous maturation or keratin pearl formation. Keratinizing squamous cell carcinomas have sig- nificant desmoplasia, with polygonal cells that show well-developed cell borders (intercellular bridges) and have abundant eosinophilic cyto- plasm. Tumours that are intermedi- ate (“with maturation”) are hybrid tumours but are much more closely associated with HPV than the kerati- nizing type. Although not absolutely correlated, immunohistochemical expression of p16 (> 75% of cells with strong nuclear and cytoplas- mic positive reaction) is considered an excellent surrogate marker for biologically active, integrated HPV infection. The interaction of HPV infection with TP53 mutations and EGFR expression requires further investigation as these markers typi- cally show a negative correlation with tumour HPV status.
Nasopharyngeal cancer WHO classifies nasopharyngeal can- cer into three major subtypes: non- keratinizing carcinoma, keratinizing carcinoma, and basaloid squamous cell carcinoma. Non-keratinizing carcinoma is the most common type and is most common in endemic populations, while the keratinizing and basaloid squamous cell carci- noma types tend to be more com- mon in non-endemic populations. Non-keratinizing carcinoma is further separated into two types: differenti- ated (stratified cells with pleomorphic, hyperchromatic nuclei surrounded by well-defined cell borders, sharply de- limited from the stroma but lacking a desmoplastic stromal response) and undifferentiated (a variable growth of cohesive cell nests to more indi- vidual dyscohesive cells within a lym- phoid stroma) (Fig. 5.8.6). The cells are large with indistinct cell borders, scant cytoplasm, and round nuclei with vesicular chromatin distribution, containing prominent nucleoli. An in situ or precursor lesion is rare. Viral infection works via many interrelat- ed somatic genetic and epigenetic changes that synergistically contrib- ute to the development of nasopha- ryngeal carcinoma. This pathway is exceedingly complex, but the meth- ylation of several tumour suppressor genes detected by nasopharyngeal brushings or in serum may aid in early detection and diagnosis.
Comprehensive genomic data are not available for nasopharyngeal cancer. One exome study based on biopsies indicated that the EBV- encoded genome maintenance protein EBNA1 along with LMP1, LMP2, and BARF1 were expressed in the majority of specimens stud- ied. Analyses also suggested that loss of the FHIT gene may be a driver of tumorigenesis [15]. In an- other study based on paraffin-em- bedded specimens, 41 microRNAs were differentially expressed be- tween nasopharyngeal tumour and corresponding non-cancer tissue, and a signature set of five was iden- tified [16].
Prospects Biomarkers For oral dysplasia, several biomark- ers have been shown to increase the risk of progression to cancer, including loss of heterozygosity, par- ticularly at the 3p ± 9p loci, survivin, matrix metalloproteinase 9 (MMP9), and DNA content. Other markers (p53, p73, MMP1, and MMP2), did not seem to predict progression [17]. HPV is a significant prognostic bio- marker in oropharyngeal squamous cell carcinoma, with the widest risk stratification for survival of any head and neck biomarker yet described, and thus has potential for important therapeutic considerations [18]. Whole-exome sequencing of 151 head and neck squamous cell car- cinomas revealed the phosphati- dylinositol 3-kinase (PI3K) pathway to be the most frequently mutated oncogenic pathway (30.5%) [19]. In a subset of HPV-positive tumours, PIK3CA or PIK3R1 was the only mu- tated cancer gene. All tumours with concurrent mutation of multiple PI3K pathway genes were advanced, im- plicating this pathway in tumour pro- gression and suggesting its use as a predictive biomarker.
Screening Screening programmes similar to cy- tology for cervical cancer do not seem to be applicable for oropharyngeal or oral cancer. Cytological abnormalities
Fig. 5.8.5. (A) A moderately differentiated, invasive, keratinizing squamous cell car- cinoma of the oral cavity. (B) A poorly differentiated, non-keratinizing squamous cell carcinoma of the larynx.
Chapter 5.8 • Head and neck cancers 429
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .8
are rarely detected even when there is known invasive disease, perhaps due to the difficulty of sampling the deep tonsillar crypts and to the lack of an identifiable precursor lesion. Therefore, delayed diagnosis is com- mon. Direct visual inspection of the oral cavity, oropharynx, nasophar- ynx, and larynx (assisted by endos- copy) or evaluation by imaging stud- ies in high-risk individuals may lead to early diagnosis, although large-scale screening programmes remain to be established on a global or regional scale. Significantly reduced oral cancer mortality has been observed in high-risk individuals in India under- going repeated rounds of screening by inspection of the oral cavity [20].
Screening for plasma EBV DNA levels might be helpful in detecting nasopharyngeal cancer, based on pooled meta-analysis data sensi- tivity (91.4%; confidence interval, 89.0–93.4%) and specificity (93.2%; 91.2–95.0%) [11]. Screening of high- risk families (family members of na- sopharyngeal carcinoma patients) or patients in endemic areas by EBV se- rology (for immunoglobulin A viral cap- sid antigen), EBV DNA, or other tech- niques (such as nasopharyngoscopy) has a potential benefit (sensitivity of 88.9% and specificity of 87.0%) as it detects cancers at a much ear- lier stage than symptomatic cancers (by up to 10 years), which yields im- proved disease-free survival.
Identification of high-risk groups Patients with certain polymorphisms in enzymes involved in detoxifying alcohol (alcohol and aldehyde de- hydrogenases [ADH]), show an in- creased susceptibility to developing carcinoma [3]. There seems to be a protective effect for patients with the ADH1B R48H variant and the ADH7 A92G variant, with a significant modi- fication based on increasing alcohol consumption; each gene’s effect is independent of the other’s.
There is an increased risk of up- per aerodigestive tract cancers in patients with the ADH1B Arg47Arg genotype, much greater than that for patients with a His47His genotype.
Furthermore, there is a modulation of susceptibility when combined with alcohol consumption and interactions with the ALDH2 487Lys allele [21]. Additional genetic polymorphisms seen in association with head and neck cancer include those in the GSTM1, GSTT1, and EPHX1 genes, XPD Lys751Gln, and P53 codon 72 Pro/Pro.
A family history of head and neck cancer in a first-degree relative in- creased the risk (odds ratio, 1.7; confidence interval, 1.2–2.3), which was even higher when the relative was a sibling (2.2; 1.6–3.1). This risk was limited to subjects exposed to tobacco and alcohol (7.2; 5.5–9.5) [22]. Hence, if there is a family his- tory of head and neck cancer, avoid- ing exposure to tobacco and alcohol reduces risk.
There are certain inherited con- ditions that are associated with an
increased risk of developing head and neck cancer, including dys- keratosis congenita and DNA repair pathway disorders, such as Bloom syndrome (helicase gene muta- tions), Fanconi anaemia (germline mutations in FAA, FAD, and FCC caretaker genes), ataxia telangiec- tasia (homozygotes), and xeroderma pigmentosum (XP genes). IARC has collected information from 15 centres to study alcohol-related cancer and genetic susceptibility in patients en- rolled from these centres, with a clear preponderance of smokers and alco- hol drinkers in head and neck cancer cases.
Prevention Avoiding cigarettes and alcohol could prevent up to 80% of oral cancer and up to 90% of laryn- geal and hypopharyngeal cancer. Interventions targeting smoking
Fig. 5.8.6. (A) A non-keratinizing nasopharyngeal carcinoma showing prominent nucleoli and amyloid. (B) Most nasopharyngeal carcinomas have a very heavy inflammatory infiltrate, which can obscure the neoplastic cells. (C) Nearly every one of the neoplastic cells shows a strong nuclear reaction with Epstein–Barr virus-encoded small RNA.
430
prevention or cessation are most effective when modelled by physi- cians, individual counsellors, or workplace- and school-based pro- grammes (see Chapter 4.1). Further, national smoke-free policies may have public health benefits for both
alcohol- and tobacco-related health problems.
The prophylactic use of the bi- valent (HPV16 and 18) or quad- rivalent (HPV6, 11, 16, and 18) vac- cines in males and females (aged 9–26 years), if proven efficacious,
holds great promise for primary prevention of HPV-associated oro- pharyngeal cancers since such a high proportion of HPV-associated oropharyngeal carcinomas are HPV16-derived [23].
Chapter 5.8 • Head and neck cancers 431
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .8
References 1. Lubin JH, Purdue M, Kelsey K et al. (2009).
Total exposure and exposure rate effects for alcohol and smoking and risk of head and neck cancer: a pooled analysis of case-control studies. Am J Epidemiol, 170:937–947. http://dx.doi.org/10.1093/ aje/kwp222 PMID:19745021
2. Pednekar MS, Gupta PC, Yeole BB, Hébert JR (2011). Association of tobacco habits, including bidi smoking, with overall and site-specific cancer incidence: results from the Mumbai cohort study. Cancer Causes Control, 22:859–868. http://dx. doi.org/10.1007/s10552-011-9756-1 PMID: 21431915
3. Xue Y, Wang M, Zhong D et al. (2012). ADH1C Ile350Val polymorphism and cancer risk: evidence from 35 case-con- trol studies. PLoS One, 7:e37227. http:// dx.doi.org/10.1371/journal.pone.0037227 PMID:22675424
4. Goldenberg D, Lee J, Koch WM et al. (2004). Habitual risk factors for head and neck cancer. Otolaryngol Head Neck Surg, 131:986–993. http://dx.doi.org/10.1016/j. otohns.2004.02.035 PMID:15577802
5. Isayeva T, Li Y, Maswahu D, Brandwein- Gensler M (2012). Human papillomavirus in non-oropharyngeal head and neck cancers: a systematic literature review. Head Neck Pathol, 6 Suppl 1:S104–S120. http://dx. doi.org/10.1007/s12105-012-0368-1 PMID: 22782230
6. Chaturvedi AK (2012). Epidemiology and clinical aspects of HPV in head and neck cancers. Head Neck Pathol, 6 Suppl 1:S16– S24. http://dx.doi.org/10.1007/s12105-012- 0377-0 PMID:22782220
7. Gillison ML, Koch WM, Capone RB et al. (2000). Evidence for a causal associa- tion between human papillomavirus and a subset of head and neck cancers. J Natl Cancer Inst, 92:709–720. http://dx.doi. org/10.1093/jnci/92.9.709 PMID:10793107
8. Kreimer AR, Clifford GM, Snijders PJ et al.; International Agency for Research on Cancer (IARC) Multicenter Oral Cancer Study Group (2005). HPV16 semiquan- titative viral load and serologic biomark- ers in oral and oropharyngeal squamous cell carcinomas. Int J Cancer, 115:329– 332. http://dx.doi.org/10.1002/ijc.20872 PMID:15688391
9. Gillison ML, D’Souza G, Westra W et al. (2008). Distinct risk factor profiles for hu- man papillomavirus type 16-positive and human papillomavirus type 16-negative head and neck cancers. J Natl Cancer Inst, 100:407– 420. http://dx.doi.org/10.1093/ jnci/djn025 PMID:18334711
10. Hildesheim A, Wang CP (2012). Genetic predisposition factors and nasopharyngeal carcinoma risk: a review of epidemiological association studies, 2000–2011: Rosetta Stone for NPC: genetics, viral infection, and other environmental factors. Semin Cancer Biol, 22:107–116. http://dx.doi.org/10.1016/j. semcancer.2012.01.007 PMID:22300735
11. Liu Y, Fang Z, Liu L et al. (2011). Detection of Epstein-Barr virus DNA in serum or plasma for nasopharyngeal cancer: a meta-analy- sis. Genet Test Mol Biomarkers, 15:495– 502. http://dx.doi.org/10.1089/gtmb.2011. 0012 PMID:21410354
12. Jia WH, Qin HD (2012). Non-viral environ- mental risk factors for nasopharyngeal car- cinoma: a systematic review. Semin Cancer Biol, 22:117–126. http://dx.doi.org/10.1016/j. semcancer.2012.01.009 PMID:22311401
13. Stransky N, Egloff AM, Tward AD et al. (2011). The mutational landscape of head and neck squamous cell carcino- ma. Science, 333:1157–1160. http://dx. doi.org/10.1126/science.1208130 PMID: 21798893
14. Chernock RD (2012). Morphologic features of conventional squamous cell carcinoma of the oropharynx: ‘keratinizing’ and ‘non- keratinizing’ histologic types as the basis for a consistent classification system. Head Neck Pathol, 6 Suppl 1:S41–S47. http://dx. doi.org/10.1007/s12105-012-0373-4 PMID: 22782222
15. Hu C, Wei W, Chen X et al. (2012). A global view of the oncogenic landscape in nasopharyngeal carcinoma: an inte- grated analysis at the genetic and expres- sion levels. PLoS One, 7:e41055. http://dx. doi.org/10.1371/journal.pone.0041055 PMID:22815911
16. Liu N, Chen NY, Cui RX et al. (2012). Prognostic value of a microRNA signature in nasopharyngeal carcinoma: a microRNA expression analysis. Lancet Oncol, 13:633– 641. http://dx.doi.org/10.1016/S1470-2045 (12)70102-X PMID:22560814
17. Smith J, Rattay T, McConkey C et al. (2009). Biomarkers in dysplasia of the oral cavity: a systematic review. Oral Oncol, 45:647–653. http://dx.doi.org/10.1016/j.oraloncology. 2009.02.006 PMID:19442563
18. Olthof NC, Straetmans JM, Snoeck R et al. (2012). Next-generation treatment strate- gies for human papillomavirus-related head and neck squamous cell carcinoma: where do we go? Rev Med Virol, 22:88–105. http://dx.doi.org/10.1002/rmv.714 PMID: 21984561
19. Lui VW, Hedberg ML, Li H et al. (2013). Frequent mutation of the PI3K pathway in head and neck cancer defines predictive biomarkers. Cancer Discov, 3:761–769. http://dx.doi.org/10.1158/2159-8290.CD- 13-0103 PMID:23619167
20. Sankaranarayanan R, Ramadas K, Somanathan T et al. (2013). Long term ef- fect of visual screening on oral cancer in- cidence and mortality in a randomized trial in Kerala, India. Oral Oncol, 49:314–321. http://dx.doi.org/10.1016/j.oraloncology. 2012.11.004 PMID:23265945
21. Cadoni G, Boccia S, Petrelli L et al. (2012). A review of genetic epidemiology of head and neck cancer related to polymorphisms in metabolic genes, cell cycle control and alcohol metabolism. Acta Otorhinolaryngol Ital, 32:1–11. PMID:22500060
22. Negri E, Boffetta P, Berthiller J et al. (2009). Family history of cancer: pooled analysis in the International Head and Neck Cancer Epidemiology Consortium. Int J Cancer, 124:394– 401. http://dx.doi. org/10.1002/ijc.23848 PMID:18814262
23. D’Souza G, Dempsey A (2011). The role of HPV in head and neck can- cer and review of the HPV vaccine. Prev Med, 53 Suppl 1:S5–S11. http:// dx.doi.o r g /10.1016 / j.y pme d. 2 011.0 8 . 001 PMID:21962471
Websites American Society of Clinical Oncology Head and Neck Cancer home page: http://www.cancer.net/cancer-types/head- and-neck-cancer
National Cancer Institute Head and Neck Cancer home page: http://cancer.gov/cancertopics/types/head- and-neck
National Cancer Institute Throat (Laryngeal and Pharyngeal) Cancer home page: http://cancer.gov/cancertopics/types/throat
The John Hopkins Medical Institution, Head and Neck Cancer home page: www.hopkinsmedicine.org/kimmel_cancer_ center/centers/head_neck/
Summary Evidence that overweight, obesity, and physical inactivity are caus- ally related to cancer is sufficiently strong to support strong actions to reduce these hazards. Multiple and increasingly intensive strategies will be needed to reverse the obesity epidemic, and actions are needed in many sectors and at all levels of society. Efforts to promote an overall healthy diet, including increases in fruits, vegetables, and whole grains and reductions in red meat, are well justified, but effects on risk of cancer specifically are likely to be modest. Continued research on diet, nutrition, and cancer is needed to expand the scientific basis for future progress to justify public health efforts.
A fundamental goal of research on diet and cancer is to identify con- stituents of food – including both classic nutrients and other aspects of diet – that increase or decrease the risk of cancer. This quest in- cludes epidemiological investiga- tions, experiments in animals, and in vitro studies. When the weight of ev- idence becomes sufficiently strong, this knowledge, which has usually been published piecemeal in scien- tific journals, needs to be translated
Diet, nutrition, and cancer: where next for public health?
Walter C. Willett
Walter C. Willett is a professor
of epidemiology and nutrition at the
Harvard School of Public Health and
a professor of medicine at Harvard
Medical School. His research uses
epidemiological approaches to
investigate the effects of dietary
factors on the cause and prevention
of cancer, cardiovascular disease,
and other conditions. Dr Willett
earned his medical degree from
the University of Michigan Medical
School and a Doctorate of Public
Health in epidemiology from the
Harvard School of Public Health.
His work emphasizes the long time
frames and large cohorts required to
obtain reliable data about diseases
that may take years to develop.
He has applied both questionnaire
and biochemical approaches in the
Nurses’ Health Studies I and II and the
Health Professionals Follow-up Study.
Dr Willet wrote the classic textbook
Nutritional Epidemiology and also
writes books on diet and nutrition for
general audiences.
into actions to prevent cancer. As discussed elsewhere in this Report, the process of research and transla- tion leading to reductions in cancer rates has been highly successful for many types of exposures, including tobacco use, radiation, pharmaceu- ticals, and occupational hazards. From these experiences, we have learned much about the process of cancer prevention that can be applied to dietary factors. Here, I briefly review our current state of knowledge on diet, nutrition, and cancer, public health approaches to translation of knowledge on diet and cancer, and suggestions for future public health directions.
Current state of knowledge Research on diet, nutrition, and cancer was limited until the late 1970s, when the large internation- al differences in cancer rates and some animal studies suggested that some aspects of diet might play major roles in the cause and prevention of cancer. In response, many case–control studies were conducted, and large prospec- tive studies were launched and are now producing abundant data. Progress has been more difficult than anticipated by many research- ers, partly because of the complex- ity of human diets but also due to the nature of cancer and its origins.
432
In particular, we have come to ap- preciate that many cancers consist of distinct diseases with different causes, and also that the process- es that lead to development of can- cer can operate over a lifetime, thus making the study of diet and can- cer exceedingly challenging. In our own research within several large prospective studies, we have also been investigating dietary factors in relation to risks of cardiovascular disease, diabetes, and many other outcomes, and clear associations between dietary factors and these diseases have emerged much more rapidly than for cancer, probably because the influences of dietary factors on the occurrence of these diseases are more immediate. This experience with other diseases has been important for research on cancer because it has documented that the methodologies to measure diet and to adjust for other factors do work well.
Despite the challenges of study- ing diet and cancer, several impor- tant conclusions have emerged. Most importantly, overweight and obesity have become established causes of many common cancers; this represents a major achieve- ment in cancer research. Although the risks of cancer for an individual who is overweight or obese are not as great as they are for a tobacco smoker, in the USA and some other countries the much higher prevalence of overweight and obe- sity than of smoking means that the numbers of cancer deaths caused by these two factors are now simi- lar [1]. Because the health conse- quences of obesity are not mani- fested immediately, the impact of the recent, rapid increases in the prevalence of obesity on cancer rates will continue to grow even if there is no further increase in prev- alence. Physical inactivity is also now well established as a risk factor for several cancers, in part through its contribution to overweight, but also directly. Consumption of red meat, particularly processed red meat, is related to modestly higher risks, and of fruits and vegetables
to modestly lower risks of some forms of cancer.
Given currently available knowl- edge, control of overweight and obesity must be a high priority for cancer prevention, and increas- ing physical activity should be part of this effort. The dietary factors related to obesity are many and complex, but sugar-sweetened bev- erages (e.g. soda) have emerged as a particularly important contrib- uting factor in many places (see Chapter 2.6).
Importantly, overweight or obe- sity, inadequate physical activity, and low intakes of fruits and vegeta- bles, together with tobacco, are also major risk factors for cardiovascular disease and type 2 diabetes. Thus, programmes for the prevention of cancer should be closely integrated with activities for the prevention of these other diseases.
Public health approaches for cancer prevention through improved diets To have a substantial impact on cancer rates, specific actions be- yond the publication of papers in scattered scientific journals are needed. Knowledge is now often transmitted widely by the gener- al media, such as television and newspapers, and some people will change their diets, become more active, or stop smoking based on this information alone. However, experience from tobacco control suggests that the impact of this will be limited. Also, with the Internet and other new channels of infor- mation, the public is being deluged with information on diet and health that is often sensationalist and out of context with other data, leav- ing many people more confused than they were with earlier, limited sources. Thus, a careful and coor- dinated public health approach for translation of nutritional knowledge will be needed to have an optimal impact; six levels of action with in- creasing intensity of intervention are described here [2]. Frieden has noted that, as compared with
individual counselling and clini- cal interventions, the greatest and most cost-effective public health impact will usually be achieved by changing the context so that the decisions of individuals are by default healthy ones, and by im- proving underlying socioeconomic factors [3].
1. Education and awareness This process often begins with sys- tematic reviews and summaries of the scientific literature on a topic, such as those on diet, nutrition, and cancer conducted by the World Cancer Research Fund/American Institute for Cancer Research [4], WHO [5], or national governments [6]. These are often accompanied by dietary guidelines based on the available evidence, and should ap- propriately consider all health out- comes simultaneously. Although the guidelines are usually developed by committees and are presumably bet- ter than judgements of individuals, collective biases and external influ- ences can lead to recommendations that are not optimal. Thus, dietary guidelines are now being evaluated by determining whether adherence to them is actually associated with lower risk of cancer or other health outcomes [7]; this is a practice that should continue, to optimize our guidance. Although education alone often has modest impacts on behav- iours, it is fundamentally important, in part because it can provide the foundation of support for more inten- sive policies, such as taxation.
2. Food and menu labelling Labelling is currently a topic of much debate and research that re- quires integration of nutritional and behavioural sciences; the effects can be mediated by changes in con- sumer choices and by motivating food suppliers to reformulate prod- ucts or modify serving sizes. Given the importance of overweight and obesity, labelling of energy (caloric) content has been a major focus; the impacts of this are not yet clear, and continued research is needed.
Diet, nutrition, and cancer: where next for public health? 433
3. Economic strategies These include taxation and subsi- dies. Increasing the prices of soda has a clear effect on consumption [8] and for this reason has been fought by the powerful beverage in- dustry. Taxes on soda can be read- ily justified because the true costs of consumption to others, such as the health consequences, have not been internalized in the price. The scientific evidence base to support soda taxation has become much more solid in the past several years, and this should be pursued vigor- ously as a public health strategy.
4. Promoting or limiting availability Limiting sales of tobacco has been an effective strategy, and this inter- vention is now becoming increas- ingly used for food and soda. In many places, soda is now no longer available in schools, and the city of Boston, USA, does not allow the sale of soda on any city property. Like tobacco, sale of soda in hospi- tals and other health-care facilities has become irrational, and is thus increasingly being eliminated in the USA. On the other hand, subsidies for whole grains, fruits, and vegeta- bles can remove an important barri- er to access for low-income popula- tions and will promote consumption.
5. Fortification Fortification has been an effective nutritional strategy to address many conditions, such as rickets, pellagra, goitre, and more recently congenital neural tube defects. Until now, this strategy has not been used specifi- cally for cancer prevention, and con- cerns have even been raised that folic acid fortification for prevention of neural tube defects may have in- creased incidence of colorectal can- cer. However, the concordance in time between increases in screening for colorectal cancer by colonosco- py and folic acid fortification almost certainly accounts for an apparent increase in incidence [9]. Other evi- dence suggests that additional folic acid may reduce incidence of colo- rectal cancer with a latency of more
than 10 years. At this time, there is not sufficient evidence to support fortification specifically for cancer prevention, but this might become an option if data suggesting a ben- efit for vitamin D in reducing cancer risk become stronger.
6. Banning We have a long regulatory history of banning specific food additives or colouring agents because of po- tential human carcinogenicity, and more recently partially hydrogen- ated oils have been banned in many countries, cities, and states because of their effects on cardiovascular disease. Banning whole foods or beverages is more difficult, and the experience of the USA with banning alcohol was notably unsuccessful. Short of absolute bans, a limit on the serving size of soda has been im- plemented in New York City and is being considered elsewhere, and a proposal to limit the amount of sugar added to beverages has been sug- gested. These strategies deserve consideration and evaluation.
Multisectoral approach for control of obesity The rapid increases in obesity glob- ally have led many organizations to develop strategies for controlling the epidemic. Because multiple factors contribute to unhealthy diets and inactivity, it is clear that interven- tions to address single aspects of the problem will have modest im- pacts, and multiple approaches will be needed. In most situations ac- tions will be needed in the following sectors: • schools and childcare settings; • health-care facilities and systems; • worksites; • the food environment (to ensure
availability and affordability of healthy food);
• the built environment (to promote physical activity);
• mass media (which is usually used to promote obesogenic foods);
• the economic sector (including taxes and subsidies, but also the analysis of cost and cost-effec- tiveness of interventions).
Strategic plans encompassing the above-mentioned sectors can be developed at almost every level, from global to national to local com- munities. The most effective level will vary depending on political re- alities; many of these actions would be most effective on a national level, but in the USA, political gridlock and the powerful influences of the food and beverage industries often make national actions impossible. Thus, progress is frequently much easier at the city or state levels, where ex- ternal influences may be less.
Indicators of effectiveness Unlike smoking, for which some can- cers are almost specific, nutritional factors are primarily associated with cancers that have many causes, so it will be difficult to evaluate effec- tiveness of interventions by declines in cancer incidence. Instead, we will usually need to evaluate interven- tions by changes in diet or activity or prevalence of overweight and obe- sity. Some progress has been docu- mented; in the USA, consumption of sugar-sweetened beverages has declined in recent years [8]. Also, in the past several years obesity rates among children have declined slightly in New York City and other cities where multilayered interven- tions have been developed [10].
Conclusions Continued research on diet, nutri- tion, and cancer is needed to ex- pand the scientific basis for future progress. However, evidence that overweight, obesity, and inactivity are causally related to cancer is suf- ficient to justify strong public health actions to reduce these hazards. These actions are further justified by the many other adverse health effects of these risk factors. Multiple and increasingly intensive strategies will be needed to reverse the obesity epidemic, and actions are needed in many sectors and at all levels of society. We have now begun to see some evidence of success, but sustained efforts will be needed for many years.
434
References 1. van Dam RM, Li T, Spiegelman D et al.
(2008). Combined impact of lifestyle factors on mortality: prospective cohort study in US women. BMJ, 337:a1440. h t t p: //d x . d o i . o r g /10 .113 6 / b m j . a14 4 0 . PMID:18796495
2. Willett WC (2013). Policy applications. In: Willett WC, ed. Nutritional Epidemiology, 3rd ed. New York: Oxford University Press, pp. 357–379.
3. Frieden TR (2010). A framework for public health action: the health impact pyramid. Am J Public Health, 100:590–595. http:// dx.doi.org/10.2105/AJPH.2009.185652 PMID:20167880
4. World Cancer Research Fund/American Institute for Cancer Research (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research.
5. Joint WHO/FAO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases (2003). Diet, Nutrition and the Prevention of Chronic Diseases: Report of a Joint WHO/FAO Expert Consultation. Geneva: WHO (WHO Technical Report No. 916).
6. U.S. Department of Agriculture and U.S. Department of Health and Human Services (2010). Dietary Guidelines for Americans, 2010. Washington, DC: U.S. Government Printing Office.
7. Chiuve SE, Sampson L, Willett WC (2011). The association between a nutri- tional quality index and risk of chronic dis- ease. Am J Prev Med, 40:505–513. http:// dx.doi.org/10.1016/j.amepre.2010.11.022 PMID:21496749
8. Brownell KD, Farley T, Willett WC et al. (2009). The public health and economic benefits of taxing sugar-sweetened bever- ages. N Engl J Med, 361:1599–1605. http:// dx.doi.o rg /10.10 5 6 / N E J M hpr 0 9 0 5723 PMID:19759377
9. Willett WC, Lenart E (2013). Folic acid and neural tube defects. In: Willett WC, ed. Nutritional Epidemiology, 3rd ed. New York: Oxford University Press, pp. 468–486. http://dx.doi.org/10.1093/acprof: oso/9780199754038.003.0020
10. Centers for Disease Control and Prevention (2011). Obesity in K-8 stu- dents - New York City, 2006-07 to 2010-11 school years. MMWR Morb Mortal Wkly Rep, 60:1673–1678. PMID:22169977
435Diet, nutrition, and cancer: where next for public health?
436
Summary • Kidney cancer is most common
in countries with higher levels of human development.
• Most kidney cancers (70%) are clear cell renal carcinomas; related tumours are papillary (10–15%), chromophobe (about 5%), and collecting duct (< 1%) renal cell carcinomas.
• Tobacco smoking is a cause of kidney cancer.
• Overweight and obesity are estab- lished risk factors. Occupational exposure to trichloroethylene causes kidney cancer.
• Hypertension and acquired cys- tic kidney disease with dialysis increase risk of renal cancer. Some renal cancer subtypes are associated with specific risk factors.
• Mutation or altered expression of the VHL gene has been ex- tensively implicated in familial and spontaneous renal cell cancer.
• PBRM1 has been proposed as a second major clear cell renal carcinoma gene, with truncat- ing mutations in about 40% of cases. Mutations in PBRM1 and BAP1 anticorrelate, and
5.9 5 ORGAN SITE
Kidney cancer Holger Moch Lawrence H. Lash (reviewer)
Ghislaine Scelo (reviewer)
PBRM1 and BAP1 apparently regulate different gene expres- sion programmes.
Most renal cancers are renal cell carcinomas, a heterogeneous class of tumours arising from different cell types within the renal parenchyma. Most are clear cell renal carcinomas (about 70% of renal cancer cases), followed by papillary (10–15%), chro- mophobe (about 5%), and collecting duct (< 1%) renal cell carcinomas. Each of these renal cell tumour subtypes has distinct genetic char- acteristics [1,2]. Other subtypes in the renal parenchyma include on- cocytoma, which is a benign tu- mour, and nephroblastoma (Wilms tumour), which occurs in children. Mesenchymal, mixed epithelial and mesenchymal, as well as other pri- mary tumours or metastases are rare. The epidemiology of cancers of the renal pelvis differs from that of the renal parenchyma. Cancers of the renal pelvis are urothelial car- cinomas and are similar to those of bladder urothelium.
Etiology Lifestyle factors Cigarette smoking causes renal cancer. Different meta-analyses con- firmed that ever-smoking increases
the risk of renal cancer compared with never-smoking [3,4]. There is also a dose-dependent increase in risk related to the number of ciga- rettes smoked per day. Risk decreas- es in the 5-year period after smoking cessation.
Using terminology adopted by the World Cancer Research Fund, there is convincing evidence that body fatness is a cause of kidney cancer. It is unlikely that coffee has a substantial effect, or that alcoholic drinks have an adverse effect, on the risk of this cancer [5]. Overweight, especially obesity, is a risk factor for renal cancer in both women and men [6]. The proportion of all cases of re- nal cancer attributable to overweight and obesity has been estimated to be about 40% in the USA and up
Fig. 5.9.1. The incidence of renal cell carcinoma increases significantly with in- creasing body mass index, the most com- mon measure of overweight and obesity.
Chapter 5.9 • Kidney cancer 437
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .9
Epidemiology
Kidney cancer • Kidney cancer is the ninth most
common cancer in men (214 000 cases) and the 14th most com- mon in women (124 000 cases) worldwide in 2012; 70% of the new cases occurred in countries with high and very high levels of human development, with 34% of the estimated new cases in Europe and 19% in North America.
• There were an estimated 143 000 deaths from kidney cancer in 2012 (91 000 in men, 52 000 in women); kidney can- cer is the 16th most common cause of death from cancer worldwide.
• The highest incidence rates are found in the Czech Republic. Elevated rates are also found in northern and eastern Europe, North America, and Australia. Low rates are estimated in much of Africa and East Asia.
• The case fatality rate is lower in highly developed countries (overall mortality-to-incidence ratio, 0.4) than in countries with low or medium levels of human development (0.5). Only 3.1% of the cases were diagnosed in Africa, but 5.7% of the deaths occurred in this region.
• Incidence and mortality rates have been increasing in many countries, across different lev- els of human development.
Map 5.9.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for kidney cancer in men, 2012.
Map 5.9.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for kidney cancer in women, 2012.
to 40% in European countries [7,8]. The mechanisms by which obesity influences renal carcinogenesis are unclear. Sex steroid hormones may affect renal cell proliferation by direct endocrine receptor-mediated effects. Obesity with the combined endocrine disorders, such as de- creased levels of sex hormone- binding globulin and progesterone,
insulin resistance, and increased levels of growth factors such as insulin-like growth factor 1 (IGF-1), may contribute to renal carcinogen- esis. Recently, a case–control study has reported a stronger association of clear cell carcinoma with obesity.
Some case–control studies sup- port a positive association between red meat intake and risk of renal
cancer. The elevated risk of renal cancer is presumably due to the fat and protein content of meat. Several case–control studies found a protec- tive effect of vegetables and fruits. However, the associations of meat and fruit with increased or reduced renal cancer risk are controversial [9]. There are studies showing an inverse association between alcohol intake
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
438
Chart 5.9.1. Estimated global number of new cases and deaths with proportions by major world regions, for kidney cancer in both sexes combined, 2012.
Chart 5.9.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for kidney cancer in men, circa 1975–2012.
Chart 5.9.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for kidney cancer in women, circa 1975–2012.
Chapter 5.9 • Kidney cancer 439
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .9
and renal cancer risk [10]. Recent analyses of 13 prospective studies from North America and Europe found no association between intakes of red meat, processed meat, poultry, or seafood and renal cell carcinoma risk [9,11]. No histological subtype dif- ferences were observed for associa- tions with smoking, hypertension, or family history of kidney cancer [12].
Predisposing medical condi- tions, use of pharmaceutical drugs, and hormonal and environmental factors Hypertension or its treatment has been associated with risk of renal cancer [13]. Use of hypertensive medication, including diuretics, has been associated with an elevated risk. The associations between risk of renal cancer and hypertension are independent of obesity. The biological mechanism for the asso- ciation is not known. A few cohort studies have reported a significantly increased incidence of renal cancer among diabetic patients. However, diabetes may not be an independent risk factor because of its strong rela- tion to obesity and hypertension [14]. Elevated levels of growth factors and growth factor receptors and insulin may mediate the possible relation- ship between diabetes and renal cancer.
Acquired cystic kidney disease usually develops in patients on long- term haemodialysis due to end-stage renal disease. The incidence of renal cancer is reported to be markedly increased in patients with end-stage disease (3–7%) [15]. Renal cancer occurring in end-stage renal disease has specific characteristics different from those of classic renal cancer. Papillary renal cell carcinoma was believed to be the most common subtype in end-stage renal disease. Currently, renal cancer associated with acquired cystic kidney disease is regarded as its own histological subtype, but all other subtypes (clear cell, papillary, and chromophobe re- nal cell cancer) also occur in cystic and non-cystic end-stage kidneys. Potentially, renal hyperplastic cysts
are the precursor lesions of some of these renal tumour types [16].
Phenacetin-containing analge- sics are involved in the develop- ment of urothelial cancer of the renal pelvis. Several studies have found also an elevated risk of re- nal cancer with long-term use of phenacetin [17]. Among other an- algesics, a few studies have found a positive association between acetaminophen – a metabolite of phenacetin – and renal cancer risk. There is also a potential risk of developing renal cancer for persons using non-aspirin, non-steroidal anti- inflammatory drugs (e.g. ibuprofen and naproxen). Aspirin and acet- aminophen use were not associated with renal cell carcinoma risk [18].
Hormone-related factors have been suggested to play a role in renal cancer development, but several studies found no associa- tions with use of oral contraceptives or hormone replacement therapy. Hysterectomy has been associ- ated with increased risk of renal cancer in case–control studies [19]. The possible role of reproduc- tive factors in renal cancer etiology remains poorly understood; cer- tain hormone-related factors are
associated with risk, but these as- sociations cannot explain the lower incidence in women than in men.
Studies in Taiwan (China) and Chile have also linked arsenic in drinking-water with cancer of the kidney.
Occupation Renal cell cancer is not considered an occupation-related cancer. There are associations with asbestos, gas- oline and other petroleum products, lead, cadmium, and trichloroethy- lene. Trichloroethylene is a solvent that has been widely used as a metal degreaser and chemical additive [20]. A meta-analysis of the associa- tion between exposure to trichloro- ethylene and clear cell renal cancer reported significant relative risks of kidney cancer: 1.3 overall and 1.6 for high-exposure groups [21]. IARC has classified trichloroethylene as Group 1 (carcinogenic to humans) on the basis of causing kidney cancer; the determination involved relevant epidemiological data, together with carcinogenicity of the solvent in both rats and mice. Biotransformation of trichloroethylene, well character- ized in humans and animals, occurs
Fig. 5.9.2. A boy collects rainwater for drinking and cooking near an arsenic-contam- inated tube well in a village in Rajbari District, Bangladesh. Arsenic in drinking-water causes cancers of the bladder, lung, and skin; a positive association has been observed between such exposure to arsenic and cancers of the kidney, liver, and prostate.
440
primarily through oxidative metabo- lism by cytochrome P450 enzymes, resulting in the generation of geno- toxic metabolites [22].
Pathology Clear cell renal cell carcinomas have a very vascular tumour stroma, frequently resulting in haemorrhagic areas [2]. The typical yellow tumour surface is due to the lipid content of the cells; cholesterol, neutral lipids, and phospholipids are also abundant (Figs 5.9.3, 5.9.4). Most renal cell carcinomas have little inflammatory response, but sometimes an intense lymphocytic or neutrophilic infiltrate with natural killer cells is present, and there is an association between a strong lymphocytic infiltration and worse outcome. Clear cell renal cell carcinomas most commonly metastasize haematogenously via the vena cava primarily to the lung.
Papillary renal cell carcinomas are characterized by epithelial cells forming papillae and tubules. Chromophobe renal cell carcinoma is distinguished by large polygonal cells with reticulated cytoplasm and
prominent cell membranes. Some cells are irregular and multinucle- ated. Perinuclear halos are com- mon. Collecting duct renal cell car- cinoma is extremely aggressive, with frequent metastasis already at diagnosis. These tumours are usu- ally located in the central region of the kidney. Histologically, collecting duct cancer is characterized by a tubulopapillary architecture with a characteristic desmoplastic stroma reaction.
Genetics Heritable tumours Although most renal carcinomas are sporadic, 2– 4% have a familial cause. Several genetic diseases are associated with renal cancer [1]. The risk of renal cancer for a first-degree relative of a patient with renal cancer is increased about 2-fold. Each of the common his- tological subtypes of renal cancer has a corresponding familial cancer syndrome. Table 5.9.1 provides a complete list of syndromes known to be associated with kidney can- cer risk. The most common ones include von Hippel–Lindau (VHL) syndrome, hereditary papillary renal
cell carcinoma syndrome, heredi- tary leiomyomatosis and renal cell carcinoma syndrome, and Birt– Hogg–Dubé syndrome.
In VHL syndrome, patients de- velop haemangioblastoma of the central nervous system, retinal angiomas, and phaeochromocy- tomas. Renal cancer is found in 40–50% of VHL mutation carriers. The VHL gene produces a protein that targets hypoxia-inducible fac- tor, a transcription factor. Loss of function leads to accumulation of hypoxia-inducible factor and sub- sequent upregulation of vascular endothelial growth factor and other factors that promote angiogenesis and tumour growth. The risk of de- veloping kidney cancer associated with different germline VHL alleles correlates with the degree to which their protein products are impaired in their ability to regulate hypoxia- inducible factor.
Hereditary papillary renal cell carcinoma is caused by activa- tion of the c-MET proto-oncogene. Mutations lead to constitutive activa- tion of the receptor, which promotes tumour growth. This syndrome is characterized by multifocal papil- lary renal cell carcinomas. Patients with hereditary leiomyomatosis and
Fig. 5.9.3. Clear cell renal cell carcinoma. Typical cross-section of yellowish, spheri- cal neoplasm in the upper pole of the kid- ney. Note the tumour in the dilated throm- bosed renal vein (arrow). In addition, there is a benign renal angiomyolipoma in the lower pole with a characteristic yellow cut surface due to fat cells (asterisk).
Fig. 5.9.4. Microscopic appearance of clear cell renal cell carcinoma.
Chapter 5.9 • Kidney cancer 441
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .9
Table 5.9.1. Hereditary renal cell tumours
Syndrome Gene (chromosome) Protein Tumour type Extrarenal manifestations
Dermis Other organs
von Hippel–Lindau syndrome
VHL (3p25) pVHL Multiple, bilateral clear cell RCC, renal cysts
– Haemangioblastoma of retina/central nervous system, phaeochromocytoma, pancreatic/renal cysts, neuroendocrine tumours, epididymal/ parametrial cysts, tumours of the inner ear
Hereditary papillary RCC syndrome
c-MET (7p31) HGFR Multiple, bilateral papillary RCC (type 1)
– –
Hereditary leiomyomatosis and RCC syndrome
FH (1q42) FH Papillary RCC (non- type 1)
Leiomyoma Uterine leiomyoma/ leiomyosarcoma
Familial papillary thyroid carcinoma
? (1q21) ? Papillary RCC, oncocytomas
Papillary thyroid carcinoma
Hyperparathyroidism– jaw tumour syndrome
HRPT2 (1q25) Epithelial–stromal mixed tumours, papillary RCC
– Tumours of the parathyroid glands, fibro-osseous jaw tumours
Birt–Hogg–Dubé syndrome
BHD (17p11) Folliculin Multiple chromophobe RCC, oncocytomas, papillary RCC
Facial fibrofolliculoma
Pulmonary cysts, spontaneous pneumothorax
Tuberous sclerosis TSC1 (9q34) Hamartin Multiple, bilateral angiomyolipomas, lymphangioleiomy- omatosis
Angiofibroma, subungual fibroma
Cardiac rhabdomyo- ma, adenomatous pol- yps of the small intes- tine, pulmonary/renal cysts, cortical tubers, subependymal giant cell astrocytomas
TSC2 (16p13) Tuberin Rare clear cell RCC
Constitutional chromosome 3 translocations
? (3p13-14) ? Multiple, bilateral clear cell RCC
– –
RCC, renal cell carcinoma.
renal cell carcinoma syndrome fre- quently have cutaneous leimyomas, and women often have a history of hysterectomy for uterine leimyomas at an early age. These patients show specific papillary renal cell carcino- mas. In this syndrome, the FH gene shows loss-of-function mutations.
Birt–Hogg–Dubé syndrome is caused by germline loss-of-function mutations of the BHD gene and is characterized by fibrofolliculomas, lung cysts, and a spectrum of renal carcinomas including chromophobe renal cell carcinomas and oncocyto- mas. This syndrome has been cor- related also with mutations in the folliculin tumour suppressor gene.
Spontaneous tumours In sporadic renal cell carcinoma, there is a genotype–phenotype correlation. Clear cell renal cell car- cinoma is characterized by a loss of chromosome 3p and alterations of the VHL tumour suppressor gene by mutation or hypermethylation. pVHL is a multifunctional protein and a substrate recognition subunit of a ubiquitin ligase complex that binds directly to hypoxia-inducible factor. Papillary renal cell carcino- mas demonstrate frequent trisomy or polysomy of chromosome 7 and 17. Interestingly, activating mutations of the c-MET oncogene on chromo- some 7 are rare in the sporadic forms of papillary renal cell cancer.
Chromophobe renal cell cancer is characterized by multiple cytogenet- ic changes, including monosomies of chromosomes 1, 2, 5, 10, 13, 17, and 21.
The role of a particular VHL mu- tation in trichloroethylene-induced malignancy has been highlighted. Model studies suggest that the mutation, P81S, may result in cells less likely to initiate apoptosis in response to a range of stimuli, and hence provide a pro-tumour benefit [23].
Results of exome sequenc- ing are becoming available. One study based on 101 cases identi- fied inactivating mutations in two genes encoding enzymes involved
442
in histone modification – SETD2, a histone H3 lysine 36 methyltrans- ferase, and JARID1C (also known as KDM5C), a histone H3 lysine 4 demethylase – as well as mutations in UTX (KMD6A), a histone H3 ly- sine 27 demethylase [24]. A separate investigation resulted in the identi- fication of the SWI/SNF chromatin remodelling complex gene PBRM1 as a second major clear cell renal carcinoma cancer gene, with trun- cating mutations in 92 of 227 cases (41%) (Fig. 5.9.5) [25]. More recently, whole-genome and exome sequenc- ing and other investigations have identified several putative two-hit tumour suppressor genes, includ- ing BAP1 [26]. The BAP1 protein, a nuclear deubiquitinase, is inacti- vated in 15% of clear cell renal cell carcinomas. Mutations in BAP1 and PBRM1 are inversely correlated in tumours; BAP1 and PBRM1 regulate seemingly different gene expression programmes.
Recently, a survey of more than 400 clear cell renal cell carcinomas using different genomic platforms identified 19 significantly mutated genes, of which the eight most ex- treme were VHL, PBRM1, SETD2, KDM5C, PTEN, BAP1, MTOR, and TP53. The PI3K/AKT pathway was recurrently mutated. Widespread
DNA hypomethylation was associ- ated with mutation of SETD2; muta- tions involving the SWI/SNF chro- matic remodelling complex were also implicated [27].
Kidney cancer has been charac- terized as a metabolic disease be- cause many of the known genes as- sociated with kidney cancer – VHL, MET, FLCN, TSC1, TSC2, TFE3, TFEB, MITF, FH, SDHB, SDHD, and PTEN – are involved in the cell’s abil- ity to sense oxygen, iron, nutrients, or energy. Understanding the meta- bolic basis of kidney cancer may en- able the development of novel forms of therapy for this disease [28].
Prospects The main avoidable causes of kid- ney cancer include cigarette smok- ing, obesity, and hypertension, rep- resenting opportunities for primary prevention [29]. Recent studies have shown associations between spe- cific renal tumour subtypes and es- tablished risk factors, for example end-stage kidney disease or specific genetic syndromes. Based on a spe- cific molecular background, novel tu- mour entities will be recognized as new distinct epithelial tumours within the future WHO classification system [2]. Further reports of these entities
are required to better understand the nature and behaviour of rare and highly unusual tumours. The num- ber of biomarkers in renal cell car- cinoma is extensive and can range from diagnostic biomarkers, aiding in the classification of different tumour entities, to prognostic biomarkers for use as tools to stratify patient cohorts [30]. Systemic treatment of renal cell carcinoma has changed in the past years, with the development of new agents that target complex molecular pathways regulating tu- mour angiogenesis, cell prolifera- tion, and survival [31]. Predictive biomarkers allow the choice of the appropriate therapy in patients who present with advanced disease. They provide information on which patients will be responsive or resis- tant to a targeted therapy. However, there are currently no available renal cancer biomarkers with enough reli- ability for treatment prediction [32]. Potential candidates for novel prog- nostic/predictive biomarkers may include microRNA profiles, different proteins, and/or multigene/protein assays. More novel molecular analy- sis is being used to identify specific molecular pathways involved in vari- ous tumour types and may identify potential new therapeutic targets.
Fig. 5.9.5. PBRM1 somatic mutations. Representation of PBRM1 transcript; boxes indicate the positions of the bromodomains 1–6 (BD1–BD6), the bromo-adjacent homology domains (BAH1, BAH2), and the high-mobility group domain (HMG). Relative positions of mutations are indicated by symbols: nonsense mutations (stars), missense mutations (dots), frameshift deletions (red triangles), frameshift insertions (black triangles), and in-frame deletions (green triangles). Splice-site mutations are not depicted.
Chapter 5.9 • Kidney cancer 443
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .9
References
1. Eble JN, Sauter G, Epstein JI, Sesterhenn IA, eds (2004). Tumours of the kidney. In: Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. Lyon: IARC.
2. Moch H (2013). An overview of renal cell cancer: pathology and genetics. Semin Cancer Biol, 23:3–9. http://dx.doi. org/10.1016/j.semcancer.2012.06.006 PMID:22722066
3. Cho E, Adami HO, Lindblad P (2011). Epidemiology of renal cell cancer. Hematol Oncol Clin North Am, 25:651–665. http:// dx.doi.org/10.1016/j.hoc.2011.04.002 PMID: 21763961
4. Hunt JD, van der Hel OL, McMillan GP et al. (2005). Renal cell carcinoma in relation to cigarette smoking: meta-analysis of 24 stud- ies. Int J Cancer, 114:101–108. http://dx.doi. org/10.1002/ijc.20618 PMID:15523697
5. World Cancer Research Fund/American Institute for Cancer Research (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research.
6. Ljungberg B, Campbell SC, Choi HY et al. (2011). The epidemiology of renal cell carcinoma. Eur Urol, 60:615–621. http:// dx.doi.org/10.1016/j.eururo.2011.06.049 PMID:21741761
7. Renehan AG, Soerjomataram I, Tyson M et al. (2010). Incident cancer burden at- tributable to excess body mass index in 30 European countries. Int J Cancer, 126: 692–702. http://dx.doi.org/10.1002/ijc.24 803 PMID:19645011
8. Renehan AG, Tyson M, Egger M et al. (2008). Body-mass index and incidence of cancer: a systematic review and meta- analysis of prospective observational stud- ies. Lancet, 371:569–578. http://dx.doi.org/ 10.1016/S0140-6736(08)60269-X PMID: 18280327
9. Lee JE, Männistö S, Spiegelman D et al. (2009). Intakes of fruit, vegetables, and ca- rotenoids and renal cell cancer risk: a pooled analysis of 13 prospective studies. Cancer Epidemiol Biomarkers Prev, 18:1730–1739. http://dx.doi.org/10.1158/1055-9965.EPI- 09-0045 PMID:19505906
10. Lee JE, Hunter DJ, Spiegelman D et al. (2007). Alcohol intake and renal cell can- cer in a pooled analysis of 12 prospective studies. J Natl Cancer Inst, 99:801–810. http://dx.doi.org/10.1093/jnci/djk181 PMID: 17505075
11. Lee JE, Spiegelman D, Hunter DJ et al. (2008). Fat, protein, and meat consump- tion and renal cell cancer risk: a pooled analysis of 13 prospective studies. J Natl Cancer Inst, 100:1695–1706. http://dx.doi. org/10.1093/jnci/djn386 PMID:19033572
12. Purdue MP, Moore LE, Merino MJ et al. (2013). An investigation of risk factors for renal cell carcinoma by histologic subtype in two case-control studies. Int J Cancer, 132:2640–2647. http://dx.doi.org/10.1002/ ijc.27934 PMID:23150424
13. Weikert S, Boeing H, Pischon T et al. (2008). Blood pressure and risk of renal cell carcinoma in the European prospective investigation into cancer and nutrition. Am J Epidemiol, 167:438– 446. http://dx.doi. org/10.1093/aje/kwm321 PMID:18048375
14. Schlehofer B, Pommer W, Mellemgaard A et al. (1996). International renal-cell- cancer study. VI. The role of medical and family history. Int J Cancer, 66:723–726. ht tp: //dx.doi.org /10.10 02 /(SICI)10 97- 0215(19960611)66:6<723::AID-IJC2>3.0. CO;2-1 PMID:8647639
15. Denton MD, Magee CC, Ovuworie C et al. (2002). Prevalence of renal cell carcinoma in patients with ESRD pre-transplantation: a pathologic analysis. Kidney Int, 61:2201– 2209. http://dx.doi.org/10.1046/j.1523-1755. 2002.00374.x PMID:12028461
16. Montani M, Heinimann K, von Teichman A et al. (2010). VHL-gene deletion in single renal tubular epithelial cells and renal tu- bular cysts: further evidence for a cyst-de- pendent progression pathway of clear cell renal carcinoma in von Hippel-Lindau dis- ease. Am J Surg Pathol, 34:806–815. http:// dx.doi.org/10.1097/PAS.0b013e3181ddf54d PMID:20431476
17. McCredie M, Pommer W, McLaughlin JK et al. (1995). International renal-cell cancer study. II. Analgesics. Int J Cancer, 60:345–349. http://dx.doi.org/10.1002/ijc. 2910600312 PMID:7829242
18. Cho E, Curhan G, Hankinson SE et al. (2011). Prospective evaluation of analgesic use and risk of renal cell cancer. Arch Intern Med, 171:1487–1493. http://dx.doi.org/ 10.1001/archinternmed.2011.356 PMID: 21911634
19. Lindblad P, Mellemgaard A, Schlehofer B et al. (1995). International renal-cell cancer study. V. Reproductive factors, gynecolog- ic operations and exogenous hormones. Int J Cancer, 61:192–198. http://dx.doi. org/10.1002/ijc.2910610209 PMID:7705947
20. Kelsh MA, Alexander DD, Mink PJ, Mandel JH (2010). Occupational trichloroethylene exposure and kidney cancer: a meta-analy- sis. Epidemiology, 21:95–102. http://dx.doi. org/10.10 97/EDE.0b013e3181c 3 0e92 PMID:20010212
21. Scott CS, Jinot J (2011). Trichloroethylene and cancer: systematic and quantita- tive review of epidemiologic evidence for identifying hazards. Int J Environ Res Public Health, 8:4238– 4272. http:// dx.doi.org/10.3390/ijerph8114238 PMID: 22163205
22. Guha N, Loomis D, Grosse Y et al.; International Agency for Research on Cancer Monograph Working Group (2012). Carcinogenicity of trichloroethylene, tetra- chloroethylene, some other chlorinated sol- vents, and their metabolites. Lancet Oncol, 13:1192–1193. http://dx.doi.org/10.1016/S14 70-2045(12)70485-0 PMID:23323277
23. Desimone MC, Rathmell WK, Threadgill DW (2013). Pleiotropic effects of the trichloro- ethylene-associated P81S VHL mutation on metabolism, apoptosis, and ATM-mediated DNA damage response. J Natl Cancer Inst, 105:1355–1364. http://dx.doi.org/10.1093/ jnci/djt226 PMID:23990666
24. Dalgliesh GL, Furge K, Greenman C et al. (2010). Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes. Nature, 463:360–363. http://dx.doi.org/10.1038/nature08672 PMID:20054297
25. Varela I, Tarpey P, Raine K et al. (2011). Exome sequencing identifies frequent mu- tation of the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature, 469:539–542. http://dx.doi.org/10.1038/nature09639 PMID:21248752
26. Peña-Llopis S, Vega-Rubín-de-Celis S, Liao A et al. (2012). BAP1 loss defines a new class of renal cell carcinoma. Nat Genet, 44:751–759. http://dx.doi.org/10.1038/ng. 2323 PMID:22683710
27. Creighton CJ, Morgan M, Gunaratne PH et al.; Cancer Genome Atlas Research Network (2013). Comprehensive molecular characterization of clear cell renal cell car- cinoma. Nature, 499:43– 49. http://dx.doi. org/10.1038/nature12222 PMID:23792563
28. Linehan WM, Ricketts CJ (2013). The meta- bolic basis of kidney cancer. Semin Cancer Biol, 23:46–55. http://dx.doi.org/10.1016/j. semcancer.2012.06.002 PMID:22705279
29. Weikert S, Ljungberg B (2010). Contemporary epidemiology of renal cell carcinoma: per- spectives of primary prevention. World J Urol, 28:247–252. http://dx.doi.org/10.1007/ s00345-010-0555-1 PMID:20390283
30. Eichelberg C, Junker K, Ljungberg B, Moch H (2009). Diagnostic and prognostic mo- lecular markers for renal cell carcinoma: a critical appraisal of the current state of research and clinical applicability. Eur Urol, 55:851–863. http://dx.doi.org/10.1016/j.eur uro.2009.01.003 PMID:19155123
31. Fisher R, Gore M, Larkin J (2013). Current and future systemic treatments for renal cell carcinoma. Semin Cancer Biol, 23:38– 45. http://dx.doi.org/10.1016/j.semcancer. 2012.06.004 PMID:22705280
32. Algaba F, Akaza H, López-Beltrán A et al. (2011). Current pathology keys of renal cell carcinoma. Eur Urol, 60:634–643. http://dx. doi.org/10.1016/j.eururo.2011.06.047 PMID: 21741159
444
Summary • Bladder cancer is the ninth most
common cancer worldwide, and urothelial carcinoma is the most frequent histological type (> 90%). About 70–80% of patients are diagnosed with non-invasive and low-grade tumours.
• Tobacco smoking remains the most important cause of blad- der cancer. Arsenic and some occupational exposures also cause bladder cancer. Chronic infection and inflammation asso- ciated with Schistosoma haema- tobium is a cause of squamous cell carcinoma of the bladder.
• Non-invasive and invasive cancers show distinct genetic profiles, reflecting different mo- lecular pathways of develop- ment. Low-grade non-invasive tumours frequently show acti- vating mutations of the FGFR3 gene, which encodes fibroblast growth factor receptor 3, and inactivation of multiple tumour suppressors at chromosome 9. High-grade and invasive dis- ease is accompanied by a high degree of genetic instability, aneuploidy, and alterations of a multitude of tumour suppressor genes and oncogenes, often in- cluding TP53.
5.10 5 ORGAN SITE
Bladder cancer Guido Sauter Mahul B. Amin (reviewer)
Manolis Kogevinas (reviewer) Ronald Simon (contributor)
• Despite improvement in diagno- sis and therapy, the likelihood of recurrence and progression to invasively growing tumours must be recognized. As many as 50–70% of these tumours recur, and about 10–20% progress to muscle-invasive disease. Only 30–40% of patients with mus- cle-invasive cancers survive 5 years or longer.
There are five major histological types of bladder cancer, including non-invasive and invasive urothelial carcinoma, as well as squamous cell carcinoma, adenocarcinoma, and small cell carcinoma. All sub- types develop from the urothelium and show characteristic morphologi- cal features. Invasive cancers often show mixed features including more than one phenotype.
Etiology The most important known cause of bladder cancer is tobacco smok- ing [1]. The risk of developing blad- der cancer is increased 2–6-fold in smokers compared with never- smokers. Black tobacco smoke (com- mon in southern Europe and Latin America) contains higher levels of N-nitrosamines and aromatic amines and poses a risk that is about twice
that posed by the blond tobacco that is typically smoked in the USA and northern Europe. Tobacco smoke contains aromatic amines such as 4-aminobiphenyl and 2-naphthyl- amine, which are human bladder carcinogens in the context of occu- pational exposure.
Arsenic contamination in drink- ing-water is proven to cause bladder
Fig. 5.10.1. Villagers surround a shallow tube well in a village in Comilla District, Bangladesh. The pump is marked with red paint to indicate arsenic-contaminated drinking-water.
Chapter 5.10 • Bladder cancer 445
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
0
Epidemiology
Bladder cancer • With an estimated 430 000
new cases and 165 000 deaths in 2012, bladder cancer is the ninth most frequent cancer worldwide (in both sexes) and the 13th most common cause of cancer death.
• Incidence and mortality rates are elevated in North America, Europe, North Africa, the Middle East, and Australia and New Zealand. Rates tend to be rather low in many African and Asian countries and some Latin American countries.
• Incidence rates vary 10-fold between geographical regions, and 72% of the new cases oc- curred in countries that have at- tained high or very high levels of human development. Incidence and mortality rates in men are commonly 2–4 times those seen in women.
• Trends in bladder cancer inci- dence are difficult to interpret without precise information on how cancer registries have dealt with papillomas over time. There have generally been declines in incidence and mortality in de- veloped countries, but some ris- ing trends have been observed in eastern Europe and in cer- tain countries in developmental transition.
• While smoking- and occupation- related urothelial bladder cancer tends to be the dominant form of bladder cancer in developed countries, squamous cell car- cinomas are common in the Middle East and parts of Africa, as a result of chronic infection with Schistosoma haematobium.
Map 5.10.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for bladder cancer in men, 2012.
Map 5.10.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for bladder cancer in women, 2012.
cancer, particularly in parts of South- East Asia. There is evidence that disinfection by-products – most typi- cally, chemicals generated by water chlorination – could increase the risk of the disease [2]. Bladder cancer,
along with cancers of the lung and skin, has been widely recognized as an occupational disease. Male precision metalworkers and met- alworking machine operators, au- tomobile mechanics, plumbers,
workers in electronic components manufacturing, and landscape industry workers have signifi- cantly elevated risks, increasing with the duration of employment. Historically, occupational exposure
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
446
Chart 5.10.1. Estimated global number of new cases and deaths with proportions by major world regions, for bladder cancer in both sexes combined, 2012.
Chart 5.10.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for bladder cancer in men, circa 1975–2012.
Chart 5.10.3. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for bladder cancer in women, circa 1975–2012.
Chapter 5.10 • Bladder cancer 447
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
0
to aromatic amines was established as a causative agent for much oc- cupational bladder cancer; more re- cently, specific causative agents have been difficult to identify, but exposure to metalworking fluids is strongly im- plicated [3].
Chronic inflammation of the bladder is a major risk factor, espe- cially for squamous cell carcinomas. Chronic infection with the parasitic trematode Schistosoma haemato- bium causing schistosomiasis has been associated with bladder cancer development in endemic countries, with a 2–15-fold increased risk of cancer in infected compared with non-infected subjects. Continuous mucosal irritation and chronic inflam- mation from, for example, indwelling catheters or Schistosoma infection, can cause squamous metaplasia, dysplasia, and eventually squamous cell carcinomas of the bladder. Patients with long-term paraplegia are at markedly elevated risk for this tumour type due to chronic bladder infection.
Recent genome-wide association studies have identified multiple chro- mosomal regions of which alterations are linked to the risk of developing bladder cancer [4]. Polymorphisms of the NAT2 and GSTM1 metabolic genes have emerged as the most consistent findings; the NAT2 slow acetylator phenotype is linked with a 2–3-fold increased risk of bladder cancer in former or current heavy smokers, and with a 1.3–1.5-fold in- creased risk in patients with inacti- vation of one of two GSTM1 alleles. Other recurrently identified candidate
susceptibility loci include the urea transporter SLC14A1 and the pros- tate stem cell antigen (PSCA) gene.
Pathology and genetics The most important element in pathological evaluation of bladder cancer is recognition of the pres- ence and extent of invasion. Non- invasive tumours are restricted to the urothelial cell layers and do not penetrate the lamina pro- pria. The current classification of non-invasive neoplasms of the bladder includes four groups: (i) papillary urothelial neoplasms of low malignant potential, which are
characterized by an excellent prog- nosis and often not identified as “cancer”, (ii) non-invasive papillary urothelial carcinomas (low-grade), (iii) non-invasive papillary urothelial carcinomas (high-grade), and (iv) flat carcinoma in situ. Invasive tu- mours are subdivided according to the depth of invasion.
The nomenclature of bladder cancer was previously confusing because non-invasive and minimally invasive bladder tumours had been combined into the subgroup of “su- perficial bladder cancers”, which was distinguished from the subgroup of “invasive bladder cancers”, a term
Fig. 5.10.2. Urothelial carcinoma of the bladder. Transurethral biopsy.
Fig. 5.10.3. Putative model of bladder cancer development and progression, based on genetic findings. Thick arrows indicate the most frequent pathways, and dotted lines the rarest events. CIS, carcinoma in situ; NIHGC, non-invasive papillary urothelial carcinoma (high-grade); NILGC, non-invasive papillary urothelial carcinoma (low-grade); pT1, primary tumour, stage 1; pT2-4 primary tumour, stage 2–4; PUNLMP, papillary urothelial neoplasm of low malignant potential.
448
restricted to muscle-invasive tu- mours. Most non-high-grade non- invasive papillary bladder carcinomas rarely progress to invasive cancer, and most invasive bladder cancers are not preceded by a non-invasive papillary carcinoma.
Bladder urothelial carcinoma These tumours constitute more than 90% of invasive bladder cancers and can exhibit either a papillary or a solid growth pattern. There is no single characteristic morphological or immunohistochemical feature that is specific for urothelial carcinoma. Therefore, unequivocal diagnosis of urothelial carcinoma relies on iden- tification of precursor lesions such as carcinoma in situ. The histology of invasive urothelial carcinoma is highly variable. These carcinomas often contain areas of divergent differentiation.
Urothelial carcinoma consti- tutes two genetically distinct sub- groups with marked differences in their degree of genetic instability (Fig. 5.10.3). The first category in- cludes the genetically stable tumours: papillary urothelial neoplasms of low malignant potential, non-invasive papillary urothelial carcinomas (low- grade), and probably more than half of non-invasive papillary urothelial carcinomas (high-grade). The sec- ond category (genetically unstable cancers) includes carcinoma in situ, about half of non-invasive papillary urothelial carcinomas (high-grade), and invasively growing cancers. Non- invasive papillary urothelial carcino- mas account for 70–80% of all blad- der cancers and are characterized by a papillary tumour growth pattern and cells showing mild or moderate atypia.
Papillary urothelial neoplasms of low malignant potential and non-inva- sive papillary urothelial carcinomas (low-grade) are characterized by few genomic alterations, typically includ- ing partial or complete deletions of chromosome 9 as well as mutation of the FGFR3 growth factor receptor [5,6]. Chromosome 9 losses occur in about 50% of tumours, and at compa- rable frequency also in tumours of all
other grades and stages. Since chro- mosome 9 deletions are also found in hyperplasia and normal-appearing urothelium, chromosome 9 aberra- tions are seen as early key events in bladder cancer development. The frequent loss of the entire chromo- some 9 suggests that multiple tumour suppressor genes on 9p and 9q might be jointly inactivated by one genetic event. Two tumour suppressor genes have been identified at chromosome 9p21: the cell-cycle control genes CDKN2A (p16/p14ARF) and CDKN2B (also known as p15INK4B or p15).
Additional tumour suppressor loci have been mapped to 9q, including the tuberous sclerosis 1 (TSC1) gene at 9q34 and the deleted in bladder cancer 1 (DBC1) gene at 9q32-q33. Deletion and mutation of TSC1 have been reported to occur in more than 50% and in 15% of bladder cancers, and inactivation of this gene results in the loss of mTOR-mediated control of cell growth. The putative tumour suppressor DBC1 is often affected by homozygous deletions, although hypermethylation has been suggest- ed as one of the earliest events in bladder cancer development. FGFR3 mutations occur in 80–90% of papil- lomas. In neoplastic lesions, the frac- tion of FGFR3 mutations is inversely correlated with tumour stage and grade: FGFR3 mutations are pres- ent in more than 60% of low-grade non-invasive tumours, 35% of high- grade non-invasive tumours, 25% of minimally invasive (pT1) cancers, and 16% of muscle-invasive (pT2–4) cancers [7]. Gene amplifications and TP53 mutations are rare. DNA aneuploidy occurs in less than 50% of cases.
Invasive urothelial carcinomas differ markedly from non-invasive low-grade cancers. They are ge- netically unstable, often exhibiting high-level amplifications and TP53 mutations. Aneuploidy is strongly as- sociated with tumour stage and grade and is found in more than 90% of cas- es. Earlier studies reported frequent deletions involving chromosomes 2q, 5q, 8p, 9p, 9q, 10q, 11p, and 18q and the Y chromosome, as well as gains of 1q, 5p, 8q, and 17q. In addition,
high-level amplifications are frequent in these cancers. The most frequent amplification sites occurring in more than 10% of high-grade/invasive can- cers include 11q13, the locus of the cyclin-dependent kinase 1 (CCND1) gene; 6p22, with the transcription factor gene E2F3; and 17q21, which harbours the human epidermal growth factor receptor 2 (HER2) gene (Table 5.10.1).
Methylation studies have re- vealed – as for other cancers – that invasive bladder cancers exhibit a plethora of differentially methylated areas with hypermethylation in pro- moter regions and hypomethylation in the gene body [8,9]. Genes sus- pected to be dysregulated through methylation include 14-3-3 sigma, SYK, CAGE-1, PTEN, FOXO1, MAPK1, and PDK1, as well as vari- ous HOX family transcription factors, inactivation of which has been linked to aggressive forms of the disease. Several microRNAs are differentially expressed in urothelial carcinomas, including miR-10a, miR-21, miR-30b,
Fig. 5.10.4. An outdoor leather tannery in Fez, Morocco. Men work in the vats curing and dying leather from camels, cattle, and goats. Skins are first dipped in lye, then washed and placed in sulfuric acid. The solvents used in the tanning process have been associated with an increased risk of bladder cancer.
Chapter 5.10 • Bladder cancer 449
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
0
Table 5.10.1. The most frequent chromosomal aberrations found in larger studies of bladder cancer (≥ 50 tumours analysed)
Aberration Chromosome arm Non-invasive Invasive
(pT1– 4)
Peak of gain (amplification) or deletiona
Candidate gene(s)bLow-grade High-grade
Gains 1q 13% 17% 11% 1q21.2 (35%) SETDB1
2p 8–30% 10% 2p25 (10%)
3p 10% 3p25 (10–12%) RAF1
3q 1% 5% 10–25% 3q26 (10%) PIK3CA
4p 5% 4p16 (5%)
5p 2% 28% 25% 5p15 (1–25%) TRIO, SKP2
6p 1–5% 28% 16–24% 6p22 (3–25%) E2F3
7 5% 5% 10–20%
8p 8p11-p12 (2– 4%) FGFR1
8q 5–10% 20% 20–50% 8q21 (4–7%) 8q24 (1–33%)
TPD52 MYC
10p 3% 5% 10–20% 10p11-p14 (2–15%) MAP3K8
11q 5% 15% 20% 11q13 (10%) CCND1
12p 1% 5% 5%
12q 1–15% 5% 5–30% 12q13-q15 (3–5%) MDM2 CDK4
16p 5% 16p13 (5%)
16q 5% 16q22 (5%)
17q 10–30% 33% 10–50% 17q12-q21 (10–20%) HER2, TOP2A
18p 10% 18p11 (1–10%) YES1
19q 10% 19q13 (2%) CCNE
20q 7–15% 33% 20–30% 20q11 (25%) BCL2L1
Deletions 2q 4–5% 39% 17–30% 2q36 (25%) CUL3
4p 2–5% 22% 10–30%
4q 1–10% 17% 10–30%
5q 4–20% 33% 16–30%
6q 1–10% 33% 19% 6q21-q24 (5%)
8p 5–15% 28% 30% 8p11-p12 (30%)
9p 36– 45% 45% 31– 47% 9p21 (30–50%) CDKN2A
9q 45% 38% 23– 47% 9q34 (30–50%) DBC1
10q 5% 28% 18–28% 10q23 (20%) 10q26 (30%)
PTEN
11p 10% 17% 24– 43%
11q 6% 23% 22–34% 11q23 (10%)
13q 0–20% 17% 19–29% 13q14 (20%) RB1
14q 1% 10% 14q23-q24 (10%)
15q 8% 15q15 (8%)
16p 8%
16q 8% 16q13-q21 (8%)
17p 1–5% 11% 19–24% 17p13 (24%) TP53
18q 7–10% 39% 13–30% 18q21 (10%)
Y 10–20% 28% 15–37%
pT1–4, primary tumour, stage 1–4. a In invasive cancers. b Candidate oncogenes for gains, candidate tumour suppressor genes for deletions.
450
miR-31, miR-100, miR-141, miR-143, mi-145, miR-192, miR-195, miR- 200a/b/c, miR-205, miR-452, and miR-708, and some of these have been linked to invasive growth or poor prognosis. Such findings em- phasize the potential of microRNA measurements for diagnostic pur- poses. Large-scale next-generation sequencing studies are lacking so far, but anecdotal reports on druggable FGFR fusions in bladder cancers have already suggested that these powerful new methods have the po- tential to significantly advance our knowledge of bladder cancer.
Only few studies have analysed genetic changes in carcinoma in situ and in those high-grade non-inva- sive papillary urothelial carcinomas with a particularly high degree of atypia (pTaG3). FGFR3 mutations, which are frequently found in papil- lomas and papillary urothelial neo- plasms of low malignant potential, are virtually absent in carcinoma in situ, whereas TP53 mutations, a hallmark of muscle-invasive bladder cancer, can frequently be found in carcinoma in situ already. The pat- tern of deletions and amplifications found was similar to that of invasive cancers [10].
Bladder squamous cell carcinoma These tumours account for about 2–5% of all bladder cancers and represent a cancer with histologi- cally pure squamous cell phenotype, sometimes accompanied by kerati- nizing squamous metaplasia in the adjacent flat epithelium. There do not appear to be major differences between muscle-invasive urothelial carcinomas and schistosomiasis-as- sociated squamous cell carcinomas. Frequent cytogenetic alterations in- clude gains of 5p, 6p, 7p, 8q, 11q, 17q, and 20q as well as deletions of 4q, 5q, 8p, 13q, 17p, and 18q. Other key alterations of urothelial carci- noma, such as mutation of TP53 or overexpression of EGFR and HER2, are also found at compara- ble frequencies in squamous cell carcinomas.
The genetics of non-schistoso- miasis-associated squamous cell carcinomas indicate alterations not markedly different from those in muscle-invasive urothelial carcino- mas. Differences may exist in ex- pression levels. One study reported differential expression of keratin-10 and caveolin-1 between the two subtypes [10], and a more recent validation of cytokeratin expression revealed a pattern of CK5/6 positiv- ity and CK5/14 positivity plus CK20 negativity and uroplakin negativity that identifies squamous cell differ- entiation [11]. A particularly high in- cidence of 14-3-3 gene methylation, but relatively infrequent methylation of the SKY gene, was reported for squamous cell carcinomas com- pared with urothelial cancers, and loss of expression of both genes has been suggested to contribute to the aggressive clinical behaviour of non- urothelial bladder tumours [12].
Bladder adenocarcinoma Adenocarcinomas constitute less than 2% of all bladder cancers and are derived from the urothelium, showing a pure glandular pheno- type. The growth pattern is variable, with enteric (colonic), signet ring cell, mucinous, clear cell, hepatoid, and mixed types.
Most adenocarcinomas found in bladder biopsies (90%) represent tumours of another origin infiltrating the bladder, such as adenocarcino- mas of the prostate, colon, or ovary. Molecular data on pure primary ad- enocarcinomas of the bladder are rare. Few cases of adenocarcinoma have been analysed cytogenetically, revealing a spectrum of chromosom- al alterations that seemed similar to that found in urothelial carcinomas. An early study on 8 schistosomiasis- associated adenocarcinomas report- ed deletions of 3p, 4p, 4q, 9p, 9q, 17p, 8p, 11p, and 18q [13]. In another series of 13 adenocarcinomas, 4 car- ried mutations of the TP53 tumour suppressor [14], an alteration that is also frequently (30–50%) found in invasive urothelial carcinomas. More recently, epigenetic analysis identi- fied frequent promoter methylation
of the 14-3-3 and CAGE-1 genes in 10 non-schistosomiasis-associated adenocarcinomas and 6 signet ring cell carcinomas [15].
Bladder small cell carcinoma These tumours, accounting for about 1% and 3% of all bladder cancers in men and women, respectively, are malignant neuroendocrine neo- plasms that histologically mimic their pulmonary counterparts. All small cell cancers are invasive at diagno- sis and typically appear as a large solid mass of small, uniform cells. About 50% of cancers have areas of urothelial carcinoma and, in rare cases, squamous cell carcinoma and/or adenocarcinoma.
Classic cytogenetic and mo- lecular cytogenetic studies report frequent gains and losses at the same chromosomal loci often af- fected in muscle-invasive urothelial carcinomas. The most frequent ab- errations include deletions at chro- mosomes 3p25-p26 (VHL), 4q, 5q, 9p21 (p16), 9q32-q33 (DBC1), 10q, 13q, and 17p13 (TP53) as well as gains at 5p, 6p, 8q, and 20q. High- level amplifications affect 1p22-p32, 3q26.3 (PIK3CA), 8q24 (MYC), and 12q14-q21 (MDM2, CDK4). In ad- dition, promoter methylation and expression studies suggested that silencing of the tumour suppres- sors RASSF1, MLH1, DAPK1, and MGMT might contribute to the malig- nant behaviour of small cell bladder cancers [16]. Recent genetic studies demonstrate transition from urothe- lial carcinoma to small cell carci- noma during the course of tumour progression. For example, identical patterns of loss of heterozygosity at five polymorphic loci and inactiva- tion of the same copy of the X chro- mosome were seen in the urothe- lial and the small cell component, suggesting a common clonal origin [17]. In addition, identical point mu- tations of TP53 were found in the invasive small cell component and in the urothelial in situ carcinoma, suggesting that small cell carcinoma might develop from in situ carcinoma of the bladder [18].
Chapter 5.10 • Bladder cancer 451
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
0
Prospects Prevention Cessation of smoking is the most effective measure for prevention of bladder cancer. Intervention in re- lation to proven or likely workplace exposures will prevent occupational bladder cancer, particularly in rela- tion to work practices in low-income countries. Avoidance of arsenic-con- taminated drinking-water should also be implemented. For squamous cell carcinoma, avoidance of infestation with Schistosoma haematobium is an effective intervention.
Screening Attempts to perform population- based screening have not been suc- cessful due to the low incidence of bladder cancers and the low speci- ficity (and sensitivity) of tests [19,20].
Targeted therapies Urothelial carcinoma is one of the most heterogeneous cancer types. Molecular heterogeneity may com- promise the efficiency of targeted drugs in this malignancy. Several drugs targeting genes frequently al- tered in bladder cancer have exhib- ited disappointing efficacy compared with other cancer types. Thus, in ad- vanced bladder cancer, addition of gefitinib to cisplatin and gemcitabine did not improve response rates or survival [21]. In another phase II study (ClinicalTrials.gov: NCT00380029), neoadjuvant administration of anoth- er EGFR inhibitor, erlotinib, showed some beneficial effects. Although mostly heterogeneous, HER2 over- expression and amplification is found in about 10–30% of advanced blad- der cancers. In a clinical trial investi- gating the effects of combining che- motherapy (paclitaxel, carboplatin, and gemcitabine) with trastuzumab, there was a 73% response rate and a median survival of 15.2 months [22]. Targeting FGFR3 – which is frequent- ly mutated, particularly in low-grade non-invasive bladder cancers – has
shown encouraging results in multi- ple myeloma and bladder cancer cell lines [23]. At present, phase II clini- cal trials are under way for several additional novel targeted agents, in- cluding bevacizumab and aflibercept, which target VEGF, and the multiple tyrosine kinase inhibitors sunitinib, sorafenib, and lapatinib.
The prevalence and co-occur- rence of actionable genomic alter- ations in patients with high-grade bladder cancer have been assessed to underpin therapeutic drug discov- ery [24]. Of the tumours examined,
61% harboured potentially actionable genomic alterations. A core pathway analysis of the integrated data set revealed a non-overlapping pattern of mutations in the RTK/RAS/RAF (Fig. 5.10.5) and PI3K/AKT/mTOR signalling pathways and regulators of G1–S cell-cycle progression. The findings suggested that optimal de- velopment of target-specific agents will require pre-treatment genomic characterization.
Fig. 5.10.5. Co-occurrence of alterations within the RTK/RAS/RAF signalling pathway in high-grade bladder cancer. Incidence (%) of amplifications, deletions, and mutations of selected receptor tyrosine kinases and downstream targets. The heat map (bottom) compares the distribution of each alteration across tumour samples.
452
References 1. Freedman ND, Silverman DT, Hollenbeck
AR et al. (2011). Association between smoking and risk of bladder cancer among men and women. JAMA, 306:737–745. http://dx.doi.org/10.1001/jama.2011.1142 PMID:21846855
2. Costet N, Villanueva CM, Jaakkola JJ et al. (2011). Water disinfection by-products and bladder cancer: is there a European specificity? A pooled and meta-analy- sis of European case-control studies. Occup Environ Med, 68:379–385. http:// dx.doi.org/10.113 6/oem. 2010.0 62703 PMID:21389011
3. Colt JS, Karagas MR, Schwenn M et al. (2011). Occupation and bladder cancer in a population-based case- control study in Northern New England. Occup Environ Med, 68:239–249. http:// dx.doi.org/10.113 6/oem. 20 0 9.052571 PMID:20864470
4. Rothman N, Garcia-Closas M, Chatterjee N et al. (2010). A multi-stage genome-wide association study of bladder cancer identi- fies multiple susceptibility loci. Nat Genet, 42:978–984. http://dx.doi.org/10.1038/ng. 687 PMID:20972438
5. Lott S, Wang M, Zhang S et al. (2009). FGFR3 and TP53 mutation analysis in in- verted urothelial papilloma: incidence and etiological considerations. Mod Pathol, 22:627–632. http://dx.doi.org/10.1038/mod pathol.2009.28 PMID:19287463
6. Castillo-Martin M, Domingo-Domenech J, Karni-Schmidt O et al. (2010). Molecular pathways of urothelial development and bladder tumorigenesis. Urol Oncol, 28:401– 408. http://dx.doi.org/10.1016/j. urolonc.2009.04.019 PMID:20610278
7. Knowles MA (2007). Role of FGFR3 in urothelial cell carcinoma: biomarker and potential therapeutic target. World J Urol, 25:581–593. http://dx.doi.org/10.1007/s00 345-007-0213-4 PMID:17912529
8. Pu RT, Laitala LE, Clark DP (2006). Methylation profiling of urothelial carci- noma in bladder biopsy and urine. Acta Cytol, 50:499–506. http://dx.doi.org/10. 1159/000326003 PMID:17017434
9. Wolff EM, Chihara Y, Pan F et al. (2010). Unique DNA methylation patterns distin- guish noninvasive and invasive urothe- lial cancers and establish an epige- netic field defect in premalignant tissue. Cancer Res, 70:8169–8178. http://dx.doi. org/10.1158/0 0 08 - 5 472.CAN -10 -1335 PMID:20841482
10. Zhao J, Richter J, Wagner U et al. (1999). Chromosomal imbalances in noninvasive papillary bladder neoplasms (pTa). Cancer Res, 59:4658– 4661. PMID:10493521
11. Gaisa NT, Braunschweig T, Reimer N et al. (2011). Different immunohistochemical and ultrastructural phenotypes of squamous differentiation in bladder cancer. Virchows Arch, 458:301–312. http://dx.doi.org/10. 1007/s00428-010-1017-2 PMID:21136076
12. Kunze E, Wendt M, Schlott T (2006). Promoter hypermethylation of the 14-3-3 sigma, SYK and CAGE-1 genes is related to the various phenotypes of urinary blad- der carcinomas and associated with pro- gression of transitional cell carcinomas. Int J Mol Med, 18:547–557. PMID:16964403
13. Shaw ME, Elder PA, Abbas A, Knowles MA (1999). Partial allelotype of schistosomiasis- associated bladder cancer. Int J Cancer, 80:656–661. http://dx.doi.org/10.1002/(SICI) 1097-0215(19990301)80:5<656::AID-IJC 4>3.0.CO;2-A PMID:10048962
14. Warren W, Biggs PJ, el-Baz M et al. (1995). Mutations in the p53 gene in schistosomal bladder cancer: a study of 92 tumours from Egyptian patients and a comparison between mutational spectra from schis- tosomal and non-schistosomal urothelial tumours. Carcinogenesis, 16:1181–1189. http://dx.doi.org/10.1093/carcin/16.5.1181 PMID:7767983
15. Kunze E, Schlott T (2007). High frequen- cy of promoter methylation of the 14-3-3 sigma and CAGE-1 genes, but lack of hy- permethylation of the caveolin-1 gene, in primary adenocarcinomas and signet ring cell carcinomas of the urinary bladder. Int J Mol Med, 20:557–563. PMID:17786288
16. Zhao X, Flynn EA (2012). Small cell car- cinoma of the urinary bladder: a rare, ag- gressive neuroendocrine malignancy. Arch Pathol Lab Med, 136:1451–1459. http:// dx.doi.org/10.5858/arpa.2011-0267-RS PMID:23106592
17. Cheng L, Jones TD, McCarthy RP et al. (2005). Molecular genetic evidence for a common clonal origin of urinary blad- der small cell carcinoma and coexist- ing urothelial carcinoma. Am J Pathol, 166:1533–1539. http://dx.doi.org/10.1016/ S0002-9440(10)62369-3 PMID:15855652
18. Gaisa NT, Tilki D, Losen I et al. (2008). Insights from a whole cystectomy speci- men – association of primary small cell carcinoma of the bladder with transi- tional cell carcinoma in situ. Hum Pathol, 39:1258–1262. http://dx.doi.org/10.1016/j. humpath.2007.12.017 PMID:18547617
19. Chou R, Dana T (2010). Screening adults for bladder cancer: a review of the evidence for the U.S. preventive services task force. Ann Intern Med, 153:461–468. http://dx.doi. org/10.7326/0003-4819-153-7-201010050- 00009 PMID:20921545
20. Kaufman DS, Shipley WU, Feldman AS (2009). Bladder cancer. Lancet, 374:239– 249. http://dx.doi.org/10.1016/S0140-6736 (09)60491-8 PMID:19520422
21. Philips GK, Halabi S, Sanford BL et al.; Cancer and Leukemia Group B (2009). A phase II trial of cisplatin (C), gemcitabine (G) and gefitinib for advanced urothelial tract carcinoma: results of Cancer and Leukemia Group B (CALGB) 90102. Ann Oncol, 20:1074–1079. http://dx.doi.org/10.1093/ annonc/mdn749 PMID:19168670
22. Hussain MH, MacVicar GR, Petrylak DP et al.; National Cancer Institute (2007). Trastuzumab, paclitaxel, carboplatin, and gemcitabine in advanced human epider- mal growth factor receptor-2/neu-positive urothelial carcinoma: results of a multi- center phase II National Cancer Institute trial. J Clin Oncol, 25:2218–2224. http:// dx.doi.org/10.1200/JCO.2006.08.0994 PMID:17538166
23. Lamont FR, Tomlinson DC, Cooper PA et al. (2011). Small molecule FGF receptor inhibitors block FGFR-dependent urothe- lial carcinoma growth in vitro and in vivo. Br J Cancer, 104:75–82. http://dx.doi.org/ 10.1038/sj.bjc.6606016 PMID:21119661
24. Iyer G, Al-Ahmadie H, Schultz N et al. (2013). Prevalence and co-occurrence of action- able genomic alterations in high-grade bladder cancer. J Clin Oncol, 31:3133– 3140. http://dx.doi.org/10.1200/JCO.2012. 46.5740 PMID:23897969
Chapter 5.11 • Cancers of the male reproductive organs 453
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
1
Summary • Prostate cancer is the second
most common cancer in men worldwide. Risk factors for pros- tate cancer are age, family his- tory, and race.
• Prostate cancers are mainly ad- enocarcinomas and are indolent in many men, but there is also a lethal form. Gleason histo- logical grade, serum prostate- specific antigen (PSA) level, and stage are significant prognostic indicators.
• Recurrent protein-altering muta- tions are rare in prostate can- cer and are most often seen in the androgen receptor, PTEN, AKT1, and TP53 genes, usually in advanced-stage disease.
• Testicular cancer mostly affects young men. Risk factors include cryptorchidism, a prior testicu- lar germ cell tumour, a family history of germ cell tumour, an- drogen insensitivity syndrome, and gonadal dysgenesis with a Y chromosome.
• Testicular cancers are typically germ cell neoplasms, with his- topathological classification into seminomatous and non-semino- matous types, which has clinical significance. Testicular cancer is
5.11 5 ORGAN SITE
Cancers of the male reproductive organs Peter A. Humphrey Joachim Schüz (reviewer)
a very curable malignancy, even after metastatic spread.
• Single gene mutations are un- common in testicular germ cell tumours, but KIT, TP53, KRAS/ NRAS, and BRAF are the genes most commonly mutated and im- plicated in pathogenesis.
Prostate cancer Prostate cancer in most cases is acinar adenocarcinoma, a glandular malignancy. Primary sarcomas and other types of malignancy do arise in the prostate gland, but are rare.
Etiology The major identified risk factors for development of prostate cancer are age, family history, and race. Age is strongly related to the detection of prostate cancer, either clinically or at autopsy. The incidence of pros- tate cancer rises extremely steeply with age. Clinically, carcinoma of the prostate is most often detected in men older than 60 years. The likelihood of detection is nearly 40 times greater for men older than 65 than for men younger than 65. Only about 1% of prostate cancers are detected in men younger than 50 years. Prostate cancer in chil- dren is rare and is almost always rhabdomyosarcoma.
A man with a family history of prostate cancer has a significantly higher risk than one without such a history [1]. About 25% of men with prostatic carcinoma have a known family history [2]. The degree of risk is related to the age of the rela- tives at diagnosis and the number of relatives affected. Men with either a father or a brother with a diagno- sis of prostate cancer have an ap- proximately 2–3 times greater risk of developing prostate cancer than men without a family history of the disease. The greatest risk, with an 11-fold increase, is that for the father or the brother of a man diagnosed at about age 40 with prostate can- cer, with an additional first-degree relative who has prostate cancer. Genetic factors clearly make a ma- jor contribution to prostate cancer susceptibility, as highlighted by a twin cohort study indicating that 42% of prostate cancer cases exhibited heritable risk [3].
The role of race and ethnicity in prostate cancer etiology has been most intensively investigated in the USA, where the African American population is 1.6 times more likely to develop prostate cancer than the Caucasian population. Adjustment for recognized environmental, so- cioeconomic, and health-care fac- tors does not fully account for this difference.
454
Epidemiology
Prostate cancer • Globally, prostate cancer is the
second most frequently diag- nosed cancer and the fifth most common cause of cancer death among men, with an estimated 1.1 million new cases (15% of all cancers in men) and 0.3 million cancer deaths (7% of all cancer deaths in men) in 2012.
• In 2012, 60% of the estimated new cases and 41% of the deaths occurred in Europe and North America.
• Mortality rates are highest in countries and areas with pre- dominantly Black populations – in the Caribbean and in parts of sub-Saharan Africa – but mortality rates are also high in certain northern European countries, such as the Nordic countries.
• Incidence rates of prostate cancer vary by more than 25-fold in different parts of the world. The highest rates are in Australia and New Zealand (111.6 per 100 000), northern and western Europe, and North America.
• Prostate cancer was the most frequent type of cancer among men in 84 countries worldwide in 2012, largely in countries that have attained high or very high levels of human development, but also in several countries in Central and Southern Africa.
• Incidence rates increased dra- matically in the late 1980s in North America as prostate- specific antigen (PSA) testing became available; a similar pat- tern developed in many of the highest-resource countries dur- ing the 1990s. Incidence rates have levelled off in some of these countries but continue to uniformly increase in countries transitioning towards higher lev- els of human development.
Map 5.11.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for prostate cancer, 2012.
Map 5.11.2. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for prostate cancer, 2012.
• Mortality rates, which histori- cally increased with incidence, began to decrease in the 1990s in some of the highest-resource countries, likely as a result of a combination of curative treat- ment and earlier detection of the disease.
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.11 • Cancers of the male reproductive organs 455
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
1
Chart 5.11.1. Estimated global number of new cases and deaths with proportions by major world regions, for prostate cancer, 2012.
Chart 5.11.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for prostate cancer, circa 1975–2012.
Chart 5.11.3. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for prostate cancer, circa 1975–2012.
456
Testicular cancer • Testicular cancer is a relatively
rare cancer overall but a com- mon malignancy among young adult men (aged 15–39 years) in the most developed settings. Hence, while it is the 21th most frequently occurring cancer in men globally, with 55 000 new cases estimated in 2012 for all ages, it is by far the most com- mon cancer in young men in countries that have attained high or very high levels of hu- man development.
• Incidence rates of testicular cancer vary at least 25-fold in different parts of the world. The highest incidence rates are found in Caucasian populations in Europe (notably in Denmark, Norway, and Switzerland), Aus- tralia and New Zealand, and North America. Incidence rates are lowest in South-East Asia and throughout Africa.
• In 2012, 55.2% of all new cases of testicular cancer occurred in Europe and North America, but only 20.3% of the deaths. The fatality rate is one of the low- est of all forms of cancer, al- though it is considerably higher in countries classified as having low or medium levels of human development.
• Over the past 50 years, tes- ticular cancer incidence rates have uniformly increased in many countries with predomi- nantly Caucasian populations, and incidence appears to have peaked in several countries with the highest rates. In contrast, mortality rates have declined in line with improvements in treat- ment, notably with the introduc- tion of cisplatin therapy.
Map 5.11.3. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for testicular cancer, 2012.
Map 5.11.4. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for testicular cancer, 2012.
Chapter 5.11 • Cancers of the male reproductive organs 457
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
1
Chart 5.11.4. Estimated global number of new cases and deaths with proportions by major world regions, for testicular cancer, 2012.
Chart 5.11.5. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for testicular cancer, circa 1975–2012.
Chart 5.11.6. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for testicular cancer, circa 1975–2012.
458
Molecular factors that distinguish between races and are possibly causally related to the difference in incidence include the hormonal milieu of the tumour (with genetic mutations contributing to a higher dihydrotestosterone-to-testosterone ratio), decreased apoptosis due to lower BCL-2 levels, and overexpres- sion of epidermal growth factor re- ceptor [4]. Prostate cancer risk loci at 8q24 and 17q21 appear to be more frequent in men of African de- scent; together, these two loci could explain up to 60% of the increased risk in men of African descent [4].
Probable risk factors for prostate cancer include diet and nutrition, and hormonal factors [5]. A diet typical of industrialized countries has been proposed to be an important risk factor for prostate cancer, given the large worldwide variation in the in- cidence of prostate cancer and the
increased risk in migrants and their offspring who move from low-risk to high-risk countries. Dietary and saturated fat, red/processed/grilled meat, and milk and dairy products may increase risk, whereas toma- toes/lycopene, fish/marine omega-3 fatty acids, soy, and cruciferous vegetables (such as broccoli, cau- liflower, and Brussels sprouts) may decrease risk [5]. Studies on vitamin E are equivocal, and a large random- ized intervention trial with vitamin E and selenium, the SELECT trial, did not demonstrate reduced prostate cancer incidence. Data on calcium and zinc are conflicting. Vitamin D intake does not seem to be related to prostate cancer risk. Vitamin A as β-carotene in vegetable sources is not consistently associated with risk.
Hormonal influences on pros- tate cancer risk include androgens, the androgen receptor, and the
insulin-like growth factor axis. The importance of androgens is sub- stantiated by clinical trials on use of 5α-reductase inhibitors (e.g. fin- asteride, dutasteride), which block the conversion of testosterone to the more active molecule in the prostate, dihydrotestosterone. These trials demonstrated a decrease in low- grade prostate cancer incidence [6]. Circulating testosterone level is not, however, associated with pros- tate cancer risk. Short CAG repeats in the androgen receptor gene – a structural variation that increases its activity – have been connected in some, but not all, studies with a higher risk of prostate cancer. Finally, circulating levels of insulin- like growth factor, a polypeptide hor- mone that increases cell proliferation and decreases apoptosis of prostate cells, are associated with increased prostate cancer risk.
Lifestyle factors including physi- cal activity, cigarette smoking, and alcohol consumption have not been definitively linked to prostate cancer risk. Obesity does not appear to be linked to total prostate cancer inci- dence, but it may be associated with the development of advanced-stage or fatal prostate cancer [5].
No infection has been defini- tively linked to the development of prostate cancer, although there is a suggestion from two studies that Trichomonas vaginalis infection may be associated with an increased risk of prostate cancer, especially more aggressive disease [5]. Findings on the retrovirus XMRV, despite initial excitement, do not support a role for it in the etiology of prostate cancer [7]. Inflammation in the prostate is very common, and several separate lines of research, including epidemi- ological, genetic, histopathological, molecular pathological, and animal studies, have pointed towards a po- tential role for inflammation in pros- tate carcinogenesis [8].
Cholesterol levels may have a role in prostate cancer risk. Lower total cholesterol is associated with a lower risk of high-grade prostate cancer, and use of statins, which lower serum cholesterol, may be
Fig. 5.11.1. Few, if any, exogenous causes of prostate cancer have been clearly es- tablished. Accordingly, old age is recognized as one of a very limited number of proven risk factors. Prostate cancer is most common in North America, northern and western Europe, and Australia and New Zealand.
Chapter 5.11 • Cancers of the male reproductive organs 459
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
1
related to a lower risk of all prostate cancers and advanced or high-grade prostate cancer.
Pathology Adenocarcinoma of the prostate is thought to arise most commonly from a precursor proliferation known as high-grade prostatic intraepithe- lial neoplasia, which is an in situ proliferation of neoplastic prostatic epithelial cells (Fig. 5.11.2) [9]. The initial phase of the natural history of prostatic adenocarcinoma is growth into prostatic stroma. Of note, the prevalence of prostate cancer at au- topsy, i.e. latent prostate cancer, is extremely high and increases with age. Rare latent cases are detected in the third decade, and it is of inter- est that there is a frequency of about 30% in the fourth and fifth decades in the USA. There is variation of the frequency of latent prostate cancer depending on the world region, with the highest frequency in industrial- ized countries. Pathologically, most of these latent cancers are small and well differentiated, but a small minor- ity of latent prostate cancers are of a similar grade, size, and stage to clinical prostate cancer.
For clinical prostate cancer, a pathological diagnosis of malig- nancy is most often made based on light microscopic examination of fine-needle aspirates, sections of prostate needle biopsy tissue (Fig. 5.11.3), or sections of prostatic chips from transurethral resection of the prostate. Performance of an 18-gauge needle biopsy of the
prostate is a common clinical ap- proach when the patient has a pal- pable prostatic nodule and/or an elevated serum prostate-specific antigen (PSA) level.
Key histological features of ma- lignancy include an abnormal archi- tectural growth pattern of the glands (Fig. 5.11.4), loss of basal cells, and nuclear atypia. The growth patterns depicted in Fig. 5.11.4 indicate the Gleason grade of the tumour, which is one of the most powerful prognos- tic indicators for men with clinically localized prostate cancer. An ad- ditional histopathological finding of clinical importance in needle biopsy tissue is the extent of the cancer. Gleason histological grade, serum PSA level, clinical stage, and extent of cancer in needle core tissue are often used to stratify patients with clinically localized cancer as to risk for aggressive disease and for man- agement/treatment considerations.
The local spread of prostatic adenocarcinoma is into peripros- tatic adipose tissue, the bladder neck, and the seminal vesicles. Such direct extension of the cancer is diagnosed in radical prostatec- tomy specimens, and it is the light microscopic examination of these tissue sections that allows for as- signment of pathological stage, Gleason grade in the whole gland, and surgical margin status, which are significant indicators of clinical outcome. Metastasis of prostatic ad- enocarcinoma is mainly to the pelvic lymph nodes and to bone; for bone, the radiological and histological
picture is typically of osteoblastic metastasis.
Genetics The molecular genetics of prostate cancer may be separated into genet- ic susceptibility in the germline and the genetics of established sporadic tumours. Susceptibility has been in- vestigated in high-risk men identified with reference to African American heritage or familial disease. Genetic variants associated with prostate cancer susceptibility include more than 30 single-nucleotide polymor- phisms identified by genome-wide association studies, with replicated single-nucleotide polymorphisms most often located in the 8q24 re- gion [10]. However, these single- nucleotide polymorphisms together account for only an estimated 25%
Fig. 5.11.2. Prostatic intraepithelial neo- plasia. There is a proliferation of neoplastic epithelial cells in pre-existing glands.
Fig. 5.11.4. Gleason grading diagram, modified version. This diagram displays the different growth patterns of adenocar- cinoma of the prostate. Pattern 1 is the lowest grade and the least aggressive ad- enocarcinoma, while pattern 5 is the high- est grade and the most aggressive. The Gleason score is the sum of the grades for the two most common patterns.
Fig. 5.11.3. Adenocarcinoma of the pros- tate in needle core tissue. This is an infil- trative small gland malignancy.
460
of familial risk. A susceptibility locus that appears to be specific to men of African descent has been identi- fied at 17q21. Mutations in several specific susceptibility genes, includ- ing RNASEL/HPC1, ELAC2/HPC2, MSR1, and HOXB13, have been identified. These highly penetrant genes from high-risk families likely account for only a small fraction of the genetic predisposition to pros- tate cancer.
Another approach to suscepti- bility has involved whole-genome sequencing of 1795 Icelanders. A new low-frequency variant at 8q24 associated with prostate cancer in European populations was iden- tified. This variant was only very weakly correlated with previously reported risk variants at 8q24, and
its association remained significant after adjustment for all known risk- associated variants [11].
The molecular genetic profile of established prostate cancer is characterized by structural genomic changes in oncogenes and tumour suppressor genes such as NKX3.1 and PTEN, and amplification of the androgen receptor and MYC genes [12]. The prostate cancer genome undergoes frequent large-scale genomic rearrangements, including a common rearrangement produc- ing ETS gene fusions. Although several ETS and non-ETS family members have been observed to be fused with TMPRSS2 or other 5′ fusion partners, most fusions involve TMPRSS2-ERG. In contrast, recur- rent protein-altering mutations are
rare in prostate cancer and are most often seen in the androgen recep- tor, PTEN, AKT1, and TP53 genes, usually in advanced-stage disease. A molecular diversity model, where a wide range of possible pathways exist, has been proposed as a basis for pathological and genetic progres- sion of prostate cancer (Fig. 5.11.5) [13].
Metastatic castration-resistant prostate cancer has been the subject of specific investigation. One study [14] reported low overall mutation rates even in such tumours subject to intensive therapy and confirmed the monoclonal origin of lethal dis- ease. The analysis identified disrup- tions of CHD1 that defined a subtype of ETS gene family fusion-negative prostate cancer. ETS2, which was deleted in approximately one third of metastatic castration-resistant pros- tate cancers, was also dysregulated through mutation. Furthermore, re- current mutations in multiple chro- matin and histone modifying genes were evident, including MLL2 (mutat- ed in 8.6% of prostate cancers), as was interaction of the MLL complex with the androgen receptor, which is required for androgen receptor-me- diated signalling. There were novel recurrent mutations in the androgen receptor collaborating factor FOXA1.
Prospects The most immediately applicable prevention strategies could include increased physical activity, intake of tomatoes and other foods contain- ing lycopene, intake of foods con- taining selenium, and avoidance of diets high in calcium. Hence, it would seem beneficial to empha- size a diet consisting of a wide va- riety of plant-based foods and fish. Chemoprevention with pharmaceu- tical agents is not recommended at this time.
Screening for prostate cancer is controversial. Screening with serum PSA level has increased detection of all prostate cancers, including indo- lent prostate cancers. PSA emerged as one of the most-used serum tests to screen for cancer, particularly in the USA but also in Europe. Recent
Fig. 5.11.5. Two models of prostate cancer progression: a linear model compared with a molecular diversity model. In the molecular diversity model, not all pathways result in progression. Accumulation of molecular alterations associated with aggressive disease, such as the overexpression of EZH2 or PTEN mutations, may lead to invasive disease that progresses to metastatic disease, whereas other lesions, such as those with 5q or 6q gain and overexpression of AZGP1, might be seen most often in indolent disease. Mutations and alterations associated with p53 and the androgen receptor (AR) are probably late events and may play a key role in the development of castration-resistant disease. SNP, single-nucleotide polymorphism.
Chapter 5.11 • Cancers of the male reproductive organs 461
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
1
data from the Prostate, Lung, Colon, and Ovarian Cancer Screening Trial showed no benefit to screening, and the European Randomized Study of Screening for Prostate Cancer showed a 20% reduction in relative risk of cancer-specific mortality, which translated into an absolute reduction of prostate cancer-related deaths of 0.71 per 1000. Each trial has criticisms that may or may not have affected power and outcome, although the rate ratios comparing screening to not screening are simi- lar. Definitive evidence for or against screening is still lacking [15].
The detection via screening of both clinically significant and poten- tially clinically insignificant prostate cancer has created a dilemma as to which patients should receive aggres- sive treatment [16]. Overtreatment of screen-detected tumours is a major concern. In the USA, most men with screen-detected cancers undergo aggressive treatment. Such treat- ment is unlikely to yield a survival benefit in those with indolent disease or in men older than 65 years, but it can result in a considerable decrease in quality of life as a result of poten- tially persistent urinary, sexual, and bowel dysfunction [16].
Active surveillance of such low- risk disease is one strategy that could potentially counteract some of this overtreatment and has shown favourable outcomes in carefully se- lected patients. However, the criteria for enrolment into an active surveil- lance programme are not standard- ized [17]. Inclusion criteria typically
include low serum PSA level, low clinical stage, low Gleason grade (typically, Gleason score 6 or less), and a relatively small amount of cancer in needle core tissue [17]. Molecular markers in needle biopsy tissue hold promise for further re- finement in identification of patients with potentially indolent versus ag- gressive prostate cancer, and could help to stratify patients into active surveillance versus active treatment protocols.
Therapeutic targeting of prostate cancer based on genetic alterations is challenging since there are few recurrent genetic mutations. One promising target is the ETS gene fusions, seen in about one half of all prostate cancers [13].
Prostate cancer clinical biomark- ers of benefit are under active inves- tigation. Recent trials in patients with castration-resistant prostate cancer have incorporated the detection of circulating tumour cells, imag- ing, and patient-reported outcome biomarkers including pain and elevated PSA level [18]. Application of any such putative biomarkers will require a series of prospectively de- signed studies with adequate statisti- cal support to demonstrate a clinical effect within a well-defined context.
Testicular cancer Testicular cancer is in most cases a germ cell neoplasm. Sex cord stro- mal tumours and lymphomas may also involve the testis but are dis- tinctly uncommon.
Etiology Established risk factors for tes- ticular germ cell tumours include cryptorchidism, a prior testicular germ cell tumour, a family history of germ cell tumour, and various inter- sex syndromes such as androgen insensitivity syndrome (testicular feminization) and gonadal dysgen- esis with a Y chromosome [19,20]. Cryptorchidism carries about a 4-fold increased risk of germ cell tumour development. Patients with a history of testicular germ cell tu- mour have an increased incidence of
contralateral testicular cancer, with an incidence of 2–5%, which trans- lates into a 5–10-fold increased risk.
A small minority of testicular cancers appear to be familial, with approximately 2% of patients hav- ing an affected family member. First- degree male relatives of patients with testicular germ cell tumour have a 3–10-fold increased risk of being di- agnosed with testicular cancer. The mode of transmission is not known but may be autosomal recessive in nature [20]. Patients with androgen insensitivity syndrome have a 15- fold increased risk of developing a testicular germ cell tumour, while the risk increase is 50-fold for patients with gonadal dysgenesis with a Y chromosome.
Several prenatal factors, includ- ing maternal endogenous estrogen levels, and low birth weight have been suggested as risk factors, but the data are inconclusive [20]. One hypothesis is that in utero exposure to exogenous estrogens could result in increased risk of testicular cancer due to increased cryptorchidism and testicular dysgenesis. Several post- natal factors, such as body size (es- pecially height), nutrition, chemical exposure, occupation, infection, and physical activity, have been hypoth- esized to be potentially associated with testicular cancer, but the data appear mixed [20]. The peak age at diagnosis (twenties and thirties) ap- pears related to sex hormone activ- ity and may reflect promotion of cells initiated during gestation. Finally, mi- grant studies suggest the importance of as-yet-uncharacterized environ- mental/dietary factors. For example, the incidence of testicular cancer in Finland is half that found in Sweden, and second-generation Finnish mi- grants to Sweden have a testicular cancer incidence that approaches that of the Swedish population.
Pathology The precursor for all types of germ cell tumours that originate in the post-pubertal testis is intratubular germ cell neoplasia. This lesion is seen in about 4% of cryptorchid tes- tes, in approximately 5% of testes
Fig. 5.11.6. Intratubular germ cell neo- plasia in the testis. This is a precursor for most testicular cancers.
462
contralateral to a testis with a germ cell tumour, and almost always adja- cent to an invasive germ cell tumour. Histologically, intratubular germ cell neoplasia is characterized by the presence of large atypical germ cells inside seminiferous tubules (Fig. 5.11.6), which often have a thickened basement membrane. Cytologically, intratubular germ cell neoplasia cells often resemble seminoma cells. Most patients with isolated intratubular germ cell neo- plasia develop an invasive germ cell tumour within 7 years. For patients with gonadal dysgenesis with a Y chromosome, gonadoblastoma acts as the in situ precursor proliferation.
Invasive germ cell tumours are broadly classified using histopatho- logical criteria as seminomatous versus non-seminomatous [21]. Microscopically, classic seminoma comprises a solid growth of primi- tive-appearing cells, separated by fibrous bands infiltrated by lym- phocytes (Fig. 5.11.7). In infants and children, yolk sac tumour and teratoma are the most common histological types. Of note, pre- pubertal teratomas are benign, whereas post-pubertal teratomas are malignant. In adults, pure semi- nomas account for about 50% of all testicular germ cell tumours. Non- seminomatous germ cell tumour types include embryonal carcino- ma, yolk sac tumour, trophoblastic tumours (mainly choriocarcinoma), and teratoma. Mixed malignant germ cell tumours composed of dif- ferent germ cell types are common, representing 30% of all testicular
cancers. For these mixed tumours, the presence of embryonal carci- noma and the percentage of the tu- mour that is embryonal carcinoma are important prognostic indicators.
In addition to histological type, important histopathological attri- butes of germ cell tumours in radi- cal orchiectomy specimens include pathological stage and presence of lymphovascular invasion by the germ cell tumour. Serum tumour marker levels are important in stag- ing and patient management and correlate with specific cell/tumour types. α-Fetoprotein is produced mainly by yolk sac tumours, whereas β-human chorionic gonadotropin is produced by syncytiotrophoblasts in choriocarcinoma and syncytio- trophoblasts seen in seminoma with syncytiotrophoblasts or mixed malig- nant germ cell tumours.
Testicular germ cell tumours spread by lymphatics, first to retro- peritoneal para-aortic lymph nodes and then to mediastinal and supra- clavicular lymph nodes. Haema- togenous dissemination is primar- ily to the lungs, but the liver, brain, and bones may be involved. Clinical stage is critical in defining treatment and prognosis.
Genetics Recently reported genome-wide association studies implicate six gene loci that predispose to testicu- lar germ cell tumour development. These include KIT ligand (KITLG; also known as SCF), SPRY4, BAK1, TERT, ATF7IP, and DMRT1. KITLG, SPRY4, and BAK1 are involved in KIT signalling, TERT and ATF7IP maintain telomere length and are re- activated in a range of tumour types, and DMRT1 is responsible for male sex determination [20,22]. Multiple new susceptibility loci for testicular germ cell tumours continue to be re- ported [23,24].
The genetics of established pae- diatric and adult germ cell tumours are distinct [19]. Paediatric tumours are mostly diploid, particularly tera- tomas, and typically show deletion of 1p, loss of 6q, and structural abnor- malities of chromosome 2 and 3p.
In contrast, invasive adult germ cell tumours are consistently aneuploid with gain of chromosomal material on 12p, usually in the form of iso- chromosome 12p.
Parallels between the origin of testicular germ cell tumours and malignant ovarian germ cell tu- mours have been considered with reference to germ cell develop- ment, endocrinological influences, and pathogenesis [25]. The basis of analysis was genomic aberrations assessed by ploidy, cytogenetic banding, and comparative genomic hybridization.
The development of intratubu- lar germ cell neoplasia may involve an aberrantly activated KITLG/KIT pathway and overexpression of em- bryonic transcription factors such as NANOG and POU5F1, which leads to suppression of apoptosis, increased proliferation, and accu- mulation of mutations in gonocytes. Single gene mutations are uncom- mon in testicular germ cell tumours, but KIT, TP53, KRAS/NRAS, and BRAF are the genes most com- monly mutated and implicated in pathogenesis. Different histologi- cal subtypes of testicular germ cell tumours have different gene expres- sion profiles, which reflect different directions of differentiation. Their distinct gene expression profiles are likely caused by epigenetic regula- tion, in particular DNA methylation, but not by gene copy number alter- ations [26].
Prospects Prevention of testicular germ cell neoplasia is not currently feasible. There are no reliable screening tests for these cancers. Testicular self- examination has not been shown to improve outcomes but is still advo- cated by some medical professional organizations. Such examination may be more beneficial in men at increased risk of testicular cancer, including men with a history of crypt- orchidism, testicular atrophy, or a family history of testicular cancer. Most patients present with a nodule or painless swelling of one gonad, or a painful testis, so these clinical
Fig. 5.11.7. Seminoma of the testis, with large primitive neoplastic cells.
Chapter 5.11 • Cancers of the male reproductive organs 463
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
1
findings should prompt evaluation. A systematic review of options for testicular cancer screening con- cluded that no published random- ized controlled trial of screening compared with no screening had established reduced testicular can- cer-specific mortality [27].
There is a high cure rate for testicular germ cell neoplasia, us- ing multidisciplinary management. The 5-year survival rates for men with localized disease, regional spread, and distant metastasis are 99%, 96%, and 72%, respectively. For patients with intermediate- and
poor-prognosis germ cell tumours, identification of molecular mecha- nisms of therapy resistance and development of novel molecularly targeted therapeutic approaches are important goals.
464
References 1. Johns LE, Houlston RS (2003). A system-
atic review and meta-analysis of familial prostate cancer risk. BJU Int, 91:789–794. http://dx.doi.org/10.1046/j.1464 - 410X. 2003.04232.x PMID:12780833
2. Walsh PC, Partin AW (1997). Family his- tory facilitates the early diagnosis of pros- tate carcinoma. Cancer, 80:1871–1874. http://dx.doi.org/10.1002/(SICI)1097-0142 (19971101)80:9<1871::AID-CNCR28>3.0. CO;2-1 PMID:9351562
3. Lichtenstein P, Holm NV, Verkasalo PK et al. (2000). Environmental and heritable fac- tors in the causation of cancer – analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med, 343:78–85. http://dx.doi.org/10.1056/NEJM2000071 33430201 PMID:10891514
4. Henderson BE, Lee NH, Seewaldt V, Shen H (2012). The influence of race and ethnicity on the biology of cancer. Nat Rev Cancer, 12:648–653. http://dx.doi. org/10.1038/nrc3341 PMID:22854838
5. Giovannucci E, Platz EA, Mucci L (2011). Epidemiology of prostate cancer. In: Scardino PT, Linehan WM, Zelefsky MJ et al., eds. Comprehensive Textbook of Genitourinary Oncology, 4th ed. Baltimore: Lippincott Williams & Wilkins, pp. 1–17.
6. Andriole GL, Bostwick DG, Brawley OW et al.; REDUCE Study Group (2010). Effect of dutasteride on the risk of prostate can- cer. N Engl J Med, 362:1192–1202. http:// dx.doi.org /10.10 5 6 / N E J M oa 0 9 0 8127 PMID:20357281
7. Sfanos KS, Aloia AL, De Marzo AM, Rein A (2012). XMRV and prostate cancer – a ‘final’ perspective. Nat Rev Urol, 9:111–118. http://dx.doi.org/10.1038/nrurol.2011.225 PMID:22231291
8. Sfanos KS, De Marzo AM (2012). Prostate cancer and inflammation: the evidence. Histopathology, 60:199–215. http://dx.doi. org/10.1111/j.1365 -2559.2011.04033.x PMID:22212087
9. Epstein JI, Cubilla AL, Humphrey PA (2011). Tumors of the Prostate Gland, Seminal Vesicles, Penis, and Scrotum (AFIP Atlas of Tumor Pathology Series 4, Fascicle 14). Washington, DC: American Registry of Pathology and Armed Forces Institute of Pathology.
10. Ishak MB, Giri VN (2011). A systematic review of replication studies of prostate cancer susceptibility genetic variants in high-risk men originally identified from genome-wide association studies. Cancer Epidemiol Biomarkers Prev, 20:1599–1610. http://dx.doi.org/10.1158/1055-9965.EPI- 11-0312 PMID:21715604
11. Gudmundsson J, Sulem P, Gudbjartsson DF et al. (2012). A study based on whole- genome sequencing yields a rare variant at 8q24 associated with prostate cancer. Nat Genet, 44:1326–1329. http://dx.doi. org/10.1038/ng.2437 PMID:23104005
12. Barbieri CE, Demichelis F, Rubin MA (2012). Molecular genetics of prostate cancer: emerging appreciation of genetic complexity. Histopathology, 60:187–198. ht tp://dx.doi.org/10.1111/j.1365 -2559. 2011.04041.x PMID:22212086
13. Rubin MA, Maher CA, Chinnaiyan AM (2011). Common gene rearrange- ments in prostate cancer. J Clin Oncol, 29:3659–3668. http://dx.doi.org/10.1200/ JCO.2011.35.1916 PMID:21859993
14. Grasso CS, Wu YM, Robinson DR et al. (2012). The mutational landscape of lethal castration-resistant prostate can- cer. Nature, 487:239–243. http://dx.doi. org/10.1038/nature11125 PMID:22722839
15. Rove KO, Crawford ED (2012). Randomized controlled screening trials for prostate can- cer using prostate-specific antigen: a tale of contrasts. World J Urol, 30:137–142. http:// dx.doi.org/10.1007/s00345 -011-0799- 4 PMID:22116599
16. Sandhu GS, Andriole GL (2012). Overdiagnosis of prostate cancer. J Natl Cancer Inst Monogr, 2012:146–151. http:// dx.doi.org/10.1093/jncimonographs/lgs031 PMID:23271765
17. Buethe DD, Pow-Sang J (2012). Enrollment criteria controversies for active surveil- lance and triggers for conversion to treat- ment in prostate cancer. J Natl Compr Canc Netw, 10:1101–1110. PMID:22956809
18. Scher HI, Morris MJ, Larson S, Heller G (2013). Validation and clinical util- ity of prostate cancer biomarkers. Nat Rev Clin Oncol, 10:225–234. http:// dx.doi.org/10.1038/nrclinonc. 2013.30 PMID:23459624
19. Reuter VE (2005). Origins and molecu- lar biology of testicular germ cell tumors. Mod Pathol, 18 Suppl 2:S51–S60. http:// dx.doi.org/10.1038/modpathol.3800309 PMID:15761466
20. Vaughan DJ, Kanetsky PA, Nathanson KL (2011). The epidemiology and genetics of sporadic and hereditary testicular germ cell tumors. In: Scardino PT, Linehan WM, Zelefsky MJ et al., eds. Comprehensive Textbook of Genitourinary Oncology, 4th ed. Baltimore: Lippincott Williams & Wilkins, pp. 509–520.
21. Emerson RE, Ulbright TM (2007). Morphological approach to tumours of the testis and paratestis. J Clin Pathol, 60:866–880. http://dx.doi.org/10.1136/jcp. 2005.036475 PMID:17307866
22. Gilbert D, Rapley E, Shipley J (2011). Testicular germ cell tumours: predisposi- tion genes and the male germ cell niche. Nat Rev Cancer, 11:278–288. http://dx.doi. org/10.1038/nrc3021 PMID:21412254
23. Chung CC, Kanetsky PA, Wang Z et al. (2013). Meta-analysis identifies four new loci associated with testicular germ cell tu- mor. Nat Genet, 45:680–685. http://dx.doi. org/10.1038/ng.2634 PMID:23666239
24. Ruark E, Seal S, McDonald H et al.; UK Testicular Cancer Collaboration (UKTCC) (2013). Identification of nine new suscep- tibility loci for testicular cancer, includ- ing variants near DAZL and PRDM14. Nat Genet, 45:686–689. http://dx.doi. org/10.1038/ng.2635 PMID:23666240
25. Kraggerud SM, Hoei-Hansen CE, Alagaratnam S et al. (2013). Molecular characteristics of malignant ovarian germ cell tumors and comparison with testicular counterparts: implications for pathogene- sis. Endocr Rev, 34:339–376. http://dx.doi. org/10.1210/er.2012-1045 PMID:23575763
26. Sheikine Y, Genega E, Melamed J et al. (2012). Molecular genetics of testicu- lar germ cell tumors. Am J Cancer Res, 2:153–167. PMID:22432056
27. Ilic D, Misso ML (2011). Screening for tes- ticular cancer. Cochrane Database Syst Rev, 2:CD007853. PMID:21328302
Website Surveillance, Epidemiology and End Results Program. SEER Cancer Statistics Review, 1975–2009 (Vintage 2009 Populations): http://seer.cancer.gov/csr/1975_2009_pops09/
Chapter 5.12 • Cancers of the female reproductive organs 465
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
Summary • Cancer of the vulva accounts
for 3% of all female genital can- cers and it mainly includes two different types of carcinomas. Keratinizing carcinomas pre- dominate in older women and have little association with hu- man papillomavirus (HPV) infec- tion. Conversely, the majority of warty and basaloid carcinomas are caused by HPV.
• Cervical cancer, the fourth most common cancer among women worldwide, with about 70% of the cases occurring in develop- ing countries, is caused by HPV. Precancerous lesions may be detected and treated; HPV vac- cination offers the possibility of primary prevention. The most common somatic mutations were found for PIK3CA, and are associated with shorter survival.
• More than 80% of endometrial cancers occur as endometrioid carcinomas, a form that is as- sociated with estrogen expo- sure; obesity is estimated to account for up to 40% of cases worldwide. Principal genetic le- sions include microsatellite in- stability and mutations of the PTEN, PIK3CA, ARID1A, KRAS, and CTNNB1 (β-catenin) genes,
5.12 5 ORGAN SITE
Cancers of the female reproductive organs Jaime Prat Silvia Franceschi
Lynette Denny (reviewer) Eduardo Lazcano Ponce (reviewer)
whereas TP53 and PPP2R1A mutations, loss of E-cadherin, HER2/neu amplification, and loss of heterozygosity at multiple loci occur in non-endometrioid carcinomas.
• Ovarian cancer comprises sev- eral tumour types, including the predominant (70%) high-grade serous carcinomas, which ex- hibit TP53 mutations and BRCA abnormalities and are distin- guished from low-grade serous carcinomas (< 5%), which usu- ally exhibit KRAS and BRAF
mutations. Endometrioid and clear cell carcinomas of the ovary have the same genetic abnormalities as their uterine counterparts. Most mucinous carcinomas involving the ovary are metastatic tumours from the gastrointestinal tract.
Vulvar cancer Cancer of the vulva accounts for 3% of all female genital cancers and occurs mainly in women older than 60 years. Squamous cell carcinoma is the most common type (86%).
Fig. 5.12.1. A patient attends a counselling session with a service provider in the family planning unit at Orolodo primary health centre in Omuaran township in Kwara State, Nigeria.
466
Epidemiology
Cervical cancer • Cervical cancer is the fourth
most common cancer (528 000 new cases) and the fourth most common cause of cancer death (266 000 deaths) in women worldwide, as estimated in 2012.
• The disease affects predomi- nantly women in lower-resource countries; almost 70% of the global burden occurs in areas with low or medium levels of human development, and more than one fifth of all new cases of cervical cancer are diagnosed in India.
• Incidence rates of cervical can- cer vary greatly from country to country. The disease is the most common cancer among women in 39 of the 184 countries world- wide, and is the leading cause of cancer death in women in 45 countries. These countries are mainly in sub-Saharan Africa, parts of Asia, and some countries in Central and South America.
• The lowest incidence rates tend to be in western Europe, North America, Australia and New Zealand, and the eastern Mediterranean.
• Over the past 30 years, cervi- cal cancer rates have declined in many countries that are tran- sitioning towards higher levels of human development. This trend reflects changing soci- etal factors linked to economic
Map 5.12.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for cervical cancer, 2012.
Map 5.12.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for cancer of the corpus uteri (endometrial cancer), 2012.
These tumours are divided into two groups: keratinizing squamous cell carcinomas, which have little asso- ciation with human papillomavirus (HPV), and warty and basaloid carci- nomas, which are strongly associated with high-risk HPV, mainly HPV16, and are increasingly often diagnosed in younger women in countries in which cervical screening is common.
Etiology HPV infection causes a subset of squamous cell carcinomas of the vulva and vagina [1]. According to meta-analysis, HPV prevalence was 40.4% in vulvar carcinoma and 69.9% in vaginal carcinoma [2]. Also strongly related to HPV are vulvar intraepithelial and vaginal intraepi- thelial neoplasias. The meta-analysis
supported the hypothesis that two distinct subsets of carcinomas of the vulva and vagina exist: one that is strongly associated with HPV and that may be preceded by high-grade lesions as cervical carcinoma, and another that arises independently of HPV infection [3].
HPV16 was found in more than three quarters of HPV-positive
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.12 • Cancers of the female reproductive organs 467
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
progress but also points, in some countries, to the implementation of effective secondary prevention programmes.
• In contrast, rates are increasing in certain countries with high or very high levels of human devel- opment, including several eastern European countries and former Soviet states. Recent changes in sexual behaviour have led to an increase in the risk of infection with high-risk human papilloma- virus types in these populations, largely in the absence of effective screening programmes.
Cancer of the corpus uteri (endometrial cancer) • Cancer of the corpus uteri (en-
dometrial cancer) is the sixth most common cancer in women (almost 5% of all cancers in wom- en). There were an estimated 320 000 new cases and 76 000 deaths from the disease in 2012.
• Incidence rates vary 20–30-fold between countries; close to two thirds of the estimated new cas- es occur in countries with very high or high levels of human development.
• Incidence rates are elevated in northern and eastern Europe and North America, and tend to be low in Africa and West Asia.
• About 48% of the new cases occurred in Europe and North America; 41% of the estimated
new cases and 45% of the cancer deaths occurred in Asia.
Ovarian cancer • With an estimated 239 000 new
cases in 2012, ovarian cancer is the seventh most common cancer in women, representing 4% of all cancers in women.
• The fatality rate of ovarian cancer tends to be rather high relative to other cancers of the female repro- ductive organs, and case fatality is higher in lower-resource settings. As a consequence, ovarian cancer is the eighth most frequent cause of cancer death among women, with 152 000 deaths.
• In 2012, almost 55% of all new cases occurred in countries with high or very high levels of human development; 37% of the new cases and 39% of the deaths occurred in Europe and North America.
• Incidence rates are highest in northern and eastern Europe, North America, and Oceania, and tend to be relatively low in Africa and much of Asia.
• Incidence rates have been de- clining in certain countries with very high levels of human devel- opment, notably in Europe and North America.
Map 5.12.3. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for ovarian cancer, 2012.
carcinomas of the vulva and vagina. Conversely, HPV18 was rarer in car- cinomas of the vulva and vagina than in carcinoma of the cervix. HPV6 and 11 were commonly detected in vulvar intraepithelial neoplasias but not in vaginal intraepithelial neoplasias [2].
There is substantial heteroge- neity in HPV prevalence in vulvar squamous cell carcinomas. HPV prevalence was 69.4% in warty and basaloid carcinomas but only 13.2% in keratinizing carcinomas. HPV- positive vulvar carcinomas occur in significantly younger women than do
HPV-negative vulvar carcinomas [2], and HPV prevalence in vulvar carci- nomas in North America was approxi- mately twice that in other geographi- cal regions [4], possibly attributable to enhanced surveillance of the vulva during cervical screening and some misclassification, together with higher HPV prevalence.
Risk factors for carcinomas of the vulva and vagina other than HPV in- fection include sexual habits, smok- ing, and immunosuppression [1]. Current HPV vaccines may prevent
approximately 30% of vulvar carcino- mas and 60% of vaginal carcinomas.
Precursor lesions Two pathogenetic pathways, featur- ing different precursor lesions, exist for vulvar squamous cell carcinomas, depending on HPV involvement.
Squamous intraepithelial lesions of the vulva are HPV-related in 90% of cases and range from low-grade vulvar intraepithelial neoplasia grade 1 to the severe full-thickness dys- plasia of high grade (grade 3) [5]. Patients are usually in their thirties
468
Chart 5.12.1. Estimated global number of new cases and deaths with proportions by major world regions, for cervical cancer, 2012.
Chart 5.12.2. Estimated global number of new cases and deaths with proportions by major world regions, for cancer of the corpus uteri (endometrial cancer), 2012.
Chapter 5.12 • Cancers of the female reproductive organs 469
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
Chart 5.12.3. Estimated global number of new cases and deaths with proportions by major world regions, for ovarian cancer, 2012.
Chart 5.12.4. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for cervical cancer, circa 1975–2012.
Chart 5.12.5. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for cancer of the corpus uteri (endometrial cancer), circa 1975–2012.
470
or forties and often have multifocal or multicentric HPV-related disease [6,7]. Most low-grade lesions con- tain low-risk HPV virus subtypes, whereas high-grade lesions contain high-risk subtypes, most commonly HPV16. The typical progression time from incident infection to clinical dis- ease is 18.5 months [6]. Only about 6% of high-grade lesions progress to invasive carcinomas, except in older or immunosuppressed women [7].
Patients with high-grade vulvar intraepithelial neoplasia of differen- tiated or simplex type are usually postmenopausal and often have li- chen sclerosus or lichen simplex chronicus [8]. Differentiated vulvar intraepithelial neoplasia is more likely (33%) to progress to invasive carcino- ma than corresponding HPV-related disease [9]. In one study comparing the association of the two types of vulvar intraepithelial neoplasia with
invasive carcinoma, the rates were 24.2% for HPV-related grade 3 vulvar intraepithelial neoplasia and 83.3% for the corresponding differentiated lesions [6].
Pathology and genetics Whereas low-grade squamous in- traepithelial neoplasia is a rare di- agnosis [6], high-grade disease is divided into warty and basaloid sub- types. Warty lesions show a spiky or undulating surface. The thickened epithelium forms wide, deep rete pegs separated by thin dermal pa- pillae. Hyperkeratosis is prominent. In contrast to warty lesions, basaloid lesions show a flat surface and a uni- form population of immature cells. Keratinocytes undergo premature keratinization, and rete pegs often contain keratin pearls.
Superficially invasive squamous cell carcinoma is the earliest stage
(stage IA) of invasive vulvar cancer and has an excellent prognosis [10]. Most invasive carcinomas of the vul- va arise in a background of lichen sclerosus with or without differenti- ated vulvar intraepithelial neoplasia. Tumours are frequently exophytic, but some may be ulcerative. Vulvar carcinomas are composed of inva- sive nests of malignant squamous epithelium with central keratin pearls (Fig. 5.12.2). HPV-related tumours typically occur in younger patients and tend to have a similar morphol- ogy [3,6]. Vulvar carcinomas grow slowly, extending to contiguous skin, vagina, and rectum. They metasta- size initially to inguinal and femoral lymph nodes, and subsequently to pelvic lymph nodes [11].
Differentiated vulvar intraepithe- lial neoplasia shows overexpression of p53 in the basal and suprabasal epithelial layers [7]. Cases of lichen sclerosus, differentiated vulval in- traepithelial neoplasia, and invasive squamous cell carcinoma with iden- tical TP53 mutations have been re- ported [8]. TP53 mutation, however, is an uncommon and late event in vulvar carcinogenesis. Of greater utility in the diagnosis of high-grade squamous intraepithelial lesions is p16INK4A; as a reliable marker of high- risk HPV [10], it is strongly positive throughout the epithelium [12].
Cervical cancer Cervical cancer originates at the junction between the columnar epi- thelium of the endocervix and the squamous epithelium of the ectocer- vix, a site of continuous metaplastic change, especially at puberty and from after the first pregnancy until menopause. Persistent epithelial in- fection with one or more oncogenic types of HPV may lead to the devel- opment of precancerous lesions, a small proportion of which may prog- ress to invasive cervical cancer over a period of 10–20 years [1].
Etiology HPV infection is the most common sexually transmitted infection world- wide, and the majority of sexually
Chart 5.12.6. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for ovarian cancer, circa 1975–2012.
Chapter 5.12 • Cancers of the female reproductive organs 471
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
active individuals of both sexes ac- quire it at some time during their life. About 291 million women world- wide have HPV infection of the cer- vix at any time, corresponding to a prevalence of 10.4%, although the prevalence is higher among women younger than 25 years (16.9%) [13].
High HPV prevalence occurs in countries where the burden of cervi- cal cancer is high (Fig. 5.12.3), i.e. in sub-Saharan Africa, Latin America, and India, but also in countries like Mongolia and China. Large differ- ences in age-specific HPV preva- lence are evident in several less- developed countries (India, China, and African countries) [14]. More than 90% of new HPV infections at any age regress over 6–18 months [15]. Low-grade squamous intra- epithelial lesions, also known as cervical intraepithelial neoplasia
Fig. 5.12.3. Age-adjusted prevalence (percentage) of cervical human papillomavirus (HPV) DNA in sexually active women aged 15–69 years in various countries.
Fig. 5.12.2. Keratinizing squamous cell carcinoma of the vulva. Anastomosing cords of neoplastic squamous cells, some with keratin pearls, are evident.
472
grade 1 (CIN1), is an insensitive proxy of HPV infection and is not precancer. High-grade squamous intraepithelial lesions (CIN grade 2) include a heterogeneous group of lesions that are equivocal in cancer potential. In contrast, grade 3 dis- ease represents the most clinically relevant lesion and is the best sur- rogate end-point for cervical cancer in screening and vaccination trials. The probability of HPV clearance depends on the duration of the in- fection [16].
The mean age at which women develop squamous intraepithelial le- sions is 24–27 years for low-grade
lesions (CIN1) and 35–42 years for high-grade lesions (CIN3). Half of low-grade lesions regress, 10% progress to high-grade lesions, and less than 2% become invasive can- cer. The average time for low-grade squamous intraepithelial lesions to progress to high-grade lesions (CIN3) is about 10 years. At least 20% of cases of CIN3 progress to invasive carcinoma in that time.
The strongest risk factors for persistence and progression are immunodeficiency and the type of HPV involved [1]. Of 13 HPV types, HPV16 is by far the likeliest to persist and cause CIN3 and cervical cancer.
The proportion of HPV16 increases among HPV-positive women with the increase in lesion severity [17]. Conversely, the proportion of other high-risk HPV types is diminished in CIN3 and cervical cancer. Women who tested positive twice for HPV16 after a 9–21-month interval had a 3-year cumulative incidence of CIN2 or worse of 40%. The corresponding cumulative incidence was 15% for HPV18 and 8.5% for other high-risk HPV types.
The most commonly detected HPV types in cervical cancer world- wide are HPV16 (57%) and 18 (16%), followed by HPV58, 33, 45, 31, 52,
Fig. 5.12.4. Interrelations of naming systems in precursor human papillomavirus (HPV)-related cervical lesions. This chart integrates multiple aspects of the disease. It illustrates the changes in progressively more abnormal disease states and provides terminology for the dysplasia/carcinoma in situ system, the cervical intraepithelial neoplasia (CIN) system, and the Bethesda system. The scheme also illustrates the corresponding cytological smear resulting from exfoliation of the most superficial cells as well as the equivalent histopathological lesions (top). SIL, squamous intraepithelial lesion.
Chapter 5.12 • Cancers of the female reproductive organs 473
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
and 35 [17]. Therefore, current vac- cines that prevent infection by HPV16 and 18 should prevent about 75% of cervical cancers and 60% of high- grade lesions (that need treatment when detected during screening), with small variations by region [17]. Future confirmation of cross-protection of HPV16 and 18 vaccines against other high-risk types [18] or the introduc- tion of vaccines targeting additional high-risk HPV types would increase prevention (see Chapter 4.6).
Cervical cancer risk factors other than HPV infection are either weak in terms of the magnitude of association (smoking) or represent correlates of the probability of acquisition (number of sexual partners) or, possibly, the likelihood of progression of HPV in- fection to cancer (age at first sexual intercourse and first birth, parity, and oral contraceptive use) [19].
CIN is a spectrum of epithelial changes that begins with minimal atypia and progresses through stag- es of greater epithelial abnormali- ties to invasive squamous cell car- cinoma. The terms CIN, dysplasia, carcinoma in situ, and squamous intraepithelial lesion are commonly used interchangeably (Fig. 5.12.4). p16 immunohistochemistry is rec- ommended when the differential di- agnosis is between precancer (CIN2 or CIN3) and a mimic of precancer, such as immature squamous meta- plasia or reparative changes. Strong and diffuse block-positive p16 results support a diagnosis of precancerous disease [5].
Early phases of infection with all HPV types probably show episomal viral propagation throughout a poly- clonal epithelial field, with low-grade squamous intraepithelial lesions (CIN1) cytology. These lesions mark a permissive infection, i.e. episomal HPV freely replicates and causes cell death. Huge numbers of virus must accumulate in the cytoplasm before being visible as a koilocyte (Fig. 5.12.4). In contrast, in most cases of CIN2/3, viral DNA integrates into the cell genome. Some 85% of CIN1 lesions, as well as many genital warts, contain low-risk HPV6 or 11. In contrast, cells in CIN2/3 usually have
HPV types 16, 18, 58, 33, 45, 31, 52, and 35 [17].
Pathology and genetics The association with HPV infection is equally strong for the two main histological types of cervical cancer: squamous cell carcinoma and adeno- carcinoma. However, since most cer- vical abnormalities caused by HPV infection are unlikely to progress to high-grade CIN or cervical cancer, other exogenous or endogenous fac- tors acting in conjunction with HPV infection may be necessary for pro- gression of the disease [1].
The normal process by which cervical squamous epithelium ma- tures is disturbed in squamous intraepithelial lesions. In low-grade lesions (CIN1), the most pronounced changes are in the basal third of the epithelium. However, abnormal cells are present throughout the thickness of the epithelium, and nuclei in the upper levels are still abnormal [5]. Thus, the sloughed cells can be de- tected as abnormal in Pap smears. In high-grade lesions (CIN2), most cellular abnormalities are in the lower and middle thirds of the epi- thelium. High-grade squamous intra- epithelial lesions are often detected
on colposcopic examination. In the case of adenocarcinoma in situ, nor- mal columnar epithelium is replaced by dysplastic glandular epithelium. However, morphological change comparable to CIN1 does not occur in the glandular epithelium.
Superficially invasive or microin- vasive squamous cell carcinoma is the earliest stage of invasive cervical cancer. In this setting, stromal inva- sion usually arises from an overly- ing squamous intraepithelial lesion. Staging of microinvasive disease is based on width and depth of invasion [11]. The earliest invasive changes, sometimes termed early stromal invasion, appear as tiny irregular epithelial buds emanating from the base of CIN3 lesions. These small (< 1 mm) tongues of neoplastic epi- thelial cells do not affect the progno- sis of CIN3 lesions; hence, both can be treated similarly with conserva- tive surgery.
Histologically, 85–90% of inva- sive cervical cancers are squamous cell carcinoma, appearing as infiltrat- ing networks of neoplastic cells in the stroma, with varying degrees of dif- ferentiation, with or without keratini- zation (Fig. 5.12.5). Adenocarcinoma and its variants constitute 10–15%
Fig. 5.12.5. Squamous cell carcinoma, large cell, non-keratinizing type.
474
of cervical cancers. The most com- mon type of adenocarcinoma is the endocervical cell type, showing ab- normal glands with varying size and shape with budding and branching, infiltrating the stroma (Fig. 5.12.6). Cervical cancer spreads by direct extension, through lymphatic ves- sels and only rarely by the haema- togenous route. Local extension into surrounding tissues (parametrium) results in ureteric compression [11].
Genome-wide association stud- ies have identified loci associated with cervical cancer susceptibility, including those at 4q12 and 17q12 in the Han Chinese population [20] and at 6p21.3 in a Swedish study [21]. A recent genetic analysis examined 80 cervical tumours for 1250 known mutations in 139 cancer genes [22]. Validated mutations were detected in 48 of the 80 tumours (60%). The highest mutation rates were in the PIK3CA (31.3%), KRAS (8.8%), and EGFR (3.8%) genes. PIK3CA muta- tion rates did not differ significantly between adenocarcinomas and squamous cell carcinomas. In con- trast, KRAS mutations were identi- fied only in adenocarcinomas (17.5% vs 0%), and a novel EGFR mutation was detected only in squamous cell
carcinomas (0% vs 7.5%). There were no associations between HPV16 or HPV18 and somatic mutations or overall survival. In adjusted analyses, PIK3CA mutations were associated with shorter survival.
Prevention Prophylactic vaccines against HPV currently available include monova- lent (HPV16), bivalent (HPV16 and 18), and quadrivalent (HPV6, 11, 16, and 18) virus-like particle vaccines. Furthermore, clinical trials examin- ing the efficacy of a nonavalent vac- cine are under way. In large clinical trials, vaccines have shown excel- lent safety and nearly 100% efficacy in preventing persistent infections and the cervical precancers due to HPV16 and 18 [23,24]. Despite many barriers to implementation of the vaccine in resource-restricted environments, the vaccine has suc- cessfully been introduced in sev- eral developing countries [23] (see Chapter 4.6).
Early changes in the cervix, spe- cifically squamous intraepithelial le- sions, can be detected years before invasive cancer develops by screen- ing tests such as conventional cy- tology (Pap smear), liquid-based
cytology, and HPV testing (Chapter 4.7) [25]. Testing for HPV DNA of oncogenic types is gaining increas- ing interest and application in cervi- cal cancer screening. It has much greater sensitivity and only slightly lower specificity than Pap cytology. As a primary screening test followed by Pap triage of HPV-positive cases, HPV testing has the potential to im- prove the overall quality of screening programmes [23]. Women with ab- normal screening results are further investigated with colposcopy-direct- ed biopsies.
Endometrial cancer Endometrial carcinoma is a malig- nant epithelial tumour, usually exhib- iting glandular differentiation (adeno- carcinoma), capable of invading the myometrium and spreading outside the uterus. More than 80% of cases are estrogen-related, well- to mod- erately well-differentiated (endo- metrioid) adenocarcinomas, which are usually confined to the corpus uteri at diagnosis and can be cured by hysterectomy. The remaining cases, however, are non-estrogen- related high-grade carcinomas (non- endometrioid serous and clear cell adenocarcinomas) associated with poor prognosis.
Etiology Hormones play an important role in the etiology of endometrial carci- noma. The “unopposed estrogens” hypothesis is widely accepted and explains most of the risk factors for endometrial cancer: early age at menarche, late age at menopause, nulliparity, hormone replacement therapy use, and obesity.
Obesity is the most important risk factor worldwide, estimated to ac- count for 40% of endometrial cancer incidence. In premenopausal women, obesity is associated with anovula- tory cycles during which the endome- trial tissue receives continuous stim- ulation. In postmenopausal women, obesity increases the concentration of endogenous estrogens, which are mainly produced by the aromatization of androgens in the adipose tissues.
Fig. 5.12.6. Adenocarcinoma, endocervical or usual type.
Chapter 5.12 • Cancers of the female reproductive organs 475
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
Excess weight is associated with insulin resistance and chronically elevated insulin concentrations in blood, and with increasing concen- trations of sex steroids [26], factors that are associated with increased endometrial cancer risk. Type 2 and type 1 diabetes are strongly associ- ated with an increase in endometrial cancer risk.
The use of oral contraceptives is associated with a long-lasting de- crease in endometrial cancer risk, but only when they contain pro- gestogen in addition to estrogens [27]. Use of hormone replacement therapy by postmenopausal women increases the risk of endometrial cancer about 2-fold [28]. Higher en- dogenous estrogen concentrations in blood are associated with an in- crease in endometrial cancer risk mainly in postmenopausal women, whereas higher endogenous an- drogen concentrations are associ- ated with an increase in endometrial cancer risk in both premenopausal and postmenopausal women [26]. Polycystic ovary syndrome – associ- ated with increased blood androgen levels, and with infertility, amenor- rhea, hirsutism and diabetes – has been consistently associated with an increase in endometrial cancer risk. Women who develop breast cancer are at increased risk, and are more likely to develop non-endometrioid rather than endometrioid endometri- al carcinoma. Although this increase in risk could be explained partly by common risk factors between breast and endometrial malignancies, such as nulliparity or late age at meno- pause, the use of tamoxifen for the treatment of breast cancer has also been questioned: women under ta- moxifen therapy had a more than 2-fold increase in endometrial can- cer risk compared with non-users.
Pathology Endometrial carcinomas are classi- fied into two different clinicopatho- logical types (Fig. 5.12.7) [29]. More than 80% are type I tumours, which are estrogen-related, low-grade en- dometrioid carcinomas, often pre- ceded by endometrial hyperplasia.
They are usually confined to the uter- us and have a favourable outcome. In contrast, approximately 10% of cases are type II tumours, which are non-endometrioid (mainly se- rous and, less frequently, clear cell) carcinomas, arising occasionally
in endometrial polyps or atrophic endometria. Type II tumours are not associated with estrogen stimulation or hyperplasia; they often, but not always, invade the myometrium and lymphovascular spaces, and they have a high mortality rate.
Fig. 5.12.7. Adenocarcinoma of the endometrium. (A) Endometrioid carcinoma (type I). Polypoid endometrial tumour with only superficial myometrial invasion. (B) Non- endometrioid carcinoma (type II). Large haemorrhagic and necrotic tumour with deeper myometrial invasion. (C) Well-differentiated (grade 1) endometrioid adenocarcinoma. The neoplastic glands resemble normal endometrial glands. (D) Non-endometrioid (se- rous) carcinoma exhibiting stratification of markedly atypical tumour cells with numerous mitoses. (E) Endometrioid carcinoma. MLH1 inactivation by promoter hypermethylation is the most common cause of the microsatellite instability phenotype in endometrioid endometrial carcinoma. Progressive accumulation of alterations secondary to micro- satellite instability affects important regulatory genes and promotes carcinogenesis. (F) Non-endometrioid (serous) carcinoma usually shows a strong p53 immunoreaction as a result of (G) TP53 mutation.
A
C
E
B
D
F
G
476
Key surgical and pathological prognostic indicators include his- tological type, histological grade, surgical/pathological stage, depth of myometrial invasion, lymphovas- cular invasion, and cervical involve- ment [30]. In 2009, a new staging classification of endometrial car- cinoma was proposed [12]. Most endometrial cancers are sporadic, but 2–5% are familial (Table 5.12.1). Endometrial carcinoma is the most common extracolonic cancer in women with hereditary non-polypo- sis colon cancer syndrome, a defect in DNA mismatch repair that is also associated with breast and ovarian cancers.
Genetics A dualistic model of endometrial carcinogenesis has been proposed (Fig. 5.12.7) [31]. According to this model, normal endometrial cells would transform into endometrioid carcinoma through replication er- rors, so-called “microsatellite insta- bility”, and subsequent accumula- tion of mutations in oncogenes and tumour suppressor genes. In con- trast, alterations of TP53 and loss of heterozygosity on several chro- mosomes would drive the process of neoplastic transformation into the acquisition of a non-endometrioid carcinoma phenotype.
Molecular genetic alterations in type I, i.e. endometrioid, carcino- mas (Table 5.12.2) differ in some respects from those detected in type II (non-endometrioid) carci- nomas (Table 5.12.3) [31]. Non- endometrioid carcinomas may also derive from endometrioid carcinoma with microsatellite instability through
tumour progression and subsequent TP53 mutations (Fig. 5.12.8) [31].
In sporadic endometrioid car- cinomas, microsatellite instability results from promoter hypermeth- ylation of hMLH1 and leads to mutations in several target genes (containing microsatellites) that are involved in apoptosis, cell prolifera- tion, and cell differentiation. The wide range of mutations would be responsible for tumour heteroge- neity. Inactivation of the tumour suppressor gene PTEN may result from microsatellite instability (45%), promoter hypermethylation (16%), or loss of heterozygosity (24%). PTEN
inactivation releases the PI3K/AKT pathway, inhibiting apoptosis and resulting in tumour growth advan- tage. Recent genomic and related analyses have indicated that most endometrioid tumours have few copy number alterations or TP53 mutations but frequent mutations in PTEN, CTNNB1, PIK3CA, ARID1A, and KRAS, and novel mutations in the SWI/SNF chromatin remodelling complex gene ARID5B [32].
β-Catenin gene mutations occur in 20% of endometrioid carcinomas, correlate with MMP-7 and cyclin D1 overexpression, and are associated with good prognosis. Mutation of the ARID1A gene and loss of the cor- responding protein BAF250a have been found in 29–40% of endome- trioid carcinomas, 18% of serous carcinomas, and 26% of clear cell carcinomas [33,34]. In contrast, non- endometrioid carcinomas show TP53 and PPP2R1A mutations, inactiva- tion of p16 and E-cadherin, c-erbB2 amplification, STK15 alterations, and
Table 5.12.1. Hereditary syndromes with lifetime risk of endometrial cancer
Hereditary condition Gene (chromosome) Lifetime risk of endometrial cancer
Hereditary non-polyposis colorectal cancer Lynch syndrome II Muir–Torre syndrome
MSH2 (2p21) MLH1 (3p21.3) PMS2 (7p22.2) MSH6 (2p16)
40–60%
Cowden syndrome PTEN (10q23.3) 5–10%
BRCA1 syndrome BRCA1 (17q) 2–3%
Table 5.12.2. Altered genes in endometrioid endometrial carcinoma
Gene Chromosome Mechanism of alteration Percentage of
cancers
Oncogenes
PIK3CA 3q26.3 Point mutation 24–39%
KRAS 12p12.1 Point mutation 10–30%
CTNNB1 (β-catenin) 3p21 Point mutation 14– 44%
PIK3R1 5q13.1 Point mutation 43%
FGFR2 10q26 Point mutation 16%
Tumour suppressor genes
PTEN 10q23.3 Mutation, loss of heterozygosity
37–61%
ARID1A 1p35.3 Mutation 40%
MLH1 (MSI) 3p21.3 Promoter hypermeth- ylation (epigenetic silencing)
30%
TP53 17p13.1 Mutation 10–20%
CDKN2A (p16) 9p21 Loss of heterozy- gosity, epigenetic silencing
10%
SPRY2 13q31.1 Epigenetic silencing 20%
RASSF1A 3p21.3 Epigenetic silencing 48%
Chapter 5.12 • Cancers of the female reproductive organs 477
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
loss of heterozygosity at multiple loci (Table 5.12.3) [29–31].
Targeted therapy Endometrioid carcinomas with PTEN mutations have activation of the PI3K/AKT pathway and ge- nomic instability. Accordingly, these patients are treated with poly (ADP- ribose) polymerase (PARP) inhibitors [35]. Mammalian target of rapamycin (mTOR) is a kinase that regulates cell growth and apoptosis [36]. mTOR in- hibitors have recently been developed as potential anticancer agents, and PTEN-mutated tumours are particu- larly susceptible [37]. Angiogenesis is essential to the development and maintenance of any tissue. Increased production of vascular endothelial growth factor (VEGF), frequently ob- served in hypoxia or inflammation, results in increased endothelial cell proliferation and decreased apopto- sis [38]. VEGF overexpression is
associated with a poor prognosis in endometrial cancer. The role of drugs inhibiting angiogenesis, such as bevacizumab and tyrosine kinase inhibitors, is being studied.
Ovarian cancer The most common ovarian cancers are ovarian carcinomas, which are also the most lethal gynaecological malignancies. Based on histopathol- ogy and molecular genetics, ovarian carcinomas are divided into five main types: high-grade serous (70%), en- dometrioid (10%), clear cell (10%), mucinous (3%), and low-grade se- rous carcinomas (< 5%), which to- gether account for more than 95% of cases (Table 5.12.4; Fig. 5.12.9). These types are essentially distinct diseases, as indicated by differenc- es in epidemiological and genetic risk factors, precursor lesions, pat- terns of spread, molecular events during oncogenesis, response to chemotherapy, and prognosis [39]. Much less common are malignant germ cell tumours (dysgerminomas, yolk sac tumours, and immature
Fig. 5.12.8. Pathogenesis of endometrial cancer: an alternative to the dualistic model. On the basis of molecular analysis, it has been suggested that non-endometrioid carci- nomas may result from tumour progression from pre-existing endometrioid carcinomas. This would explain why these high-grade tumours often retain the molecular alterations of endometrioid carcinomas. Ca, carcinoma; MI, microsatellite instability; NE, normal endometrium.
Table 5.12.3. Altered genes in non-endometrioid endometrial carcinoma
Gene Chromosome Mechanism of alteration Percentage of
cancers
Oncogenes
PIK3CA 3q26.3 Point mutation and amplification
20–30%
PIK3R1 5q13.1 Point mutation 12%
ERBB2 (HER2/neu) 17q12 Amplification 30%
CCNE1 (cyclin E1) 19q12 Amplification 55%
STK15 (aurora kinase A)
20q13.2 Amplification 60%
CCND1 (cyclin D1) 11q13 Amplification 26%
Tumour suppressor genes
TP53 17p13.1 Mutation 90%
CDH1 (E-cadherin) 16q22.1 Loss of heterozygosity
80–90%
CDKN2A (p16) 9p21 Loss of heterozy- gosity, epigenetic silencing
40%
PPP2R1A 19q13.41 Mutation 17– 41%
478
teratomas) (3% of ovarian cancers) and potentially malignant sex cord stromal tumours (1–2%), the most common of which are granulosa cell tumours.
Etiology Ovarian carcinomas most commonly affect nulliparous women and oc- cur least frequently in women with suppressed ovulation, typically by pregnancy or oral contraceptives. These tumours are generally con- sidered to originate from the cells covering the ovarian surface or the pelvic peritoneum. Malignant trans- formation of this mesothelium has been explained by the “incessant ovulation” theory [40]. Irritants, such as talc or asbestos, have also been implicated, as have hormonal fac- tors. Elevated gonadotropin levels may contribute to malignant transfor- mation. Pregnancy and oral contra- ceptives are protective factors, and both cause a reduction in pituitary gonadotropins.
Family history of ovarian cancer accounts for 10% of cases; the risk is increased 3-fold when two or more first-degree relatives have been
affected. Women with germline mu- tations in BRCA1 or BRCA2 have a 30–70% risk of developing ovarian cancer, mainly high-grade serous carcinomas, by age 70 [41]. BRCA1 and BRCA2 are essential compo- nents of the homologous recombi- nation DNA repair system, required to repair DNA double-strand breaks [42]. Women with hereditary non- polyposis colon cancer are also at greater risk for ovarian cancer, spe- cifically endometrioid carcinoma.
The cell descriptors used to clas- sify ovarian carcinomas – serous, mucinous, endometrioid, clear cell, transitional, and squamous – do not apply to cells in the normal ovary, and malignancy has long been at- tributed to Müllerian “neometapla- sia” of the ovarian surface epithelium (mesothelium). During embryonic life, the coelomic cavity is lined by mesothelium, which also covers the gonadal ridge and gives rise to Müllerian ducts, from which the fallopian tubes, uterus, and vagina develop. Thus, the tumour cells would resemble morphologically the epithelia of the fallopian tube, en- dometrium, or endocervix. Although the mesothelial origin cannot be
excluded, there is evidence that sev- eral primary ovarian cancers have originated in other pelvic organs and involve the ovary secondarily. High- grade serous carcinomas, especially in BRCA-positive patients, have been postulated to arise from pre- cursor epithelial lesions in the distal fimbriated end of the fallopian tube, whereas endometrioid and clear cell carcinomas originate from ovarian endometriosis [43].
Pathology and genetics Serous carcinoma High-grade serous carcinoma and low-grade serous carcinoma (Fig. 5.12.9, A and B) are fundamentally different tumour types. Low-grade serous carcinomas are usually as- sociated with a non-invasive serous borderline component, carry KRAS and BRAF mutations, and are unre- lated to TP53 mutations and BRCA abnormalities. In contrast, high- grade serous carcinomas are not associated with serous borderline tumours and typically exhibit TP53 mutations and BRCA abnormalities (Table 5.12.4) [44]. Most patients with high-grade serous carcinomas
Table 5.12.4. Characteristics of the main types of ovarian carcinoma
Characteristic Type
High-grade serous Low-grade serous Mucinous Endometrioid Clear cell
Usual stage at diagnosis
Advanced Early or advanced Early Early Early
Presumed tissue of origin/precursor lesion
Fallopian tube or tubal metaplasia in inclusions of ovarian surface epithelium
Serous borderline tumour
Adenoma–border- line tumour–carci- noma sequence; teratoma
Endometriosis, adenofibroma
Endometriosis, adenofibroma
Genetic risk BRCA1/2 ? ? Hereditary non- polyposis colorectal cancer
?
Significant molecular abnormalities
p53 and pRb pathways
BRAF or KRAS KRAS PTEN, β-catenin, ARID1A, PIK3CA
HNF-1β, ARID1A, PIK3CA
KRAS Microsatellite
instability
Proliferation High Low Intermediate Low Low
Response to primary chemotherapy
80% 26–28% 15% ? 15%
Prognosis Poor Favourable Favourable Favourable Intermediate
Chapter 5.12 • Cancers of the female reproductive organs 479
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
(> 80%) present with advanced dis- ease (stage III).
The genomics of serous ovar- ian adenocarcinomas has been as- sessed using DNA sequences of exons from coding genes in 316 of these tumours and other data [45]. High-grade serous ovarian cancer is characterized by TP53 mutations in almost all tumours (96%); low-prev- alence but statistically recurrent so- matic mutations in nine further genes, including NF1, BRCA1, BRCA2, RB1, and CDK12; 113 significant focal DNA copy number aberrations; and pro- moter methylation events involving 168 genes. Analyses delineated four ovarian cancer transcriptional sub- types, three microRNA subtypes, four promoter methylation subtypes, and
a transcriptional signature associated with survival duration. Other data in- volved the impact that tumours with BRCA1 or BRCA2 and CCNE1 ab- errations have on survival. Pathway analyses suggested that homologous recombination is defective in about half of the tumours analysed, and that NOTCH and FOXM1 signalling are involved in serous ovarian cancer pathophysiology.
Consistent patterns of chromo- somal change in patient cohorts suggest an underlying interdepen- dency of gains and losses of par- ticular genes in individual tumours [46], for example co-amplification of CCNE1 and the 20q11 locus involv- ing the cell-cycle regulator TPX2, among other genes. Genome-wide
association studies continue to identify susceptibility loci for ovar- ian cancer, specifically including loci at 2q31 and 8q24 [47].
Endometrioid carcinoma Endometrioid carcinoma, which re- sembles its endometrial counterpart (Fig. 5.12.9C), is thought to arise by malignant transformation of endo- metriosis and not from the ovarian surface epithelium. The most com- mon genetic abnormalities in ovar- ian endometrioid carcinoma are somatic mutations of the ARID1A, CTNNB1 (β-catenin), and PTEN genes, and microsatellite instabil- ity (Table 5.12.4) [39,48]. Between 15% and 20% of patients with en- dometrioid carcinoma of the ovary also have endometrial cancer. If ovarian and endometrial cancers coexist, they generally arise inde- pendently, although some may be metastases from one or the other, a distinction with important prognostic implications.
Mucinous carcinoma Mucinous ovarian tumours are of- ten heterogeneous. Benign, bor- derline, non-invasive, and invasive carcinoma components may coex- ist within the same tumour. Such a morphological continuum suggests that tumour progression occurs from cystadenoma and borderline tumour to invasive carcinomas. This hypothesis is supported by KRAS mutations in mucinous tumours: 56% of cystadenomas and 85% of carci- nomas express mutated KRAS, and borderline tumours are intermediate (Table 5.12.4) [39]. Mucinous carci- nomas are usually large, unilateral cystic masses (Fig. 5.12.9D). The finding of bilateral mucinous tumours should raise suspicion of metastases from a far more frequent mucinous carcinoma elsewhere (e.g. gastroin- testinal tract).
Clear cell carcinoma This enigmatic ovarian cancer is closely related to endometrioid adenocarcinoma, and often occurs in association with endometriosis.
Fig. 5.12.9. Representative examples of the five main types of ovarian carcinoma, which together account for 98% of cases: (A) high-grade serous carcinoma, (B) low- grade serous carcinoma, (C) endometrioid carcinoma, (D) mucinous carcinoma, and (E) clear cell carcinoma.
A
C
E
B
D
480
It usually occurs after menopause. The most common genetic abnor- malities are somatic mutations of the ARID1A, PTEN, and PIK3CA genes (Table 5.12.4) [39,48]. Clear cell carcinomas of the ovary resem- ble their counterparts in the vagina, cervix, and corpus uteri; they show sheets or tubules of malignant cells with clear cytoplasm (Fig. 5.12.9E) or tubules lined by “hobnail” cells.
Prognosis For patients with malignant ovarian tumours, survival is generally poor. The most important prognostic index is the surgical stage of the tumour at the time it is detected. A new staging classification of cancer of the ovary, fallopian tube, and peritoneum has been proposed by the International Federation of Gynecology and Obstetrics [49].
Targeted therapy The most promising targets in clinical trials are angiogenesis and homolo- gous recombination deficiency. To select patients for trials investigating these targets, predictive biomarkers are required. Other promising tar- gets currently being studied based on ovarian cancer biology include folate receptor, PI3K/AKT, and Ras/ Raf/MEK pathways [50].
References
1. IARC (2012). Biological agents. IARC Monogr Eval Carcinog Risks Hum, 100B:1– 441. PMID:23189750
2. De Vuyst H, Clifford GM, Nascimento MC et al. (2009). Prevalence and type distribution of human papillomavirus in carcinoma and intraepithelial neoplasia of the vulva, vagina and anus: a meta-analysis. Int J Cancer, 124:1626–1636. http://dx.doi.org/10.1002/ ijc.24116 PMID:19115209
3. Kurman RJ, Toki T, Schiffman MH (1993). Basaloid and warty carcinomas of the vul- va. Distinctive types of squamous cell car- cinoma frequently associated with human papillomaviruses. Am J Surg Pathol, 17:133– 145. http://dx.doi.org/10.1097/00000478- 199302000-00005 PMID:8380681
4. Centers for Disease Control and Prevention (CDC) (2012). Human papillomavirus-asso- ciated cancers - United States, 2004–2008. MMWR Morb Mortal Wkly Rep, 61:258– 261. PMID:22513527
5. Darragh TM, Colgan TJ, Cox JT et al.; Members of LAST Project Work Groups (2012). The lower anogenital squamous terminology standardization project for HPV-associated lesions: background and consensus recommendations from the College of American Pathologists and the American Society for Colposcopy and Cervical Pathology. J Low Genit Tract Dis, 16:205–242. http://dx.doi.org/10.1097/LGT. 0b013e31825c31dd PMID:22820980
6. Skapa P, Zamecnik J, Hamsikova E et al. (2007). Human papillomavirus (HPV) pro- files of vulvar lesions: possible implications for the classification of vulvar squamous cell carcinoma precursors and for the ef- ficacy of prophylactic HPV vaccination. Am J Surg Pathol, 31:1834–1843. http://dx.doi. org/10.10 97/PAS.0b013e318 0 6 8 6d10 PMID:18043037
7. van de Nieuwenhof HP, van der Avoort IAM, de Hullu JA (2008). Review of squa- mous premalignant vulvar lesions. Crit Rev Oncol Hematol, 68:131–156. http://dx.doi. org /10.1016/ j.cr itrevonc. 20 0 8.02.012 PMID:18406622
8. Pinto AP, Miron A, Yassin Y et al. (2010). Differentiated vulvar intraepithelial neo- plasia contains Tp53 mutations and is genetically linked to vulvar squamous cell carcinoma. Mod Pathol, 23:404–412. http:// dx.doi.org/10.1038/modpathol.2009.179 PMID:20062014
9. van de Nieuwenhof HP, Massuger LF, van der Avoort IA et al. (2009). Vulvar squamous cell carcinoma development after diagnosis of VIN increases with age. Eur J Cancer, 45:851–856. http://dx.doi.org/10.1016/j.ejca. 2008.11.037 PMID:19117749
10. Yoder BJ, Rufforny I, Massoll NA, Wilkinson EJ (2008). Stage IA vulvar squamous cell carcinoma: an analysis of tumor invasive characteristics and risk. Am J Surg Pathol, 32:765–772. http://dx.doi.org/10.1097/PAS. 0b013e318159a2cb PMID:18379417
11. Report M (2009). The new FIGO staging system for cancers of the vulva, cervix, en- dometrium and sarcomas. Gynecol Oncol, 115:325–328. http://dx.doi.org/10.1016/j.y gyno.2009.10.050
12. McCluggage WG (2009). Recent de- velopments in vulvovaginal pathology. Histopathology, 54:156–173. http://dx.doi. org/10.1111/j.1365 -2559.2008.03098.x PMID:18637148
13. de Sanjosé S, Diaz M, Castellsagué X et al. (2007). Worldwide prevalence and genotype distribution of cervical human papillomavirus DNA in women with normal cytology: a meta-analysis. Lancet Infect Dis, 7:453–459. http://dx.doi.org/10.1016/ S1473-3099(07)70158-5 PMID:17597569
14. Franceschi S, Herrero R, Clifford GM et al. (2006). Variations in the age-specific curves of human papillomavirus preva- lence in women worldwide. Int J Cancer, 119:2677–2684. http://dx.doi.org/10.1002/ ijc.22241 PMID:16991121
15. Schiffman M, Castle PE, Jeronimo J et al. (2007). Human papillomavirus and cervical cancer. Lancet, 370:890–907. http://dx.doi. org /10.1016/S0140 - 673 6(07) 61416 - 0 PMID:17826171
16. Rodríguez AC, Schiffman M, Herrero R et al. (2010). Longitudinal study of human pap- illomavirus persistence and cervical intra- epithelial neoplasia grade 2/3: critical role of duration of infection. J Natl Cancer Inst, 102:315–324. http://dx.doi.org/10.1093/jnci/ djq001 PMID:20157096
17. Guan P, Howell-Jones R, Li N et al. (2012). Human papillomavirus types in 115,789 HPV-positive women: a meta-analysis from cervical infection to cancer. Int J Cancer, 131:2349–2359. http://dx.doi.org/10.1002/ ijc.27485 PMID:22323075
18. Lehtinen M, Paavonen J, Wheeler CM et al.; HPV PATRICIA Study Group (2012). Overall efficacy of HPV-16/18 AS04-adjuvanted vaccine against grade 3 or greater cer- vical intraepithelial neoplasia: 4-year end-of-study analysis of the randomised, double-blind PATRICIA trial. Lancet Oncol, 13:89–99. http://dx.doi.org/10.1016/S1470- 2045(11)70286-8 PMID:22075171
19. Veldhuijzen NJ, Snijders PJ, Reiss P et al. (2010). Factors affecting transmission of mucosal human papillomavirus. Lancet Infect Dis, 10:862–874. http://dx.doi.org/ 10.1016/S1473-3099(10)70190-0 PMID: 21075056
20. Shi Y, Li L, Hu Z et al. (2013). A genome- wide association study identifies two new cervical cancer susceptibility loci at 4q12 and 17q12. Nat Genet, 45:918–922. http://dx.doi. org/10.1038/ng.2687 PMID:23817570
21. Chen D, Juko-Pecirep I, Hammer J et al. (2013). Genome-wide association study of susceptibility loci for cervical cancer. J Natl Cancer Inst, 105:624–633. http://dx.doi.org/ 10.1093/jnci/djt051 PMID:23482656
22. Wright AA, Howitt BE, Myers AP et al. (2013). Oncogenic mutations in cervical cancer: genomic differences between adenocarcinomas and squamous cell car- cinomas of the cervix. Cancer, 119:3776– 3783. http://dx.doi.org/10.1002/cncr.28288 PMID:24037752
Chapter 5.12 • Cancers of the female reproductive organs 481
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
2
23. Franco EL, Coutlée F, Ferenczy A (2009). Integrating human papillomavirus vaccina- tion in cervical cancer control programmes. Public Health Genomics, 12:352–361. ht t p: //d x.d o i.o r g /10.115 9 / 0 0 0 214 9 25 PMID:19684447
24. Van de Velde N, Boily MC, Drolet M et al. (2012). Population-level impact of the biva- lent, quadrivalent, and nonavalent human papillomavirus vaccines: a model-based analysis. J Natl Cancer Inst, 104:1712–1723. ht t p: //d x.d o i.o r g /10.10 9 3 / jnc i /djs 3 9 5 PMID:23104323
25. Sankaranarayanan R, Gaffikin L, Jacob M et al. (2005). A critical assessment of screening methods for cervical neoplasia. Int J Gynaecol Obstet, 89 Suppl 2:S4–S12. http://dx.doi.org/10.1016/j.ijgo.2005.01.009 PMID:15823266
26. Kaaks R, Lukanova A, Kurzer MS (2002). Obesity, endogenous hormones, and endometrial cancer risk: a synthetic re- view. Cancer Epidemiol Biomarkers Prev, 11:1531–1543. PMID:12496040
27. Cogliano V, Grosse Y, Baan R et al.; WHO International Agency for Research on Cancer (2005). Carcinogenicity of com- bined oestrogen-progestagen contracep- tives and menopausal treatment. Lancet Oncol, 6:552–553. http://dx.doi.org/10.1016/ S1470-2045(05)70273-4 PMID:16094770
28. Beral V, Bull D, Reeves G; Million Women Study Collaborators (2005). Endometrial cancer and hormone-replacement ther- apy in the Million Women Study. Lancet, 365:1543–1551. http://dx.doi.org/10.1016/ S0140-6736(05)66455-0 PMID:15866308
29. Bokhman JV (1983). Two pathogenetic types of endometrial carcinoma. Gynecol Oncol, 15:10–17. http://dx.doi.org/10.1016/0090-82 58(83)90111-7 PMID:6822361
30. Prat J (2004). Prognostic parameters of endometrial carcinoma. Hum Pathol, 35:649–662. http://dx.doi.org/10.1016/j.hum path.2004.02.007 PMID:15188130
31. Yeramian A, Moreno-Bueno G, Dolcet X et al. (2013). Endometrial carcinoma: molecu- lar alterations involved in tumor develop- ment and progression. Oncogene, 32:403– 413.http://dx.doi.org/10.1038/onc.2012.76 PMID:22430211
32. Kandoth C, Schultz N, Cherniack AD et al.; Cancer Genome Atlas Research Network (2013). Integrated genomic characterization of endometrial carcinoma. Nature, 497:67– 73. http://dx.doi.org/10.1038/nature12113 PMID:23636398
33. Guan B, Mao TL, Panuganti PK et al. (2011). Mutation and loss of expression of ARID1A in uterine low-grade endometrioid carcinoma. Am J Surg Pathol, 35:625– 632. http://dx.doi.org/10.1097/PAS.0b013 e318212782a PMID:21412130
34. Wiegand KC, Lee AF, Al-Agha OM et al. (2011). Loss of BAF250a (ARID1A) is fre- quent in high-grade endometrial carcino- mas. J Pathol, 224:328–333. http://dx.doi. org/10.1002/path.2911 PMID:21590771
35. Dedes KJ, Wetterskog D, Mendes-Pereira AM et al. (2010). PTEN deficiency in endo- metrioid endometrial adenocarcinomas pre- dicts sensitivity to PARP inhibitors. Sci Transl Med, 2:53ra75. http://dx.doi.org/10.1126/sci translmed.3001538 PMID:20944090
36. Bansal N, Yendluri V, Wenham RM (2009). The molecular biology of endometrial can- cers and the implications for pathogene- sis, classification, and targeted therapies. Cancer Control, 16:8–13. PMID:19078924
37. Slomovitz BM, Lu KH, Johnston T et al. (2010). A phase 2 study of the oral mamma- lian target of rapamycin inhibitor, everolimus, in patients with recurrent endometrial carci- noma. Cancer, 116:5415–5419. http://dx.doi. org/10.1002/cncr.25515 PMID:20681032
38. Ferrara N, Gerber HP, LeCouter J (2003). The biology of VEGF and its receptors. Nat Med, 9:669–676. http://dx.doi.org/10.1038/ nm0603-669 PMID:12778165
39. Prat J (2012). Ovarian carcinomas: five dis- tinct diseases with different origins, genetic alterations, and clinicopathological fea- tures. Virchows Arch, 460:237–249. http:// dx.doi.org/10.1007/s00428- 012-1203-5 PMID:22322322
40. La Vecchia C (2001). Epidemiology of ovarian cancer: a summary review. Eur J Cancer Prev, 10:125–129. http://dx.doi. org/10.1097/00008469-200104000-00002 PMID:11330452
41. Risch HA, McLaughlin JR, Cole DE et al. (2006). Population BRCA1 and BRCA2 mu- tation frequencies and cancer penetrances: a kin-cohort study in Ontario, Canada. J Natl Cancer Inst, 98:1694–1706. http://dx.doi. org/10.1093/jnci/djj465 PMID:17148771
42. Venkitaraman AR (2009). Linking the cel- lular functions of BRCA genes to cancer pathogenesis and treatment. Annu Rev Pathol, 4:461–487. http://dx.doi.org/10.1146/ annurev.pathol.3.121806.151422 PMID: 18954285
43. Lee Y, Miron A, Drapkin R et al. (2007). A candidate precursor to serous carcinoma that originates in the distal fallopian tube. J Pathol, 211:26–35. http://dx.doi.org/10. 1002/path.2091 PMID:17117391
44. Singer G, Stöhr R, Cope L et al. (2005). Patterns of p53 mutations separate ovar- ian serous borderline tumors and low- and high-grade carcinomas and provide sup- port for a new model of ovarian carcino- genesis: a mutational analysis with immu- nohistochemical correlation. Am J Surg Pathol, 29:218–224. http://dx.doi.org/10. 1097/01.pas.0000146025.91953.8d PMID: 15644779
45. Cancer Genome Atlas Research Network (2011). Integrated genomic analyses of ovarian carcinoma. Nature, 474:609–615. PMID:21720365
46. Etemadmoghadam D, George J, Cowin PA et al.; Australian Ovarian Cancer Study Group (2010). Amplicon-dependent CCNE1 expression is critical for clonogenic survival after cisplatin treatment and is correlated with 20q11 gain in ovarian cancer. PLoS One, 5:e15498. http://dx.doi.org/10.1371/ journal.pone.0015498 PMID:21103391
47. Goode EL, Chenevix-Trench G, Song H et al.; Wellcome Trust Case-Control Consortium; Australian Cancer Study (Ovarian Cancer); Australian Ovarian Cancer Study Group; Ovarian Cancer Association Consortium (OCAC); Ovarian Cancer Association Consortium (OCAC) (2010). A genome- wide association study identifies suscep- tibility loci for ovarian cancer at 2q31 and 8q24. Nat Genet, 42:874–879. http://dx.doi. org/10.1038/ng.668 PMID:20852632
48. Wiegand KC, Shah SP, Al-Agha OM et al. (2010). ARID1A mutations in endome- triosis-associated ovarian carcinomas. N Engl J Med, 363:1532–1543. http://dx. doi.org/10.1056/NEJMoa1008433 PMID: 20942669
49. Prat J; for the FIGO Committee on Gynecologic Oncology (2013). Staging classi- fication for cancer of the ovary, fallopian tube, and peritoneum. Int J Gynecol Obstet, [epub ahead of print]. http://dx.doi.org/10.1016/j. ijgo.2013.10.001 PMID:24219974
50. Ledermann JA, Marth C, Carey MS et al.; Gynecologic Cancer InterGroup (2011). Role of molecular agents and targeted therapy in clinical trials for women with ovarian cancer. Int J Gynecol Cancer, 21:763–770. http:// dx.doi.org/10.1097/IGC.0b013e31821b2669 PMID:21543938
Websites Cancer.Net. Uterine Cancer: http://www.cancer.net/cancer-types/ uterine-cancer/
MD Anderson Cancer Center. Endometrial Cancer: http://www.mdanderson.org/diseases/ hereditarygyn
National Cancer Institute. Cervical Cancer: http://www.cancer.gov/cancertopics/types/ cervical
National Cancer Institute. Endometrial Cancer: http://www.cancer.gov/cancertopics/types/ endometrial
National Cancer Institute. Ovarian Cancer: http://www.cancer.gov/cancertopics/types/ ovarian
National Cancer Institute. Vulvar Cancer: http://www.cancer.gov/cancertopics/types/ vulvar
National Institutes of Health. Vulvar Cancer: http://health.nih.gov/topic/VulvarCancer
Ovarian Cancer Research Program of British Columbia: http://www.ovcare.ca/
SEER Stat Fact Sheets. Cervix Uteri Cancer: http://seer.cancer.gov/statfacts/html/cervix. html#risk
SEER Stat Fact Sheets. Vulvar Cancer: http://seer.cancer.gov/statfacts/html/vulva. html#risk
482
Summary • Most non-Hodgkin lymphomas
are derived from either mature B or T lymphocytes, and they are morphologically, functionally, and genetically heterogeneous, with lymphoblastic malignancies being derived from precursor lymphoid cells.
• B-cell lymphomas are much more common than T-cell lym- phomas, which account for only about 10% of cases.
• Many B-cell lymphomas are characterized by recurrent chro- mosomal translocations that tar- get the immunoglobulin genes, whereas the molecular patho- genesis of most T-cell lympho- mas is as yet undiscovered.
• The classification of acute mye- loid leukaemias is largely based on recurrent genetic abnormali- ties, which have a major impact on prognosis and response to therapy.
• Chronic myeloid leukaemia is a myeloproliferative neoplasm that arises in a pluripotent stem cell and is invariably linked to the BCR-ABL1 fusion.
• The BCR-ABL1-negative my- eloproliferative neoplasms are
5.13 5 ORGAN SITE
Haematopoietic and lymphoid malignancies Elaine S. Jaffe Steven H. Swerdlow James W. Vardiman
Daniel A. Arber (reviewer) Eve Roman (reviewer)
heterogeneous in terms of the cellular targets – for example myeloid, erythroid, or megakar- yocytic – and are commonly as- sociated with mutations of JAK2.
Haematological malignancies in- clude all cell types derived from a pluripotent bone marrow stem cell, and as such include neoplasms of the lymphoid, myeloid, mast cell, histiocytic, and dendritic cell line- ages. The current approach to the classification of haematopoietic and lymphoid malignancies is based on the integration of morphological, phenotypic, genetic, and clinical features, which allows the identi- fication of distinct disease entities [1]. Most lymphoid malignancies have been related to normal coun- terparts in the B-cell (Fig. 5.13.1) and T-cell (Fig. 5.13.2) systems. For acute myeloid leukaemias, genet- ics has taken priority over aspects of cellular differentiation. In recent years there has been great prog- ress in the application of genomic methods, which has advanced our understanding of these neoplasms, led to the recognition of new entities, and pointed the way towards prom- ising new therapies, exploiting the knowledge of molecular pathways of transformation.
Chronic lymphocytic leukaemia/small lymphocytic lymphoma Chronic lymphocytic leukaemia (CLL), the most common adult leukaemia in industrialized countries, is a small B-cell lymphoid neoplasm composed of monoclonal memory B cells that most typically express CD23 and the T-cell-associated antigen CD5.
Etiology and pathology CLL/small lymphocytic lymphoma includes a more aggressive sub- set, with unmutated immunoglobu- lin heavy-chain genes, and a less aggressive subset, with mutated genes. The diagnosis requires more than 5 × 109 peripheral blood CLL- type cells/l, cytopenias or disease- related symptoms, or extramedullary disease. Individuals who do not fulfil any of these criteria but who have monoclonal CLL-type B cells are now considered to have monoclo- nal B-cell lymphocytosis (MBL) [2]. MBL can be divided into low-count and high-count MBL; cases of high- count MBL progress to frank CLL at a rate of 1–2% per year [2]. Low- count MBL (< 56 clonal B cells/μl) is found with sensitive techniques at a high frequency among older adults in the general population, and has a minimal propensity to progress. MBL is thought to be secondary to prolonged antigenic stimulation.
Chapter 5.13 • Haematopoietic and lymphoid malignancies 483
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
3
Epidemiology
Lymphomas • There were almost 566 000 new
cases of lymphoma worldwide in 2012, and about 305 000 deaths.
• The ICD-10 codes used for lym- phomas are: Hodgkin lymphoma (C81); non-Hodgkin lymphoma (C82-85, C96); multiple myeloma and immunoproliferative diseas- es (C88+C90).
• When the main entities (Hodgkin lymphoma, non-Hodgkin lym- phoma, and multiple myeloma) are examined separately, the respective cancer types do not rank highly, yet in combination, lymphomas would rank as the seventh most common form of cancer in terms of cancer incidence.
• Incidence rates for lymphoma tend to be elevated in the most developed areas, including North America, Australia and New Zealand, and northern and western Europe. In global terms, the incidence is greatest in Israel, followed by Australia and the USA.
• Where such data series are available, both incidence and mortality rates of non-Hodgkin lymphoma have tended to in- crease rather rapidly up until the mid-1990s in both sexes. Subsequently, incidence rates have stabilized, while mortal- ity rates have declined in some settings.
Map 5.13.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for lymphoma in men, 2012.
Map 5.13.2. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for lymphoma in men, 2012.
MBL with an atypical CLL-type phe- notype and a non-CLL phenotype are also recognized. There are no known etiological factors for CLL; however, there is a familial predis- position in 5 –10% of cases.
Genetics Genome-wide association studies have revealed multiple susceptibility
loci for CLL, many with proximity to genes involved in apoptosis [3]. There is great interest in prognostica- tion in CLL, which, in addition to mu- tational status, is based on clinical, phenotypic (CD38, ZAP-70 expres- sion), and molecular/cytogenetic fac- tors, particularly TP53 deletions and mutations. Recently, next-generation sequencing studies have demonstrat- ed numerous recurrent mutations in
genes that are part of varied genetic pathways – each, however, present in only up to about 10–15% of cases. Among those receiving the most at- tention are activating NOTCH1 mu- tations and mutations in the splicing factor SF3B1, both adverse prognos- tic indicators [4,5]. The hope is that this new knowledge may lead to more targeted and effective therapies.
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
484
Chart 5.13.1. Estimated global number of new cases and deaths with proportions by major world regions, for lymphoma in both sexes combined, 2012.
Chart 5.13.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for lymphoma in men, circa 1975–2012.
Chart 5.13.3. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for lymphoma in men, circa 1975–2012.
Chapter 5.13 • Haematopoietic and lymphoid malignancies 485
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
3 Hairy cell leukaemia Etiology and pathology Hairy cell leukaemia is a rare neo- plasm, typically of the post-germinal centre B-cell type, characterized by diffuse marrow and splenic red pulp involvement as well as often modest peripheral blood involvement by rela- tively small B cells with oval to re- niform nuclei and prominent “hairy” cytoplasmic projections. Hairy cell
leukaemia is diagnosed based on its characteristic morphological ap- pearance and its CD103+, CD25+, CD11c+, annexin A1+ phenotype. The etiology of hairy cell leukaemia is unknown.
Genetics Virtually all cases have BRAF V600E mutations, an abnormality shared with other malignancies such as
papillary thyroid carcinoma and mela- noma but found in few other B-cell neoplasms (and not in the variant form of hairy cell leukaemia) [6]. This probable driver mutation leads to constitutive activation of the RAF/ MEK/ERK mitogen-activated protein kinase pathway. Although most pa- tients with hairy cell leukaemia do well after treatment with purine ana- logues, BRAF V600E mutations pro- vide another therapeutic target with
Map 5.13.3. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for leukaemia in men, 2012.
Map 5.13.4. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for leukaemia in men, 2012.
Leukaemias • There were almost 352 000
new cases of leukaemia glob- ally in 2012, and about 265 000 deaths. The disease ranks as the 11th most frequent in terms of cancer incidence and the 10th most common cause of cancer death.
• The ICD-10 codes used for leu- kaemias are C91-95.
• As with lymphomas, the inci- dence rates of leukaemias tend to be highest in countries that have attained the highest levels of human development.
• Incidence rates are elevated in countries with high or very high levels of human develop- ment, including Australia and New Zealand, North America, and much of Europe. There is less variation in mortality, al- though rates are high in certain countries in North Africa and in Central and West Asia.
• Trends are somewhat difficult to decipher for leukaemia, but ob- served incidence rates are gen- erally stable with time, at least relative to trends in non-Hodg- kin lymphoma. Mortality rates have fallen for certain subtypes in higher-income countries due to improving therapeutics.
486
Chart 5.13.4. Estimated global number of new cases and deaths with proportions by major world regions, for leukaemia in both sexes combined, 2012.
Chart 5.13.5. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for leukaemia in men, circa 1975–2012.
Chart 5.13.6. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for leukaemia in men, circa 1975–2012.
Chapter 5.13 • Haematopoietic and lymphoid malignancies 487
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
3
BRAF inhibitors, already explored in other malignancies.
Plasma cell myeloma Plasma cell myeloma, sometimes referred to as multiple myeloma, is a bone marrow-based, disseminated neoplasm composed of monoclonal post-germinal centre long-lived plasma cells, which must be distin- guished from monoclonal gammop- athy of undetermined significance and lymphomas with plasmacytic differentiation.
Etiology and pathology Symptomatic plasma cell myelo- ma requires a serum and/or urine monoclonal paraprotein, clonal
plasma cells in the bone marrow, or a plasmacytoma and related end- organ damage (hypercalcaemia, renal insufficiency, anaemia, bone lesions). Asymptomatic (smoulder- ing) myeloma requires a serum paraprotein level of more than 30 g/l and/or at least 10% clonal bone mar- row plasma cells and no myeloma- associated end-organ damage or myeloma-associated symptoms. Other clinical variants include non- secretory myelomas, most of which are still associated with serum-free light-chain abnormalities, and plas- ma cell leukaemia, in which there are more than 2 × 109 circulating peripheral blood plasma cells/l or they make up more than 20% of all
leukocytes. The etiology of plasma cell myeloma is totally unknown in most patients. The disease is more common in African Americans than in Caucasians.
Genetics Upregulation of one of the cyclin D genes due to hyperdiploidy or translocations involving the immuno- globulin heavy-chain gene and the CCND1, C-MAF, FGFR3/MMSET, CCND3, or MAFB genes is consid- ered an important early event but is also found in monoclonal gammop- athy of undetermined significance. Other important early events seen in a significant minority of patients include monosomy 13 or 13q14
Fig. 5.13.1. Diagrammatic representation of B-cell differentiation. B-cell neoplasms correspond to stages of maturation. Most B cells are activated within the germinal centre, but T-cell independent activation can take place outside the germinal centre. This fact leads to cases of chronic lymphocytic leukaemia (CLL) that may or may not show evidence of somatic hypermutation. AG, antigen; DLBCL, diffuse large B-cell lymphoma; FDC, follicular dendritic cell; HCL, hairy cell leukaemia; GC, germinal centre; Ig, immunoglobulin; MALT, mucosa-associated lymphoid tissue; SLL, small lymphocytic lymphoma.
488
deletion and K-RAS or N-RAS acti- vating mutations. A variety of other genetic and epigenetic events and pathways are implicated in disease progression, including the recent identification of therapeutically tar- getable BRAF mutations in a small proportion of cases [7].
Follicular lymphoma Follicular lymphoma, one of the most common adult lymphomas in indus- trialized countries, is a neoplasm of germinal centre B cells with vary- ing proportions of centrocytes and transformed centroblasts and almost always at least a partially follicular growth pattern.
Etiology, pathology, and genetics Follicular lymphoma is diagnosed based on its morphological features, phenotype (CD20+, usually CD10+, BCL6+, and usually BCL2+), and fre- quent IGH/BCL2 translocations or,
much less commonly, BCL6 translo- cations. This disease is graded based on the number of centroblasts pres- ent. The more common grade 1–2 follicular lymphoma is an indolent but generally widely disseminated and in- curable neoplasm, and the grade 3 cases (A and B, based on the pres- ence or absence of centrocytes) are considered more aggressive. The eti- ology is uncertain, but healthy individ- uals with pesticide exposure have a greater prevalence of circulating cells with IGH/BCL2 translocations.
Typical cases need to be dis- tinguished from paediatric follic- ular lymphoma, a lymphoma that also occurs particularly in younger adults, based on morphological, im- munophenotypic, and cytogenetic/ molecular differences [8], and from in situ follicular lymphoma/intrafol- licular neoplasia/follicular lympho- ma-like B cells of uncertain/unde- termined significance, where there are scattered follicles with follicular lymphoma-like B cells but with an
intact underlying lymphoid tissue architecture [9]. Primary duodenal follicular lymphoma is also distinc- tive and often localized. Primary cu- taneous follicular lymphoma belongs in the separate category of primary cutaneous follicle centre lymphoma. Most cases have well-recognized secondary cytogenetic/molecular abnormalities, but the critical trans- formational events required in addi- tion to BCL2 translocation remain to be established.
Mantle cell lymphoma Mantle cell lymphoma is a mature B-cell lymphoma believed to be de- rived from usually naive B cells of the inner mantle zone.
Etiology and pathology The etiology of mantle cell lympho- ma is unknown. In general it is a very aggressive and incurable lymphoma, but the clinical spectrum of mantle cell lymphoma is very broad. Both
Fig. 5.13.2. Diagrammatic representation of T-cell differentiation. T-cell neoplasms correspond to different stages of maturation. Mature T cells include αβ and γδ T cells, both of which mature in the thymus gland. T-follicular helper (TFH) cells are a recently well-characterized subset of αβ T cells involved in many nodal peripheral T-cell lymphomas. AG, antigen; FDC, follicular dendritic cell; NK, natural killer.
Chapter 5.13 • Haematopoietic and lymphoid malignancies 489
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
3
aggressive (blastoid and pleomor- phic) and more indolent variants are recognized. One indolent and very distinctive variant is associated with blood, bone marrow, and sometimes splenic involvement without periph- eral adenopathy and typically has mutated immunoglobulin heavy- chain genes and few secondary cy- togenetic/molecular abnormalities. The other variant that may not rep- resent an overt malignancy is in situ mantle cell lymphoma/mantle cell lymphoma-like B cells of undeter- mined/uncertain significance. These cases lack architectural destruction and show partially infiltrated follicu- lar mantle zones, often at the mantle zone/germinal centre interface.
Genetics Most cases have a CCND1 transloca- tion and cyclin D1 expression, an ear- ly but insufficient pathogenetic event and a useful diagnostic tool. Rare cases are cyclin D1-negative but, like most other MCLs, are SOX11+ and usually express cyclin D2 or D3. Otherwise, the diagnosis is made based on a characteristic morpholog- ical appearance and a usual CD20+, CD5+, CD23– phenotype. In addition to CCND1 overexpression, mantle cell lymphoma is characterized by additional abnormalities leading to cell-cycle dysregulation, and assess- ment of the proliferative fraction is an extremely important prognostic factor [10]. Disruption of the DNA damage response pathways and activation of cell survival mechanisms are also important in the pathogenesis of this disease, with many potentially target- able abnormal signalling pathways. For example, NOTCH1 mutations similar to those seen in CLL/small lymphocytic lymphoma, and associ- ated with more aggressive disease, were recently reported in 12% of mantle cell lymphomas.
Marginal zone lymphoma and lymphoplasmacytic lymphoma These diseases are post-germinal centre B-cell neoplasms that show
varying degrees of plasmacytoid differentiation.
Etiology, pathology, and genetics Three distinct forms of marginal zone lymphomas are recognized: extra- nodal, nodal, and splenic. Extranodal marginal zone lymphomas of mucosa- associated lymphoid tissue (MALT lymphomas) are relatively common; they have been reported in nearly every anatomical site but are most frequent in the stomach, lung, and salivary gland. Antigen drive has
been implicated most clearly in gas- tric MALT lymphoma, associated with Helicobacter pylori infection. The most common translocations iden- tified in MALT lymphoma – t(11;18) (q21;q21), t(1;14)(p22;q32), and t(14;18)(q32;q21) – share a common pathway, which leads to the activa- tion of NF-κB and its downstream targets. Historically, it had been dif- ficult to distinguish lymphoplasma- cytic lymphoma from marginal zone lymphomas as both show evidence of plasmacytoid differentiation. Recent studies have identified mutations
Fig. 5.13.3. The main activation cascades of JAK-STAT and NF-κB signalling are in- volved in primary mediastinal B-cell lymphoma. Alternative pathway activation exists. Known gene alterations leading to constitutive pathway activity are shown in colour. The NF-κB signalling pathway is implicated in many forms of B-cell lymphoma.
490
in MYD88 involving L265P in 91% of patients with lymphoplasmacytic lymphoma, nearly all of whom had clinical Waldenström macroglobulin- aemia [11]. The mutation was shown to trigger IRAK-mediated NF-κB sig- nalling. Mutations in MYD88 were rarely encountered in marginal zone lymphomas, facilitating distinction of these groups.
Diffuse large B-cell lymphoma Diffuse large B-cell lymphomas (DLBCLs) are the most common histological group of lymphoma, representing up to 40% of cases worldwide.
Etiology, pathology, and genetics DLBCLs are heterogeneous mor- phologically, clinically, and at the genomic level. To address these is- sues, DLBCLs were among the first cases to be analysed by comple- mentary DNA array technology, and more recently also by genome-wide analysis [12]. A recent genome-wide association study identified 3q27 as a susceptibility locus for B-cell non-
Hodgkin lymphoma, and particularly for DLBCL, in the Chinese popula- tion [13]. By gene expression pro- filing, three groups were identified based on the differential expression of a large set of genes: the germi- nal centre-like group, the activated B-cell-like group, and primary me- diastinal (thymic) large B-cell lym- phoma. The third group shares many features with classic Hodgkin lym- phoma, showing activation of the NF-κB pathway (Fig. 5.13.3) [14]. More recently, frequent genetic defects have been identified in the B-cell antigen receptor (BCR) sig- nalling and NF-κB pathways in the activated B-cell-like type, providing new insight into the pathogenesis of DLBCL and new potential therapeu- tic targets [15,16]. In particular, in- hibitors of BTK have shown promise in a variety of B-cell lymphomas [17]. Recurrent mutations in the germinal centre-like type of DLBCL appear to target histone-modifying genes [12,18]. Somatic mutations in EZH2 also have been identified in follicular lymphoma, another tumour of germi- nal centre derivation.
Burkitt lymphoma Etiology, pathology, and genetics Burkitt lymphoma is the first subtype of lymphoma to be associated with a specific genetic aberration: trans- locations involving MYC and one of the immunoglobulin genes, most commonly IGH. The translocation is common to all subtypes of Burkitt lymphoma, regardless of the pres- ence of Epstein–Barr virus (EBV) or clinical features. Endemic Burkitt lymphoma is prevalent in equatorial Africa, corresponding in distribution to the malaria belt. It is universally positive for EBV, whereas sporadic and immunodeficiency-associated Burkitt lymphomas are EBV-positive in a smaller proportion of cases (15– 30%). Recent studies using genomic sequencing have identified recurrent somatic mutations that provide new insights into the pathogenesis of Burkitt lymphoma (reviewed in [19]). Three independent studies identified mutations in ID3 in a high proportion of cases, with mutations also seen in the transcription factor 3 (TCF3) gene. These studies link Burkitt lym- phoma to the cells of the dark zone of the germinal centre. ID3 mutations were absent in DLBCL. Mutations in this pathway were seen in sporadic and immunodeficiency-associated Burkitt lymphoma, and in a some- what lesser proportion of endemic cases of this disease.
Peripheral T-cell lymphoma Peripheral T-cell lymphomas are mor- phologically, immunophenotypically, and clinically heterogeneous and overall account for only about 10% of all non-Hodgkin lymphomas [20]. Classification schemes generally segregate nodal and extranodal lymphomas of this type; extranodal tumours are more often of cyto- toxic origin [21]. Anaplastic large cell lymphoma was the first to be associated with a specific genetic alteration: translocations involving ALK and a variety of gene partners, leading to overexpression of ALK. Recently, mutations involving IDH2 and TET2 have been identified in
Fig. 5.13.4. A boy with Burkitt lymphoma receives treatment at Banso Baptist Hospital in Cameroon.
Chapter 5.13 • Haematopoietic and lymphoid malignancies 491
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
3
angioimmunoblastic T-cell lympho- mas and other peripheral T-cell lymphomas of T-follicular helper cell origin. However, as yet the mo- lecular pathogenesis of most periph- eral T-cell lymphomas remains to be discovered.
Hodgkin lymphoma Hodgkin lymphomas are divided into classic Hodgkin lymphoma and nodular lymphocyte-predominant Hodgkin lymphoma. Both are de- rived from B cells, but the B-cell programme is markedly disrupted
in classic Hodgkin lymphoma, such that there is downregulation of most B-cell markers and func- tions. Epigenetic alterations may be responsible in part for loss of the B-cell programme [22]. Studies of the tumour cells implicate the NF-κB
Molecular detection of minimal residual disease in childhood leukaemia Rosemary Sutton
Fig. B5.13.1. Minimal residual disease (MRD) levels in childhood leukaemia patients who relapse compared with patients who are cured by their treatment.
Minimal residual disease (MRD) is a highly clinically relevant prog- nostic factor for patients with acute lymphoblastic leukaemia (ALL). Brüggemann et al. [1] cited 27 stud- ies that provided evidence that MRD is important in determining patient outcome in clinical trials for chil- dren or adults with newly diagnosed ALL, relapsed ALL, or bone mar- row transplants for ALL treatment. Consequently, many current clini- cal trials for childhood leukaemia patients include the stratification of patients into treatment risk groups based on initial MRD response to therapy measured in bone marrow aspirates using real-time quanti- tative polymerase chain reaction (PCR) or quantitative flow cytometry methods [1].
In essence, leukaemia patients who relapse usually have a slower response to early therapy than pa- tients who are cured by their treat- ment (Fig. B5.13.1). PCR MRD techniques can detect residual dis- ease at a sensitivity of about 1 in
100 000 cells (10–5), with quantifi- cation to 10–4 using definitions and standard procedures established by the EuroMRD group [2].
The molecular detection of sub- microscopic residual cancer cells is clinically useful in many malignan- cies and depends on having specific markers for the disease. The mark- ers most commonly used in ALL are the unique gene rearrangements present in cells of T- and B-cell lin- eage. The capacity of our immune systems to recognize millions of different antigens depends on the huge diversity of immunoglobulins and T-cell receptors, which results from the genetic rearrangement of numerous alternative gene seg- ments present in the genes coding for these antigen receptor proteins. Each leukaemic clone therefore has one or more unique gene rearrange- ments that can be used as molecular markers to distinguish the DNA of the leukaemic clone from the pa- tient’s normal DNA.
In addition to markers based on antigen receptor genes, several recurrent translocations and micro- deletions can be used as markers for smaller subsets of patients with childhood leukaemia to measure MRD in either genomic DNA or RNA samples from blood or mar- row. In patients with Philadelphia chromosome-positive leukaemia, the BCR-ABL transcripts are mea- sured by reverse-transcriptase PCR to detect disease recurrence at very low levels, allowing immediate inter- vention to treat molecular disease. Leukaemia in infants often has translocations involving the mixed lineage leukaemia (MLL) gene with unique breakpoints, which provide useful MRD markers, particularly in this patient group. MRD markers for ALL based on recurrent deletions in genes implicated in normal lymphoid development, such as IKZF1 and CRLF2, are also likely to be used in the future, especially since they oc- cur in higher-risk patients and their generic PCR MRD assays do not require DNA sequencing for each patient, making MRD diagnostics more accessible [3].
References 1. Brüggemann M et al.; European Working
Group for Adult Acute Lymphoblastic Leukemia (EWALL); International Berlin- Frankfurt-Münster Study Group (I-BFM- SG) (2010). Leukemia, 24:521–535. http://dx.doi.org/10.1038/leu.2009.268 PMID:20033054
2. van der Velden VH et al.; European Study Group on MRD Detection in ALL (ESG- MRD-ALL) (2007). Leukemia, 21:604–611. PMID:17287850
3. Venn NC et al. (2012). Leukemia, 26: 1414–1416. http://dx.doi.org/10.1038/leu. 2011.348 PMID:22157735
492
and JAK/STAT signalling pathways. A subset of classic Hodgkin lympho- ma, most often mixed cellularity sub- type, is positive for EBV, but the viral programme does not appear to pri- marily drive the neoplastic process.
Acute lymphoblastic leukaemia (ALL)/lympho- blastic lymphoma Acute lymphoblastic leukaemia (ALL)/lymphoblastic lymphoma are neoplasms of precursor B cells or T cells (lymphoblasts) with features corresponding to stages of B-cell development in the marrow and T-cell development in the thymus (Figs 5.13.1, 5.13.2). They usually involve bone marrow and blood in the case of ALL but can present in extramedullary sites, typically in lym- phoblastic lymphoma. Morphology and immunophenotyping suffice to establish the diagnosis, but recur- ring cytogenetic abnormalities oc- cur in 80% of ALL (Fig. 5.13.5), and many have proven important for risk stratification (see “Molecular detec- tion of minimal residual disease in childhood leukaemia”). However, not all cytogenetic abnormalities are
sufficient to establish a leukaemic clone, and cooperating molecular defects have long been suspected. Recently, whole-genome profiling in ALL has identified several gene mutations and submicroscopic de- letions that affect molecules crucial to B- and T-cell maturation, some of which correlate with clinical features. For example, 15% of paediatric B-cell ALL carries deletions or mutations of IKZF1 that affect transcriptional regulation of lymphoid maturation and confer a poor prognosis. Key genetic alterations in B-cell ALL, in addition to IKZF1, involve PAX5, JAK1, JAK2, CRLF2, and CREBBP. In T-cell ALL, activating mutations of NOTCH1 are found in about 50% of cases and correlate with a favourable outcome. In the future, whole-genome analyses may allow for better characterization of ALL and identification of abnormalities for tar- geted therapy [23–25].
Acute myeloid leukaemia Acute myeloid leukaemia (AML) is a heterogeneous disease charac- terized by increased blasts (usually ≥ 20% myeloblasts, monoblasts plus
promonocytes, and/or megakaryo- blasts) in blood or bone marrow. In recent years, there has been im- provement in survival, particularly in younger patients, but AML remains a fatal disease for most patients. The WHO classification of AML assigns patients to categories useful for pre- dicting outcome. Some are classified according to genetic abnormalities that determine morphological and clinical features (AML with recurrent genetic abnormalities), others have morphology and/or genetics that relate them to myelodysplastic syn- dromes (AML with myelodysplasia- related changes), some are unique because they follow prior cytotoxic therapy (therapy-related myeloid neoplasms), and the remainder are classified by lineage involvement and degree of differentiation.
Currently, karyotype and age are the most powerful predictors of prog- nosis [26,27]. Whole-genome and exome sequencing have revealed a plethora of gene mutations and submicroscopic genetic defects in this type of leukaemia. Although the prognostic significance of some al- terations, particularly mutated FLT3, NPM1, and CEBPA, has been recog- nized, the significance of many oth- ers remains to be determined [28]. A separate study involving 200 cases reported that AML genomes have fewer mutations than those of most other adult cancers, with an aver- age of only 13. A total of 23 genes were significantly mutated, and nearly all samples had at least one non-synonymous mutation in one of nine categories of genes relevant for pathogenesis, including transcrip- tion factor fusions (18% of cases), the gene encoding nucleophosmin (NPM1) (27%), tumour suppressor genes (16%), DNA methylation-re- lated genes (44%), signalling genes (59%), chromatin modifying genes (30%), myeloid transcription fac- tor genes (22%), cohesin-complex genes (13%), and spliceosome-com- plex genes (14%) (see “De novo DNA methyltransferases in normal and malignant haematopoiesis”) [29].
Fig. 5.13.5. Frequency of cytogenetic subtypes of paediatric acute lymphoblastic leu- kaemia (ALL). Pie chart includes all major B-cell ALLs (shown in green) and T-lineage subtypes of ALL (shown in red), to illustrate the relative frequency of each. The recently described BCR-ABL1-like subtype and BCR-ABL1-positive ALL are shown in yellow to illustrate the high frequency of childhood B-cell ALL cases with genetic alterations activating tyrosine kinase and cytokine receptor signalling, which may be amenable to targeted therapy. Data are derived from front-line studies of childhood ALL.
Chapter 5.13 • Haematopoietic and lymphoid malignancies 493
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
3
De novo DNA methyltransferases in normal and malignant haematopoiesis Margaret A. Goodell
DNA methylation plays a central role in regulating gene expression during development and is known to be dis- turbed in a variety of malignancies. The mechanisms through which aberrant DNA methylation contrib- utes to malignancy development, and through which hypomethylating agents exert their effects, are poorly understood. De novo DNA methyla- tion in murine haematopoietic stem cells has been studied as a model to address some of these fundamental questions.
Murine haematopoietic stem cells express high levels of both de novo DNA methyltransferases Dnmt3a and Dnmt3b. Using con- ditional knockout mice, the role of Dnmt3a in murine haematopoiesis has been investigated. The data indi- cate that in the absence of Dnmt3a, haematopoietic stem cell self-re- newal is dramatically enhanced at
the expense of differentiation. Serial stem cell transplantation augments this effect, such that phenotypically normal haematopoietic stem cells that fail to differentiate accumulate to high levels. Paradoxically, DNA methylation, examined genome- wide, is both increased and de- creased in Dnmt3a knockout hae- matopoietic stem cells, with CpG islands preferentially hypermethyl- ated, similar to the pattern of DNA methylation alterations in malignan- cies. The differentiated progeny of Dnmt3a knockout haematopoietic stem cells exhibited aberrant contin- ued expression of stem cell-specific genes that are normally repressed during differentiation. While muta- tions in DNMT3A are prevalent in human acute myeloid leukaemia, no frank leukaemia development was observed in the mice within the time frame initially examined.
Haematopoietic stem cell-spe- cific Dnmt3b knockout mice and Dnmt3a-Dnmt3b double-knockout haematopoietic stem cells have also been examined. In the absence of both de novo DNA methyltransfer- ases, the haematopoietic stem cells accumulate even more dramatically than in the Dnmt3a knockout, even though loss of Dnmt3b alone has minimal impact. Introduction of on- cogenes into the Dnmt3a knock- out stem cells decreases the time to malignant transformation com- pared with the oncogene or Dnmt3a knockout alone. These findings are pertinent to understanding the implications of mutations in DNA methyltransferases found in human haematological malignancies.
Reference Challen GA et al. (2012). Nat Genet, 44:23–31. http://dx.doi.org/10.1038/ng.1009 PMID:22138693
Chronic myeloid leukaemia, BCR-ABL1-positive Chronic myeloid leukaemia, BCR- ABL1 is a myeloproliferative neo- plasm arising in a pluripotent hae- matopoietic stem cell and is always associated with the BCR-ABL1 fusion gene resulting from the chromosomal translocation t(9;22) (q34;q11.2). BCR-ABL1 encodes an oncoprotein with constitutively ac- tivated tyrosine kinase activity that drives proliferation by interaction with the downstream pathways RAS, RAF, MYC, JUN kinase, and STAT. The natural history includes an ini- tial chronic phase characterized by granulocytic proliferation and granu- locytosis, followed by an accelerated phase and/or a myeloid or lymphoid blast phase. The development of tyrosine kinase inhibitors has dra- matically improved overall survival, from a 10-year overall survival of 15– 20% before these drugs were avail- able to 80–90% subsequent to their use. Point mutations in the tyrosine
kinase domain of BCR-ABL result in resistance to many tyrosine kinase inhibitors, and disease progression. Most patients refractory to front-line therapy usually respond to newer- generation tyrosine kinase inhibitors, but others may progress to the blast phase, which has a uniformly poor outcome [30].
BCR-ABL1-negative myelo- proliferative neoplasms These diseases, which include poly- cythaemia vera, essential thrombo- cythaemia, and primary myelofibro- sis, are clonal, stem-cell-derived neoplasms characterized by ex- cessive production of differentiated myeloid cells, which are sometimes difficult to distinguish from reactive marrow proliferations. In 2005, it was found that almost all patients with polycythaemia vera and nearly 50% with essential thrombocythae- mia or primary myelofibrosis have an activating somatic mutation, JAK2
V617F, encoding the cytoplasmic ty- rosine kinase JAK2. This discovery proved that these myeloproliferative neoplasms have abnormal cell sig- nalling pathways, providing an im- portant diagnostic tool – the JAK2 mutation – and raising expectations that tyrosine kinase inhibitors would emulate the successful therapy such agents mediated in chronic myeloid leukaemia. But additional data have shown that the JAK2 mutation is likely a secondary event and have also uncovered a complex layering of multiple genetic defects, particu- larly affecting genes such as TET2, EZH2, and IDH1/IDH2 involved in epigenetic regulation [31]. These additional abnormalities may ac- count for the only modest success to date using tyrosine kinase inhibi- tors. Further improvements to under- standing of the BCR-ABL1-negative myeloproliferative neoplasms will rely on their meticulous classifica- tion to enable correlation with the underlying genetic milieu.
494
References 1. Swerdlow SH, Campo E, Harris NL et al.,
eds (2008). WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues, 4th ed. Lyon: IARC.
2. Scarfò L, Fazi C, Ghia P (2013). MBL ver- sus CLL: how important is the distinction? Hematol Oncol Clin North Am, 27:251–265. http://dx.doi.org/10.1016/j.hoc.2013.01.004 PMID:23561472
3. Berndt SI, Skibola CF, Joseph V et al. (2013). Genome-wide association study identifies multiple risk loci for chronic lym- phocytic leukemia. Nat Genet, 45:868–876. http://dx.doi.org/10.1038/ng.2652 PMID: 23770605
4. Puente XS, Pinyol M, Quesada V et al. (2011). Whole-genome sequencing iden- tifies recurrent mutations in chronic lym- phocytic leukaemia. Nature, 475:101–105. ht tp://dx.doi.org/10.1038/nature10113 PMID:21642962
5. Wang L, Lawrence MS, Wan Y et al. (2011). SF3B1 and other novel cancer genes in chronic lymphocytic leukemia. N Engl J Med, 365:2497–2506. http://dx.doi.org/10.1056/ NEJMoa1109016 PMID:22150006
6. Tiacci E, Schiavoni G, Forconi F et al. (2012). Simple genetic diagnosis of hairy cell leukemia by sensitive detection of the BRAF-V600E mutation. Blood, 119:192– 195. http://dx.doi.org/10.1182/blood-2011- 08-371179 PMID:22028477
7. Egan JB, Shi CX, Tembe W et al. (2012). Whole-genome sequencing of multi- ple myeloma from diagnosis to plasma cell leukemia reveals genomic initiating events, evolution, and clonal tides. Blood, 120:1060–1066. http://dx.doi.org/10.1182/ blood-2012-01-405977 PMID:22529291
8. Liu Q, Salaverria I, Pittaluga S et al. (2012). Follicular lymphomas in children and young adults: a comparison of the pediatric vari- ant with usual follicular lymphoma. Am J Surg Pathol, 37:333–343. http://dx.doi. org/10.10 97/PAS.0b013e31826b9 b57 PMID:23108024
9. Jegalian AG, Eberle FC, Pack SD et al. (2011). Follicular lymphoma in situ: clinical implications and comparisons with partial involvement by follicular lymphoma. Blood, 118:2976–2984. http://dx.doi.org/10.1182/ blood-2011-05-355255 PMID:21768298
10. Jares P, Colomer D, Campo E (2012). Molecular pathogenesis of mantle cell lymphoma. J Clin Invest, 122:3416–3423. http://dx.doi.org/10.1172/JCI61272 PMID: 23023712
11. Treon SP, Xu L, Yang G et al. (2012). MYD88 L265P somatic mutation in Waldenström’s macroglobulinemia. N Engl J Med, 367: 826–833. http://dx.doi.org/10.1056/NEJMoa 1200710 PMID:22931316
12. Morin RD, Mendez-Lago M, Mungall AJ et al. (2011). Frequent mutation of histone- modifying genes in non-Hodgkin lympho- ma. Nature, 476:298–303. http://dx.doi. org/10.1038/nature10351 PMID:21796119
13. Tan DEK, Foo JN, Bei J-X et al. (2013). Genome-wide association study of B cell non-Hodgkin lymphoma identifies 3q27 as a susceptibility locus in the Chinese popula- tion. Nat Genet, 45:804–807. http://dx.doi. org/10.1038/ng.2666 PMID:23749188
14. Steidl C, Gascoyne RD (2011). The mo- lecular pathogenesis of primary medi- astinal large B-cell lymphoma. Blood, 118:2659–2669. http://dx.doi.org/10.1182/ blood-2011-05-326538 PMID:21700770
15. Ngo VN, Young RM, Schmitz R et al. (2011). Oncogenically active MYD88 mutations in human lymphoma. Nature, 470:115–119. http://dx.doi.org/10.1038/nature09 671 PMID:21179087
16. Pasqualucci L, Trifonov V, Fabbri G et al. (2011). Analysis of the coding genome of diffuse large B-cell lymphoma. Nat Genet, 43:830–837. http://dx.doi.org/10.1038/ng. 892 PMID:21804550
17. Kenkre VP, Kahl BS (2012). The future of B-cell lymphoma therapy: the B-cell re- ceptor and its downstream pathways. Curr Hematol Malig Rep, 7:216–220. http://dx. doi.org/10.1007/s11899-012-0127-0 PMID: 22688757
18. Morin RD, Johnson NA, Severson TM et al. (2010). Somatic mutations altering EZH2 (Tyr641) in follicular and diffuse large B-cell lymphomas of germinal-center origin. Nat Genet, 42:181–185. http://dx.doi.org/ 10.1038/ng.518 PMID:20081860
19. Campo E (2012). New pathogenic mecha- nisms in Burkitt lymphoma. Nat Genet, 44:1288–1289. http://dx.doi.org/10.1038/ ng.2476 PMID:23192177
20. Vose J, Armitage J, Weisenburger D; International T-Cell Lymphoma Project (2008). International peripheral T-cell and natural killer/T-cell lymphoma study: pathology findings and clinical outcomes. J Clin Oncol, 26:4124–4130. http://dx. doi.org/10.1200/JCO.2008.16.4558 PMID: 18626005
21. Jaffe ES, Nicolae A, Pittaluga S (2013). Peripheral T-cell and NK-cell lymphomas in the WHO classification: pearls and pitfalls. Mod Pathol, 26 Suppl 1:S71–S87. http:// dx.doi.org/10.1038/modpathol.2012.181 PMID:23281437
22. Küppers R (2012). New insights in the bi- ology of Hodgkin lymphoma. Hematology Am Soc Hematol Educ Program, 2012:328– 334. PMID:23233600
23. Mullighan CG (2012). Molecular genetics of B-precursor acute lymphoblastic leukemia. J Clin Invest, 122:3407–3415. http://dx.doi. org/10.1172/JCI61203 PMID:23023711
24. Iacobucci I, Papayannidis C, Lonetti A et al. (2012). Cytogenetic and molecular predictors of outcome in acute lympho- cytic leukemia: recent developments. Curr Hematol Malig Rep, 7:133–143. http://dx. doi.org/10.1007/s11899-012-0122-5 PMID: 22528731
25. Inaba H, Greaves M, Mullighan CG (2013). Acute lymphoblastic leukaemia. Lancet, 381:1943–1955. http://dx.doi.org/10.1016/ S0140-6736(12)62187-4 PMID:23523389
26. Burnett AK (2012). Treatment of acute myeloid leukemia: are we making prog- ress? Hematology Am Soc Hematol Educ Program, 2012:1–6. PMID:23233553
27. Grimwade D (2012). The changing paradigm of prognostic factors in acute myeloid leukae- mia. Best Pract Res Clin Haematol, 25:419– 425. http://dx.doi.org/10.1016/j.beha.2012. 10.004 PMID:23200538
28. Patel JP, Gönen M, Figueroa ME et al. (2012). Prognostic relevance of inte- grated genetic profiling in acute myeloid leukemia. N Engl J Med, 366:1079–1089. http://dx.doi.org/10.1056/NEJMoa1112304 PMID:22417203
29. Cancer Genome Atlas Research Network (2013). Genomic and epigenomic land- scapes of adult de novo acute myeloid leukemia. N Engl J Med, 368:2059–2074. http://dx.doi.org/10.1056/NEJMoa1301689 PMID:23634996
30. van Etten RA, Mauro M, Radich JP et al. (2013). Advances in the biology and therapy of chronic myeloid leukemia: proceedings from the 6th Post-ASH International Chronic Myeloid Leukemia and Myeloproliferative Neoplasms Workshop. Leuk Lymphoma, 54:1151–1158. http://dx.doi.org/10.3109/104 28194.2012.745524 PMID:23121619
31. Vakil E, Tefferi A (2011). BCR-ABL1– negative myeloproliferative neoplasms: a review of molecular biology, diagnosis, and treatment. Clin Lymphoma Myeloma Leuk, 11 Suppl 1:S37–S45. http://dx. doi.org/10.1016/j.clml.2011.04.002 PMID: 22035746
Chapter 5.14 • Skin cancer 495
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
4
Summary • Malignant melanoma is an ag-
gressive human cancer; over the past few decades, its incidence has increased dramatically worldwide among Caucasian populations.
• The risk of developing malig- nant melanoma varies mark- edly according to racial back- ground (skin pigmentation) and geographical location (sunlight- derived ultraviolet radiation). Ultraviolet-emitting tanning de- vices increase the risk of malig- nant melanoma.
• For patients with localized mel- anoma, prognosis is good with adequate surgical excision; metastatic melanoma is largely resistant to current therapies.
• Non-melanoma skin cancer includes squamous cell carci- noma and basal cell carcinoma, which are malignant epithelial neoplasms that often arise on sun-exposed areas of the skin in fair-skinned populations. These cancers cause local tis- sue destruction with functional and cosmetic deformity but are rarely a cause of death.
• Avoidance of excessive ultra- violet radiation from sunlight or
5.14 5 ORGAN SITE
Skin cancer Christine Guo Lian Martin C. Mihm Jr
Gérald E. Piérard (reviewer) Massimo Tommasino (reviewer)
tanning devices, particularly in young people, is the most effec- tive way to prevent melanoma and non-melanoma skin cancer.
Cutaneous melanoma Melanoma is the most aggressive form of skin cancer [1,2]. Cutaneous melanoma is a malignant prolifera- tion of melanocytes, the pigment- forming cells of skin.
Etiology Environmental factors play impor- tant roles in melanoma development and progression, including age, sex, degree of sun exposure, and ana- tomical location, as well as individ- ual susceptibility. Only 20–30% of melanomas arise in association with a melanocytic naevus, but the pres- ence of dysplastic naevi has been shown to be an independent risk factor for the development of mul- tiple primary melanomas [3]. Other risk factors include family history of melanoma and prior melanoma. In addition, rarely germline mutations have been found to confer risk; for example, mutations of the CDKN2A gene (on chromosome 9p21) have been found in 20–40% of families with dysplastic naevus syndromes in which at least three first-degree rela- tives are affected by melanoma [4].
In addition to genetic factors, ul- traviolet radiation from sun exposure also contributes to melanoma devel- opment. Most melanomas (80%) are caused by ultraviolet damage to sen- sitive skin, i.e. skin that burns easily, fair or reddish skin, skin with multiple freckles, or skin that does not tan and develops naevi in response to early sunlight exposure. Most damage caused by sunlight occurs in child- hood and adolescence. Ultraviolet radiation is particularly hazardous when it involves sporadic intense ex- posure, especially during childhood and early adolescence, for example through indoor tanning; children and adolescents are thus the most impor- tant target group for prevention pro- grammes [5].
Melanoma may occur anywhere on the skin, but in men most mela- nomas occur on the back, whereas in women most occur on the legs. In addition, patterns of age-specific incidence of melanoma at different anatomical sites in fair-skinned pop- ulations show that melanomas aris- ing on intermittently exposed body sites are significantly more common among younger and middle-aged adults, whereas melanomas of the head and neck are most common among older people. The difference in incidence by anatomical site is not completely explained by differential exposure to ultraviolet radiation.
496
Epidemiology
Melanoma • There were more than 232 000
new cases of melanoma and about 55 000 deaths estimated in 2012; the relatively low mor- tality rates are indicative of the reasonable prognosis.
• Incidence varies 100-fold be- tween countries worldwide. Rates are highest in countries with predominantly Caucasian populations, and more than 80% of the estimated new cases and close to 65% of the cancer deaths occurred in Oceania, Europe, and North America.
• Incidence rates in Australia and New Zealand are 2 times those observed in any other country. Incidence rates tend to be rather low in most African, Asian, and Latin American countries.
• Mortality tends to be higher in men than in women, related to sex differences in detection, body location of tumours, and subsequent prognosis.
• Incidence and mortality have been increasing over the past de- cades in many populations that are predominantly Caucasian. Of note is the (cohort-specific) sta- bilization and/or recent decline in incidence in certain high-risk populations (e.g. in Australia and among Whites in the USA). In low-risk populations (Asia, Africa, and to some extent Latin America), incidence rates have remained low and stable over time.
Map 5.14.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for melanoma in women, 2012.
Map 5.14.2. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for melanoma in women, 2012.
Moreover, melanoma can occur in areas of the body without sun expo- sure as well as in any ethnic group.
Pathology Melanomas occur primarily in the skin – more than 95% of cases – but are also found in the mucous membranes of the mouth, nose, anus, and vagina and, to a lesser extent, the intestine;
melanocytes are also present in the conjunctiva, the retina, and the menin- ges. Melanoma can be subtyped his- tologically into superficial spreading melanoma, nodular melanoma, acral lentiginous melanoma, and lentigo maligna melanoma. A staging sys- tem based on the histopathological parameters of the excised le- sion, including multiple prognostic
factors, is recommended [6]. The most robust independent prognos- tic factor in melanoma is tumour thickness (Breslow depth), found by measuring the vertical depth from the granular cell layer of the epidermis to the deepest detectable melanoma cell. In recent years, two additional criteria, ulceration and mitosis, have been shown to be important in
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.14 • Skin cancer 497
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
4
Chart 5.14.1. Estimated global number of new cases and deaths with proportions by major world regions, for melanoma in both sexes combined, 2012.
Chart 5.14.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for melanoma in women, circa 1975–2012.
Chart 5.14.3. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for melanoma in women, circa 1975–2012.
498
prognosis and have been includ- ed in the classification system (Table 5.14.1).
Most acquired naevi feature small round, oval, or spindled me- lanocytes. Small melanocytes with scant cytoplasm are evident in mel- anomas in severely sun-damaged skin. Large round or oval epithelioid melanocytes occur both in benign proliferations and in melanoma.
The outcome of melanoma initial- ly depends on the stage at presenta- tion. Excellent prognosis with 5-year survival rates of more than 90% can be achieved for patients who present
Table 5.14.1. Classification of melanoma and recommended surgical margins for primary melanoma
Classification Melanoma thickness Surgical excision margins
Tis In situ melanoma/no invasion of the dermis 5 mm
T1 ≤ 1 mm 10 mm
T1a Without ulceration and/or mitosis < 1/mm2
T1b With ulceration and/or mitosis > 1/mm2
T2a 1.01–2.0 mm 10 mm
T3 2.01– 4.0 mm 20 mm
T4 > 4.01 mm 20 mm
a Levels T2– 4 are classified a if ulceration is present, b if no ulceration is present.
Evaluating sunscreen as a measure to reduce the risk of melanoma Adèle C. Green
For several decades it was debated whether sunscreen use could pre- vent melanoma, even though it was biologically plausible that shielding the skin from excessive exposure to solar ultraviolet radiation in the long term would reduce the incidence of melanomas of the skin. Although many studies had examined this is- sue, human evidence was inconclu- sive because the studies were not randomized but observational. This means they were unable to distin- guish the main drivers of sunscreen use from those of melanoma cau- sation, since they largely overlap: susceptibility to sunburn, high oc- cupational or recreational sun expo- sure, and family history of melanoma [1]. Arguments against sunscreen as a melanoma preventive were also put forward, such as the example of dedicated sunbathers who misused sunscreen to prolong rather than minimize their sun exposure.
Evidence from a community- based trial has gone a long way towards resolving this debate. The Nambour Skin Cancer Prevention Trial was a randomized controlled trial involving 1621 adults aged 25– 75 years who had been randomly
selected from all adult residents of Nambour, a subtropical Australian township. In 1992 they were ran- domized either to apply a freely supplied broad-spectrum sunscreen with a sun protection factor (SPF) of 15+ to the head and arms daily until 1996, or to continue their usual level of sunscreen use or non-use. Trial participants were then followed up for a decade after cessation of the sunscreen intervention, with monitoring of all new melanomas, through pathology laboratories and the Queensland Cancer Registry, and of their ongoing sun behav- iour. Ten years after cessation of the sunscreen intervention, the 812 people randomized to daily sun- screen use had half as many new melanomas (11) as the control group of 809 people (22) [2], a difference of borderline statistical significance (P = 0.051). The decrease after regular sunscreen use was greater in invasive melanomas than in situ melanomas. During the trial, 75% of the intervention group applied sun- screen at least 3 or 4 days per week, compared with 25% of the control group; the sunscreen group also applied sunscreen more often to
the trunk and legs than did the con- trols. Otherwise, background sun exposure and protection patterns were the same in the intervention and control groups throughout the study period [2].
Use of a sunscreen with a higher SPF or greater ultraviolet A protection may have shown a stronger preventive effect, but in- creasing the participants’ average thickness of sunscreen application would more likely have had an even greater beneficial effect [3]. While this trial showed that regular use of sunscreen could halve people’s risk of developing melanoma, replication of the findings in another popula- tion would give further weight to the evidence.
References 1. Green AC, Williams GM (2007). Cancer
Epidemiol Biomarkers Prev, 16:1921– 1922. http://dx.doi.org/10.1158/1055-9965. EPI-07-0477 PMID:17932337
2. Green AC et al. (2011). J Clin Oncol, 29:257–263. http://dx.doi.org/10.1200/JCO. 2010.28.7078 PMID:21135266
3. Diffey BL (2001). Photochem Photobiol, 74:61–63. http://dx.doi.org/10.1562/0031- 8655(2001)074<0061:SAUPAM>2.0.CO;2 PMID:11460538
Chapter 5.14 • Skin cancer 499
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
4
with localized disease and primary tumours less than 1.0 mm in thick- ness. As recommended in the lat- est American Joint Committee on Cancer staging manual, melanoma patients are categorized into three groups: localized disease with no evidence of metastases (stage I– II), regional disease (stage III), and distant metastatic disease (stage IV). For localized disease (stage I and II), Breslow tumour thickness, ulceration, and mitotic rate are the three most important characteristics of the primary tumour for predicting outcome [6].
Genetics Heritable susceptibility is very im- portant in melanoma tumorigenesis. Germline mutations of two genes have been discovered in melanoma- prone families: CDKN2A (on chromo- some 9p21), which encodes p16INK4A, and CDK4 (on chromosome 12), which encodes cyclin-dependent ki- nase 4. A CDKN2A mutation causing functional inactivity of p16 has been found in up to 25% of melanoma- prone families worldwide, whereas a CDK4 mutation causing overex- pression of CDK4 protein, which is inhibited by p16, resulting in dysreg- ulation of the cell cycle, has been observed in only a few rare families [7]. In addition, loss-of-function
mutations in the human melanocor- tin 1 receptor (MC1R) gene have been associated with red hair, fair skin, freckles, and decreased ability to tan. However, susceptibility genes identified in melanoma-prone fami- lies are rarely mutated in sporadic melanomas. Numerous somatic mu- tations in melanoma vary between individuals and are different from those in other malignancies. Only 20% of melanomas harbour common TP53 mutations and inactivation of p16INK4A, associated with a poorer prognosis in melanoma.
Genes identified as playing a role in development of sporadic melanoma include CDKN2A, PTEN, and BRAF, and genes located on chromosomes 1p, 6q, 7p, 9p, and 11q. Mutations in BRAF (on chromo- some 7q) occur at a high frequency (45%) in melanomas [8], while so- matic mutations in BRAF have also been detected in 50% of patients with dysplastic naevi. The BRAF, NRAS, and KIT genes, involved in cell-cycle activating pathways, are commonly mutated in melanomas. The cur- rent data support a model in which melanoma tumorigenesis requires changes that initiate clonal expan- sion, overcome cell senescence, and reduce apoptosis. Thus, combination drug therapies with tyrosine kinase inhibitors of multiple components in the mitogen-activated protein kinase pathway, including MEK and ERK, are in clinical trials. Melanoma also is characterized by chromosomal gains, for example at 1q, 3p, 6p, 7, 8q, 11q, 12q14, 17q, and 20q, which encompass the melanoma onco- genes CDK4, cyclin D1 (CCND1), MYC, MDM2, and MITF, as well as chromosomal losses, for example at 1p, 6q, 9p, 10q, 11q, 17p, and 21q, which harbour the tumour suppres- sors p15INK4B, p16INK4A, p14ARF, and PTEN.
Whole-genome sequence data from 25 metastatic tumours identi- fied PREX2 – a PTEN-interacting protein and a negative regulator of PTEN in breast cancer – as a signifi- cantly mutated gene, with a mutation frequency of approximately 14% in an independent extension cohort of
107 human melanomas [9]. Other ge- netic changes include microsatellite instability, loss of heterozygosity, and increased activity of the telomerase enzyme.
Changes in gene expression con- tribute to progression from early be- nign melanocytic lesions to dysplas- tic naevi, to primary melanoma with a radial and then a vertical growth pattern, and to the acquisition of metastatic capacities. However, this sequence is challenged by the identi- fication of malignant melanoma stem cells [10]. Stem-like melanoma cells with expression of CD133, ABCB5, CD271, and CD166 are involved in the processes of progression and metastasis and are highly resistant to drugs and toxins [11].
Melanoma development not only requires genetic changes but is also driven by epigenetic alterations. 5-Hydroxymethylcytosine, a further indicator of modified DNA methyla- tion, is lost in melanoma genome- wide [12]. Rebuilding the 5-hydroxy- methylcytosine landscape in mela- noma cells by reintroducing active TET2 or IDH2 suppresses melanoma growth and increases tumour-free survival in animal models. Tumour suppressor small, non-coding RNAs, including miR-9, miR-34, miR-148a, and miR-375, are silenced by promot- er hypermethylation [13]. Expression of the histone H3K9 methyltrans- ferase SETDB1 accelerates mela- noma formation in zebrafish [14]. In addition, the H3K4 demethylase KDM5B (also called JARID1B) is con- sidered a melanoma stem cell marker [15]. The ability to understand and
Fig. 5.14.2. Malignant melanoma. Micro- graph showing a lesion just beneath the epidermis with pigmented and non- pigmented cells.
Fig. 5.14.1. Gross appearance of a cuta- neous pigmented lesion, which had been diagnosed as nodular melanoma. Note the dark reddish brown to black coloration of this lesion, as well as its raised nodular texture and amorphous, irregular borders, which should evoke suspicions about its classification.
500
control reversible epigenetic changes before irreversible mutations ensue may play an increasingly important role in further cancer prevention and treatment.
Prospects Prevention Prevention of melanoma is based on limiting exposure to ultraviolet radia- tion, particularly in the first 20 years of life. Sun avoidance behaviour and protection through increased shaded areas and appropriate clothing have been widely adopted as protective measures. Exposure can also be re- duced by avoiding artificial sources of ultraviolet radiation like sunbeds. Unequivocal evidence of the melano- ma-preventive effects of sunscreen use is lacking (see “Evaluating sun- screen as a measure to reduce the risk of melanoma”). Notwithstanding the inherent uncertainties and as- sumptions, the available evidence suggests that significant numbers of melanomas might be avoided by regular sunscreen use during recreational summer sun exposure [16]. The matter is of high priority to countries with a large fair-skinned community and strong lifestyle orien- tation to sun exposure, such as the USA [17] and Australia [18].
Targeted therapy Ipilimumab, a monoclonal antibody against CTLA-4, was the first immu- nomodulating antibody approved in the USA, in March 2011, for meta- static melanoma. Vemurafenib [19], a
specific inhibitor of signalling by mu- tated BRAF, was approved in August 2011 for patients with stage IV mela- noma with documented V600E or V600K mutation of the BRAF gene. These targeted agents have signifi- cantly changed the treatment regimen of metastatic melanoma. However, the use of these agents is challenged by their side-effects and unique limi- tations. For example, ipilimumab can achieve durable response but has a low overall response rate of less than 20%; vemurafenib has a higher response rate of 40–50% but with a median duration of only 5–6 months. Recent progress in the understand- ing of melanoma biology has led to the identification of novel pathways that could serve as targets for novel therapy (Fig. 5.14.4). For example, the PI3K pathway [20] and the p16– cyclin D–CDK4/6 –retinoblastoma pathway [21] could be exploited. In addition, some pilot clinical data sug- gest potential benefits with targeted therapeutic melanoma vaccines, for example infusion of lymphocytes genetically modified to express the MAGE-A3 tumour antigen in
patients with advanced melanoma. A consensus statement on tumour immunotherapy for the treatment of cutaneous melanoma has recently been adopted [22].
Non-melanoma skin cancer Basal cell carcinomas and squa- mous cell carcinomas are the two malignant epithelial neoplasms and the main forms of non-melanoma skin cancer, accounting for the large majority of all skin cancers.
Etiology Among several risk factors, sun exposure is the most recognized environmental cause, especially for squamous cell carcinoma. The association between sun exposure and basal cell carcinoma is more complex than that for squamous cell carcinoma [23]. Basal cell carcinoma seems to be associated more with intermittent exposure to high doses of solar radiation compared with sim- ilar doses delivered more continu- ously [24]. Fair-skinned populations who have substantial exposure to
Fig. 5.14.3. Sunscreen can help prevent sunburn and may contribute to a reduced risk of both melanoma and non-melanoma skin cancer.
Fig. 5.14.4. (A) Key signalling pathways in melanoma. Potentially druggable targets are indicated by purple shading. Mitogen-activated protein kinase (MAPK) signalling can be constitutively activated through alterations in membrane receptors or through mutations of RAS or BRAF. BRAF is frequently mutated in melanomas, almost in- variably by the V600E mutation, which results in an amino acid substitution at position 600, from a valine (V) to a glutamic acid (E), causing constitutive kinase activation. (B) Mechanisms of BRAF inhibitor resistance include reactivation of the MAPK pathway by CRAF, ARAF, or the p61BRAF(V600E) splice variant, mutational activation of NRAS or MEK, loss of the tumour suppressor NF1, and activation of compensatory pathways such as the PI3K network via enhanced receptor tyrosine kinase signalling (particularly through PDGFRβ, IGF1-R, FGFR, HGFR, and EGFR). Therapies approved or pending approval are shown in red. Activating mutations found are indicated by yellow shading, and inactivation by pink shading.
Chapter 5.14 • Skin cancer 501
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
4
ultraviolet radiation from sun ex- posure as well as indoor tanning, especially at a young age, have an increased risk of developing non- melanoma skin cancers [25,26]. In addition, most of these malignancies develop on sun-exposed skin sites, especially the head and neck area (80%). Actinic keratosis is a precur- sor with the potential to progress to invasive squamous cell carcinoma, and is considered to be a sun-in- duced precancerous lesion for these tumours. Besides sun exposure, therapeutic exposure to “psoralen plus ultraviolet A light” therapy for cutaneous disorders is associated with increased risk of non-melanoma skin cancers, particularly squamous cell carcinoma. Ionizing radiation used to treat childhood cancers in- creases the risk of subsequent de- velopment of non-melanoma skin cancers, including basal cell carci- nomas. Arsenic exposure, including environmental exposure, has been associated with increased risk of both tumour types.
Immunocompromised individu- als – typically organ transplant re- cipients, patients with a history of long-term steroid use, and people with HIV/AIDS – are at higher risk of non-melanoma skin cancer [27]. Non-melanoma skin cancers in organ transplant recipients are associated with human papillomavirus (HPV)
infection of skin keratinocytes. In ad- dition, detection of beta HPV species in cutaneous squamous cell carcino- ma suggests that certain HPV types may be involved in the progression of squamous cell carcinoma in im- munocompetent individuals as well. Furthermore, epidermodysplasia verruciformis-associated HPV infec- tion of skin keratinocytes seems to be associated with increased risk of squamous cell but not basal cell car- cinomas [28]. This association with HPV remains confined to squamous cell carcinomas arising on chronically sun-exposed areas of the skin.
Development of non-melanoma skin cancer is associated with a range of inherited disorders. Basal cell naevus syndrome (also called naevoid basal cell carcinoma syn- drome or Gorlin syndrome) is a rare autosomal dominant disorder due to germline mutations of the PTCH gene. Developmental abnormalities, postnatal tumours, and multiple ba- sal cell carcinomas are observed in affected patients. Epidermolysis bullosa syndrome is a heteroge- neous group of inherited blistering disorders due to minor trauma. The Dowling–Meara form of epidermol- ysis bullosa simplex is associated with an increased risk of basal cell carcinoma; squamous cell carcino- mas arise primarily in autosomal recessive dystrophic epidermoly- sis bullosa and are the most seri- ous complication for these patients. Naevus sebaceous is a rare congen- ital hamartoma of the skin composed of epidermal, follicular, sebaceous, and apocrine elements. While 15% of patients with naevus sebaceous develop some types of appendage tumour, 5% are diagnosed with basal cell carcinoma.
Pathology and genetics Squamous cell and basal cell car- cinomas are the major types of ke- ratinocytic tumours derived from the proliferation of epidermal and adnexal keratinocytes. The multiple variants of basal cell carcinoma have in common lobules, columns, bands, and cords of basaloid cells associated with scant cytoplasm, and a characteristic outer
palisade of cells associated with a surrounding loose fibromucinous stroma. Squamous cell carcinoma is a malignant neoplasm of epidermal keratinocytes in which the compo- nent cells show variable squamous differentiation.
Besides the risk factors such as sun exposure, genetic variations of populations also play a role in devel- opment of non-melanoma skin cancer. Basal cell and squamous cell carci- nomas arise predominantly in sun- sensitive people with fair skin and a history of sunburn [23]. Individuals with albinism, with lack of skin pigmen- tation, and patients with xeroderma pigmentosum, which causes an ex- treme sensitivity to sunlight, have a greatly increased risk of developing these tumours because they have little or no ability to repair skin damaged by ultraviolet radiation. In the past dec- ade, significant advances have been made in terms of understanding the molecular pathogenesis of non-mela- noma skin cancer. Ultraviolet-induced mutations of the tumour suppressor gene TP53 are frequently observed in these cancers [29]. The sonic hedge- hog signalling pathway plays a critical role in the pathogenesis of basal cell carcinoma. Mutations in the PTCH gene, which codes for a receptor of sonic hedgehog, are the underlying cause of naevoid basal cell carcinoma syndrome.
Non-melanoma skin cancers are usually indolent, but certain patho- logical features indicate high risk of recurrence and the potential for me- tastasis. In basal cell carcinoma, the histological subtypes of micronodular, infiltrative, and morpheaform basal cell carcinoma are more likely to recur than are the superficial and nodular subtypes. Basosquamous carcinoma, a “collision”-type tumour, is associated with a more aggressive nature, with an increased risk of metastasis [30].
Prospects The primary approach to preventing skin cancer is protection from sun ex- posure with various techniques. See the discussion above, at the end of the section on cutaneous melanoma.
Fig. 5.14.5. A young woman uses a sun- bed. Excess exposure to ultraviolet radia- tion is the most important modifiable risk factor for both malignant melanoma and non-melanoma skin cancer.
502
References
1. Siegel R, Naishadham D, Jemal A (2012). Cancer statistics, 2012. CA Cancer J Clin, 62:10–29. http://dx.doi.org/10.3322/caac. 20138 PMID:22237781
2. Tas F (2012). Metastatic behavior in mela- noma: timing, pattern, survival, and in- fluencing factors. J Oncol, 2012:647684. http://dx.doi.org/10.1155/2012/647684 PMID:22792102
3. Kang S, Barnhill RL, Mihm MC Jr et al. (1994). Melanoma risk in individuals with clinically atypical nevi. Arch Dermatol, 130: 999–1001. http://dx.doi.org/10.1001/arch derm.1994.01690080065008 PMID:8053717
4. Kefford RF, Newton Bishop JA, Bergman W, Tucker MA (1999). Counseling and DNA testing for individuals perceived to be genet- ically predisposed to melanoma: a consen- sus statement of the Melanoma Genetics Consortium. J Clin Oncol, 17:3245–3251. PMID:10506626
5. Tsao H, Sober AJ (1998). Ultraviolet radiation and malignant melanoma. Clin Dermatol, 16:67–73. http://dx.doi.org/10.1016/S0738- 081X(97)00191-0 PMID:9472435
6. Balch CM, Gershenwald JE, Soong SJ et al. (2009). Final version of 2009 AJCC melano- ma staging and classification. J Clin Oncol, 27:6199–6206. http://dx.doi.org/10.1200/ JCO.2009.23.4799 PMID:19917835
7. Goldstein AM, Chidambaram A, Halpern A et al. (2002). Rarity of CDK4 germline mutations in familial melanoma. Melanoma Res, 12:51– 55. http://dx.doi.org/10.1097/00008390-200 202000-00008 PMID:11828258
8. Davies H, Bignell GR, Cox C et al. (2002). Mutations of the BRAF gene in human cancer. Nature, 417:949–954. http://dx.doi. org/10.1038/nature00766 PMID:12068308
9. Berger MF, Hodis E, Heffernan TP et al. (2012). Melanoma genome sequencing re- veals frequent PREX2 mutations. Nature, 485:502–506. http://dx.doi.org/10.1038/na ture11071 PMID:22622578
10. Schatton T, Murphy GF, Frank NY et al. (2008). Identification of cells initiating hu- man melanomas. Nature, 451:345–349. http://dx.doi.org/10.1038/nature06489 PMID:18202660
11. Somasundaram R, Villanueva J, Herlyn M (2012). Intratumoral heterogeneity as a therapy resistance mechanism: role of melanoma subpopulations. Adv Pharmacol, 65:335–359. http://dx.doi.org/10.1016/B978- 0-12-397927-8.00011-7 PMID:22959031
12. Lian CG, Xu Y, Ceol C et al. (2012). Loss of 5-hydroxymethylcytosine is an epigenetic hallmark of melanoma. Cell, 150:1135–1146. http://dx.doi.org/10.1016/j. cell.2012.07.033 PMID:22980977
13. Mazar J, Khaitan D, DeBlasio D et al. (2011). Epigenetic regulation of microRNA genes and the role of miR-34b in cell invasion and motility in human melanoma. PLoS One, 6:e24922. http://dx.doi.org/10.1371/journal. pone.0024922 PMID:21949788
14. Ceol CJ, Houvras Y, Jane-Valbuena J et al. (2011). The histone methyltransferase SETDB1 is recurrently amplified in mela- noma and accelerates its onset. Nature, 471:513–517. http://dx.doi.org/10.1038/na ture09806 PMID:21430779
15. Roesch A, Fukunaga-Kalabis M, Schmidt EC et al. (2010). A temporarily distinct sub- population of slow-cycling melanoma cells is required for continuous tumor growth. Cell, 141:583–594. http://dx.doi.org/10.1016/j. cell.2010.04.020 PMID:20478252
16. Diffey BL (2009). Sunscreens as a pre- ventative measure in melanoma: an evi- dence-based approach or the precaution- ary principle? Br J Dermatol, 161 Suppl 3:25–27. http://dx.doi.org/10.1111/j.1365- 2133.2009.09445.x PMID:19775353
17. Lazovich D, Choi K, Vogel RI (2012). Time to get serious about skin cancer prevention. Cancer Epidemiol Biomarkers Prev, 21:1893–1901. http://dx.doi.org/ 10.1158/1055-9965.EPI-12-0327 PMID: 22962407
18. McCarthy WH (2004). The Australian ex- perience in sun protection and screening for melanoma. J Surg Oncol, 86:236– 245. http://dx.doi.org/10.1002/jso.20086 PMID:15221930
19. Bollag G, Hirth P, Tsai J et al. (2010). Clinical efficacy of a RAF inhibitor needs broad target blockade in BRAF-mutant mel- anoma. Nature, 467:596–599. http://dx.doi. org/10.1038/nature09454 PMID:20823850
20. Kwong LN, Davies MA (2013). Navigating the therapeutic complexity of PI3K pathway inhibition in melanoma. Clin Cancer Res, 19:5310–5319. http://dx.doi. org/10.1158/1078-0432.CCR-13-0142 PMID: 24089444
21. Sheppard KE, McArthur GA (2013). The cell-cycle regulator CDK4: an emerg- ing therapeutic target in melanoma. Clin Cancer Res, 19:5320–5328. http://dx.doi. org/10.1158/1078 - 0 432.CCR-13 - 0259 PMID:24089445
22. Kaufman HL, Kirkwood JM, Hodi FS et al. (2013). The Society for Immunotherapy of Cancer consensus statement on tumour immunotherapy for the treatment of cu- taneous melanoma. Nat Rev Clin Oncol, 10:588–598. http://dx.doi.org/10.1038/nr clinonc.2013.153 PMID:23982524
23. Armstrong BK, Kricker A, English DR (1997). Sun exposure and skin cancer. Australas J Dermatol, 38 Suppl 1:S1–S6. PMID:10994463
24. Kricker A, Armstrong BK, English DR, Heenan PJ (1995). Does intermittent sun exposure cause basal cell carcinoma? A case-control study in Western Australia. Int J Cancer, 60:489–494. http://dx.doi. org/10.1002/ijc.2910600411 PMID:7829262
25. Perkins JL, Liu Y, Mitby PA et al. (2005). Nonmelanoma skin cancer in survivors of childhood and adolescent cancer: a re- port from the childhood cancer survivor study. J Clin Oncol, 23:3733–3741. http:// dx.doi.org/10.120 0/JCO. 20 05.0 6. 237 PMID:15923570
26. Wehner MR, Shive ML, Chren M-M et al. (2012). Indoor tanning and non-melanoma skin cancer: systematic review and meta- analysis. BMJ, 345:e5909. http://dx.doi. org/10.1136/bmj.e5909 PMID:23033409
27. Grulich AE, van Leeuwen MT, Falster MO, Vajdic CM (2007). Incidence of cancers in people with HIV/AIDS compared with immu- nosuppressed transplant recipients: a meta- analysis. Lancet, 370:59–67. http://dx.doi. org/10.1016/S0140 - 673 6(07) 61050 -2 PMID:17617273
28. Pfister H (2003). Chapter 8: Human papil- lomavirus and skin cancer. J Natl Cancer Inst Monogr, 2003:52–56. http://dx.doi.org/ 10.1093/oxfordjournals.jncimonographs. a003483 PMID:12807946
29. Benjamin CL, Melnikova VO, Ananthaswamy HN (2008). p53 protein and pathogenesis of melanoma and nonmelanoma skin cancer. Adv Exp Med Biol, 624:265–282. http:// dx.doi.org/10.1007/978-0-387-77574-6_21 PMID:18348463
30. Bath-Hextall F, Bong J, Perkins W, Williams H (2004). Interventions for basal cell car- cinoma of the skin: systematic review. BMJ, 329:705. http://dx.doi.org/10.1136/ bmj.38219.515266.AE PMID:15364703
Chapter 5.15 • Thyroid cancer 503
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
5
Summary • The incidence of thyroid cancer is
increasing worldwide. This trend is attributed predominantly to the improved detection and screen- ing of nodular thyroid disease, particularly involving detection of papillary microcarcinoma.
• Exposure to ionizing radiation, especially during childhood, is a cause of thyroid cancer.
• Multiple endocrine neoplasia type 2 (MEN2)-associated med- ullary thyroid carcinoma serves as a role model for personal- ized medicine as RET proto- oncogene mutation analysis enables predictive testing and genotype-based, tailored pro- phylactic treatment.
• Molecular profiling using BRAF mutation analysis can allow for risk assessment of papillary thy- roid carcinoma.
• Targeted therapy will provide nov- el options to treat advanced and de-differentiated thyroid cancer.
Thyroid cancer includes a variety of malignancies ranging from the common differentiated thyroid can- cers to the very rare but uniformly
5.15 5 ORGAN SITE
Thyroid cancer Frank Weber Dillwyn Williams (reviewer)
lethal anaplastic thyroid cancer [1,2]. Thyroid cancers develop either from the follicular epithelial cells – giving rise to papillary thyroid cancer, fol- licular thyroid cancer, and likely un- differentiated and anaplastic thyroid cancer – or from the parafollicular, calcitonin-producing C cells, which give rise to medullary thyroid cancer (Fig. 5.15.1).
Etiology Worldwide, thyroid cancer is among the less frequent malignancies. However, in the past decades the incidence has almost doubled – at- tributed for the most part to the in- creased diagnosis of papillary thy- roid cancer [3]. While the reasons for the increase in thyroid cancer incidence require further elucida- tion, several lines of evidence indi- cate that in most circumstances the increase is attributable to optimal diagnostic scrutiny aided by techni- cal advances [3,4]. Overall, the eti- ology of thyroid cancer is not fully understood.
A clear gender disparity is evi- dent for thyroid cancer; it occurs more frequently in women than in men. In the USA, this tumour type accounts for 1.7% of all malignan- cies, corresponding to 2.6% of cancers in women and 0.85% of cancers in men, whereas in Japan the female-to-male ratio may be 13
[3]. Reasons for the overall 3-fold higher incidence of thyroid cancers in women than in men remain un- clear. A role for female hormonal and reproductive factors as drivers for thyroid cancer development has been shown in vitro but has not been unequivocally established in popula- tion-based analysis [5–7]. Thyroid cancer development is strongly as- sociated with a history of benign nodules/adenoma or goitre [7]. Solitary, palpable thyroid nodules can be identified in about 7% of the adult population, and the prevalence of nodules increases with age; the lifetime risk of developing a thyroid
Fig. 5.15.1. Surgical specimen showing thyroidectomy with en bloc central lymph- adenectomy for papillary thyroid cancer.
504
Epidemiology
Thyroid cancer • With 77% of the cases occur-
ring in women in 2012, thyroid cancer is one of the few non-sex- specific cancers that are more frequent in women. There were 230 000 new cases in women, representing the eighth most frequent cancer among women, compared with 68 000 cases in men, representing the 18th most frequent cancer among men.
• The case fatality of thyroid can- cer is very low, most notably in countries with very high levels of human development (mortality- to-incidence ratio, 0.14).
• There were an estimated 40 000 deaths from thyroid cancer in 2012 (27 000 of them in wom- en). Mortality rates are highest in Melanesia, in parts of Africa, and generally in countries with lower levels of human development.
• Although 37% of the estimated new cases occurred in Europe and North America and 48% in Asia, only 21% of the thyroid can- cer deaths occurred in Europe and North America, compared with 57% in Asia.
• Elevated incidence rates are found in some European coun- tries and in North America. However, the highest incidence rates globally are in the Republic of Korea, where thyroid cancer is the most frequent cancer among women. Incidence rates tend to be low in many parts of Asia and in Africa.
Map 5.15.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for thyroid cancer in women, 2012.
Map 5.15.2. Global distribution of estimated age-standardized (World) mortality rates (ASR) per 100 000, for thyroid cancer in women, 2012.
nodule is estimated to be 10%. The relative risk of cancer development in patients with thyroid nodules is 3–6 compared with unaffected controls.
Population-based studies sug- gest that risk of thyroid cancer is associated with several conditions. For instance, height and body mass index are moderately associated with thyroid cancer risk in both men
and women [7]. Acromegaly, a rare condition caused by an overpro- duction of growth hormone, is also associated with increased risk of thyroid cancer. With regard to di- etary risk factors for thyroid cancer development, iodine deficiency may induce benign thyroid conditions that predispose for thyroid cancer. Furthermore, iodine deficiency is associated with the prevalence of
follicular thyroid cancer [3] because in iodine-deficient countries follicular thyroid cancer accounts for up to 40% of all thyroid cancer, whereas in regions with high iodine supple- mentation the same tumour type ac- counts for less than 20% of cases. However, it is uncertain to what ex- tent dietary iodine intake might serve as a risk predictor for thyroid cancer development [7]. Published results
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.15 • Thyroid cancer 505
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
5
Chart 5.15.1. Estimated global number of new cases and deaths with proportions by major world regions, for thyroid cancer in both sexes combined, 2012.
Chart 5.15.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for thyroid cancer in women, circa 1975–2012.
Chart 5.15.3. Age-standardized (World) mortality rates per 100 000 by year in selected populations, for thyroid cancer in women, circa 1975–2012.
506
on the contribution of nutritional fac- tors to thyroid cancer development are generally inconsistent and not conclusive.
Among the well-established high-risk factors for thyroid cancer development is exposure to ioniz- ing radiation, especially when this occurs during childhood. Radiation exposure accounts for cancer de- velopment in about 5% of patients. Sources of radiation include certain medical treatments as well as radia- tion fallout from nuclear power plant accidents or nuclear weapons. The Chernobyl nuclear power plant acci- dent took place in 1986, and during the following years an unprecedent- ed rise in papillary thyroid cancer development during childhood was observed. Relevant cases indicat- ed that ionizing radiation leads to chromosomal rearrangements such as RET/PTC fusion transcripts that give rise to papillary thyroid cancer development [8,9].
Radiation exposure in the con- text of medical care has included treatment for acne or enlarged ton- sils during the 1960s. While this in- tervention is no longer used, previ- ously exposed individuals are now in their fifties, the age group for which papillary thyroid cancer occurs most commonly. Other reasons for radia- tion therapy in childhood are tumour therapy for cancers such as lympho- ma, Wilms tumour, or neuroblasto- ma. Exposure to diagnostic radiation (medical diagnostic radiography) has increased over the past years and currently is the leading source of radiation exposure in the USA. It is not fully established to what extent the repeated exposure to diagnostic radiation leads to an increased risk of thyroid cancer. However, a re- cent study showed a 13% increase in thyroid cancer risk for every 10 reported dental radiographs; caution should be exercised when requiring repeated radiography – especially during childhood [10].
Several inherited conditions, such as familial adenomatous pol- yposis syndrome and Cowden syn- drome, are associated with differ- ent types of thyroid cancer [11]. For
non-syndromic familial non-medul- lary thyroid cancer, epidemiological data indicate a very high likelihood of familial aggregation and hence a strong genetic component. Indeed, about 10% of all non-medullary thy- roid cancers are hereditary, and consequently first-degree relatives of patients with thyroid cancer have an up to 10-fold increased risk com- pared with the general population. The standardized incidence ratio, an index for estimating familial risk of developing a malignancy, exceeds a notable value of 3.8 for thyroid cancer. Thus, differentiated epithe- lial thyroid carcinomas have one of the highest familial risks of all cancer sites [11].
Pathology Thyroid cancer primarily involves the following entities, in decreasing order of frequency: differentiated thyroid cancers, which comprise papillary and follicular thyroid can- cer; medullary thyroid cancer; and poorly differentiated thyroid cancer, which includes undifferentiated and anaplastic thyroid cancer [12].
These tumour types are of en- dodermal origin and arise from the thyroid follicles, except for medullary thyroid cancer, which develops from the parafollicular calcitonin-secret- ing C cells and is of neuroendocrine origin. In addition, the thyroid stroma contains lymphoid cells as well as connective tissue that might give rise to thyroid lymphomas, which are nearly always non-Hodgkin lympho- mas or sarcoma; both, however, are rather rare and account for less than 5% of all thyroid cancers [12].
Differentiated thyroid carcinoma Papillary thyroid carcinomas are ma- lignant epithelial tumours that show evidence of follicular cell differentia- tion and display characteristic nuclear features. These are large nuclei that have hypodense chromatin (ground- glass appearance) and show intranu- clear inclusions and nuclear grooves. The cellular features of papillary thy- roid cancer are amphophilic, finely granular cytoplasm, large pale nuclei, nuclear grooves, and psammoma bodies. On macroscopic inspection, papillary thyroid cancer presents as
Fig. 5.15.2. More than two years after the accident at Fukushima Daiichi nuclear power plant, work continues to ensure that the damaged units remain stable. Here, workers in protective clothing and masks are shown outside the Emergency Response Centre, the main control hub at the site. Radiation exposure during childhood is the most clearly defined environmental factor associated with benign and malignant thyroid tumours.
Chapter 5.15 • Thyroid cancer 507
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
5
an invasive neoplasm of whitish col- our lacking a tumour capsule [12]. A multifocal manifestation is common. Approximately 50% of these newly detected cancers are papillary micro- carcinomas, which measure 1 cm or less in largest diameter [4]. In autopsy studies, these small tumours have been detected in as many as 35% of cases, and meticulous examination of thyroidectomy specimens has re- vealed micropapillary thyroid cancer in as many as 24% of cases.
There is growing evidence that a subset of these overall apparently indolent micropapillary thyroid can- cers are highly aggressive tumours that harbour a tendency to recur, me- tastasize, and cause death. Thus, it is of great importance to identify prognostic markers. BRAF muta- tion analysis might identify patients at risk. Similarly, histopathological characteristics, such as are indicated by the sclerosing variant, appear to be associated with metastases from these small papillary thyroid cancers (hazard ratio, 11.8) (Fig. 5.15.5) [13].
Follicular thyroid cancers are well-encapsulated lesions that display fibrosis, haemorrhage, and cystic areas. Three types of
follicular thyroid cancer may be dis- tinguished with reference to their clinical relevance: (i) minimally in- vasive tumours without angioinva- sion, which have the best progno- sis (disease-free survival, 97%); (ii) minimally invasive follicular thyroid cancers with angioinvasion (dis- ease-free survival, 81%); (iii) widely invasive follicular thyroid cancers, for which distant metastases are found in about 50% of patients [14]. The WHO classification of thyroid cancer defines follicular thyroid carcinoma as a lesion showing evi- dence of follicular cell differentiation and lacking the diagnostic nuclear features of papillary carcinoma. Thus, the diagnosis of follicular thy- roid cancer is based on the exclu- sion of the typical nuclear features of papillary thyroid cancer [12]. The major challenge for pathology rep- resents those thyroid tumours with a follicular growth pattern that include an array of neoplastic and malignant lesions. In addition, the differentia- tion of follicular thyroid malignancy from follicular adenoma is based on the identification of vascular and/or capsular invasion [15].
Medullary thyroid carcinoma On macroscopic examination, medullary thyroid carcinomas are non-encapsulated but well-circum- scribed lesions that contain areas of calcification. In a hereditary set- ting, multifocality is rather common. The unique feature of this cancer is a characteristic deposit of amyloid. Furthermore, staining for calcito- nin and carcinoembryonic antigen (CEA) helps to differentiate this cancer from other forms [12].
Poorly differentiated thyroid carcinoma Poorly differentiated thyroid can- cer is a highly aggressive cancer. Based on morphological and clini- cal criteria, these tumours occupy an intermediate position between differentiated forms (follicular and papillary carcinomas) and ana- plastic carcinomas. Most of these cancers develop by de-differenti- ation from their well-differentiated forms, as described above. Thus, the biology and histology of poorly differentiated thyroid carcinoma show characteristics of follicular or papillary thyroid cancer. Based on the WHO classification of thyroid cancer, poorly differentiated thy- roid carcinomas are neoplasms of follicular origin but lack the typical structural presentation of follicular cell differentiation.
Anaplastic thyroid cancers are highly lethal carcinomas derived from the thyroid follicular cells that lack thyroglobulin expression. Most
Fig. 5.15.3. Molecular pathways dysregulated, and putative molecular targets for ther- apy, in differentiated thyroid carcinoma. The highly complex interaction of protein/lipid phosphatases and protein/lipid kinases indicates that molecular targeted therapy aimed at only one component might not be sufficient to stabilize the dysregulated pathway. Furthermore, molecular targeted therapy will have to be adapted to the genomic make- up of the individual as it interacts with the genome of the tumour. pAKT, phosphory- lated protein kinase B; PI3K, phosphatidylinositol 3-kinase; PIP3, phosphatidylinositol 3,4,5-triphosphate; pMAPK, phosphorylated mitogen-activated protein kinase; PPAR-γ, peroxisome proliferator-activated receptor gamma.
Fig. 5.15.4. Cervical computed tomogra- phy image of anaplastic thyroid cancer.
508
of these cancers already show local invasion and metastases at the time of diagnosis (Fig. 5.15.4).
Genetics Thyroid cancer oncogenesis is as- sociated with multiple genetic and epigenetic alterations. The activation of the MAPK and PI3K/AKT signal- ling pathways appears to be crucial for thyroid cancer initiation and pro- gression (Fig. 5.15.5) [16].
Point mutations of the BRAF and RAS genes are commonly found in thyroid cancers, as are alterations in- volving RET/PTC and PAX8/PPAR-γ chromosomal rearrangements [17,18]. Several lines of evidence indicate that certain etiological factors are asso- ciated with specific molecular alter- ations. For instance, exposure to ion- izing radiation leads to chromosomal rearrangements, while exogenous chemicals tend to mediate genomic damage by causing point mutations. The extent to which high iodine intake is associated with the development of BRAF point mutations requires clarification as current data are not consistent.
The elucidation of the molecular pathology of thyroid carcinoma has led to the identification of diagnos- tic and prognostic markers. Among these, a BRAF point mutation may be a valuable indicator for the man- agement of patients with thyroid nodules. Growing evidence shows that BRAF V600E-positive micro- papillary thyroid cancer is more ag- gressive, and surgical management should be adopted accordingly [19].
Genetic alterations found in un- differentiated and anaplastic thyroid cancer involve the tumour suppres- sor gene TP53 and cell-cycle check- point genes such as p27 or p21. In addition, the Wnt pathway genes and the PTEN/PI3K/AKT pathway appear to play an important role in advanced thyroid cancer.
Heritable disease About 25% of all medullary thyroid cancers occur as part of the mul- tiple endocrine neoplasia type 2 (MEN2) syndrome [20]. MEN2 is an autosomal dominant transmit- ted tumour syndrome caused by germline mutation in the RET pro- to-oncogene and comprises the key
endocrine neoplasia components of phaeochromocytoma, hyperparathy- roidism, and, importantly, medullary thyroid carcinoma – the life-limiting scenario. MEN2-associated medul- lary thyroid cancer can be consid- ered the role model for the practice of genomic medicine. The presence of mutations in the RET medullary thyroid cancer susceptibility gene has enabled the implementation of a powerful molecular diagnostic test to identify mutation carriers at a pre- morbid stage (i.e. predictive testing) and the delineation of a genotype– phenotype correlation that allows tailored surgical management based on the distinct amino acid altered (Fig. 5.15.5).
Heritable non-medullary thyroid cancer occurs not as a single entity but as part of different tumour syn- dromes such as Cowden syndrome, Carney complex, Gardner syndrome (familial adenomatous polyposis syndrome), and Werner syndrome, all of which are transmitted in an autosomal dominant fashion [11]. In these tumour syndromes, differen- tiated thyroid cancer occurs in up to 20% of patients. A small portion of non-medullary thyroid cancer, termed familial non-medullary thy- roid cancer (mostly papillary thyroid cancer), appears to be inherited without other associated patholo- gies. Non-syndromic familial non- medullary thyroid cancer is charac- terized by two or more first-degree relatives who are affected by thyroid carcinomas but lack signs of other hereditary syndromes or exposure to other risk factors (i.e. radiation). Despite extensive research over the past 10 years to identify the gene or genes associated with non-syn- dromic familial non-medullary thy- roid cancer, no susceptibility genes have been identified. Nonetheless, at least five putative susceptibility loci have been identified, at 1q21, 2q21, 8p23.1-p22, 14q31, and 19p13.2. Of these, only 14q31 and 19p13.2, which correspond to the location of NMTC1 and TCO, respectively, have been replicated in independent fam- ily sets, but the susceptibility genes
Fig. 5.15.5. Genotype–phenotype correlation in multiple endocrine neoplasia type 2 (MEN2). Germline RET mutations are specific for subgroups in MEN2. Codons of the RET oncogene affected by germline mutations are depicted in correlation to their physi- cal location within the RET tyrosine kinase (TK) receptor. The horizontal bars indicate the frequency of these mutations in the three clinical subtypes MEN2A, MEN2B, and familial medullary thyroid carcinoma (FMTC).
Chapter 5.15 • Thyroid cancer 509
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
5
mapped to these loci have not yet been identified [21–23].
Prospects Ionizing radiation causes thyroid cancer and, as mentioned above, ex- posure to such radiation – especially in childhood – should be avoided
whenever possible. Since there are no other established causes, one challenge is to identify those patients with benign nodular thyroid disease who have an increased risk of thy- roid cancer development. Screening for thyroid cancer in these patients would be optimized by improvement in diagnostic modalities such as ul- trasound elastography of the thyroid and molecular classification using fine-needle aspirates (Fig. 5.15.6).
Most types of thyroid cancer are associated with excellent long-term survival: the 10-year overall survival exceeds 90% for papillary thyroid cancer and declines to 85% for fol- licular thyroid cancer and 70% for medullary thyroid cancer; for ana- plastic tumours, the long-term out- come is very poor [1]. Metastasis commonly affects the regional lymph nodes as well as involving spread to the bone, lung, and liver. Tumour re- currence within the thyroid bed and distant recurrence are frequent and
occur in about 30% of patients within 10 years after initial diagnosis [2].
Because of the availability of RET gene mutation analysis, famil- ial cases of medullary thyroid carci- noma as part of the MEN2 tumour syndrome can be prevented by early, prophylactic resection of the thyroid gland. Elucidation of dysregulated pathways involved in thyroid onco- genesis has enabled the identifica- tion of potential new biomarkers and therapeutic targets. Small-molecule therapies targeting receptor tyrosine kinases, as well as their intracellu- lar downstream proteins such as mTOR, BRAF, AKT, or MEK, are being investigated in clinical trials and appear to be promising for the treatment of advanced thyroid can- cer [19]. These investigations have opened up a new aspect of cancer therapy and tumour re-differentiation in the setting of advanced tumours that are not amenable to treatment by surgery or radio-iodine therapy.
Fig. 5.15.6. Ultrasound elastography im- age of a suspicious thyroid nodule, indicat- ing a hard nodule (Rago score, 3 or 4).
510
References 1. Sherman SI (2003). Thyroid carcinoma.
Lancet, 361:501–511. http://dx.doi.org/ 10.1016/S0140-6736(03)12488-9 PMID: 12583960
2. Mazzaferri EL, Kloos RT (2001). Current approaches to primary therapy for papil- lary and follicular thyroid cancer. J Clin Endocrinol Metab, 86:1447–1463. http://dx. doi.org/10.1210/jc.86.4.1447 PMID:11297567
3. Wartofsky L (2010). Increasing world incidence of thyroid cancer: increased detection or higher radiation expo- sure? Hormones (Athens), 9:103–108. PMID:20687393
4. Mazzaferri EL (2012). Managing thyroid microcarcinomas. Yonsei Med J, 53:1–14. http://dx.doi.org/10.3349/ymj.2012.53.1.1 PMID:22187228
5. Peterson E, De P, Nuttall R (2012). BMI, diet and female reproductive factors as risks for thyroid cancer: a system- atic review. PLoS One, 7:e29177. http:// dx.doi.org/10.1371/journal.pone.0029177 PMID:22276106
6. Rahbari R, Zhang L, Kebebew E (2010). Thyroid cancer gender disparity. Future Oncol, 6:1771–1779. http://dx.doi.org/10. 2217/fon.10.127 PMID:21142662
7. Dal Maso L, Bosetti C, La Vecchia C, Franceschi S (2009). Risk factors for thyroid cancer: an epidemiological re- view focused on nutritional factors. Cancer Causes Control, 20:75–86. http:// dx.doi.org/10.1007/s10552- 008-9219-5 PMID:18766448
8. Nikiforov YE (2006). Radiation-induced thyroid cancer: what we have learned from Chernobyl. Endocr Pathol, 17:307–317. http://dx.doi.org/10.1007/s12022- 00 6 - 0001-5 PMID:17525478
9. Gandhi M, Evdokimova V, Nikiforov YE (2010). Mechanisms of chromosomal rearrangements in solid tumors: the model of papillary thyroid carcinoma. Mol Cell Endocrinol, 321:36– 43. http:// dx.doi.org/10.1016/j.mce.2009.09.013 PMID:19766698
10. Neta G, Rajaraman P, Berrington de Gonzalez A et al. (2013). A prospective study of medical diagnostic radiography and risk of thyroid cancer. Am J Epidemiol, 177:800–809. http://dx.doi.org/10.1093/aje/ kws315 PMID:23529772
11. Weber F, Eng C (2008). Update on the molecular diagnosis of endocrine tumors: toward –omics-based personalized health- care? J Clin Endocrinol Metab, 93:1097– 1104. http://dx.doi.org/10.1210/jc.2008- 0212 PMID:18390809
12. DeLellis RA, Lloyd RV, Heitz PU, Eng C, eds. (2004). Pathology and Genetics of Tumours of Endocrine Organs. Lyon: IARC.
13. Pellegriti G, Scollo C, Lumera G et al. (2004). Clinical behavior and outcome of papillary thyroid cancers smaller than 1.5 cm in diameter: study of 299 cases. J Clin Endocrinol Metab, 89:3713–3720. http://dx.doi.org/10.1210/jc.2003-031982 PMID:15292295
14. O’Neill CJ, Vaughan L, Learoyd DL et al. (2011). Management of follicular thyroid carcinoma should be individualised based on degree of capsular and vascular inva- sion. Eur J Surg Oncol, 37:181–185. http:// dx.doi.org/10.1016/j.ejso. 2010.11.0 05 PMID:21144693
15. Baloch ZW, Livolsi VA (2002). Follicular- patterned lesions of the thyroid: the bane of the pathologist. Am J Clin Pathol, 117:143– 150. http://dx.doi.org/10.1309/8VL9-ECXY- NVMX-2RQF PMID:11789719
16. Xing M (2013). Molecular pathogenesis and mechanisms of thyroid cancer. Nat Rev Cancer, 13:184–199. http://dx.doi. org/10.1038/nrc3431 PMID:23429735
17. Weber F, Eng C (2005). Gene-expression profiling in differentiated thyroid cancer – a viable strategy for the practice of genomic medicine? Future Oncol, 1:497–510. http:// dx.doi.org/10. 2217/1479 6 69 4.1.4.497 PMID:16556026
18. Fagin JA, Mitsiades N (2008). Molecular pathology of thyroid cancer: diagnostic and clinical implications. Best Pract Res Clin Endocrinol Metab, 22:955–969. http:// dx.doi.org/10.1016/j.beem.2008.09.017 PMID:19041825
19. Xing M, Haugen BR, Schlumberger M (2013).). Progress in molecular-based management of differentiated thyroid can- cer. Lancet, 381:1058–1069. http://dx.doi. org /10.1016/S0140 - 673 6(13) 6 010 9 - 9 PMID:23668556
20. Moline J, Eng C (2011). Multiple endocrine neoplasia type 2: an overview. Genet Med, 13:755–764. http://dx.doi.org/10.1097/GIM. 0b013e318216cc6d PMID:21552134
21. Eng C (2010). Mendelian genetics of rare – and not so rare – cancers. Ann N Y Acad Sci, 1214:70–82. http://dx.doi.org/10.1111/ j.1749-6632.2010.05789.x PMID:20946573
22. Eng C (2010). Common alleles of pre- disposition in endocrine neoplasia. Curr Opin Genet Dev, 20:251–256. http://dx. doi.org/10.1016/j.gde.2010.02.004 PMID: 20211557
23. Dammann M, Weber F (2012). Personalized medicine: caught between hope, hype and the real world. Clinics (Sao Paulo), 67 Suppl 1:91–97. http://dx.doi.org/10.6061/ clinics/2012(Sup01)16 PMID:22584712
Websites Cancer Research UK thyroid cancer home page: http://www.cancerresearchuk.org/cancer-help/ type/thyroid-cancer/
Online Mendelian Inheritance in Man, Familial Medullary Thyroid Carcinoma: http://omim.org/entry/155240
Chapter 5.16 • Tumours of the nervous system 511
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
6
Summary • Tumours of the nervous system
account for less than 2% of all cancers but have a marked im- pact on cancer morbidity and mortality.
• Nervous system tumours oc- cur throughout life. Diffusely infiltrating gliomas manifest pre- dominantly in adults; embryonal malignancies, including medul- loblastoma and neuroblastoma, develop typically in children.
• Therapeutic ionizing radiation is the only proven cause of brain cancer. The use of mobile phones remains under investigation.
• Glioblastomas are the most com- mon and most malignant central nervous system neoplasms. Most of them (90%) manifest after a short clinical history in elderly patients (mean age, 60 years). Genetic hallmarks in- clude EGFR amplification and PTEN mutations.
• Secondary glioblastomas de- velop in younger patients (mean age, 45 years) through progres- sion from low-grade or anaplastic astrocytoma. Genetic hallmarks include IDH1 and TP53 muta- tions. Distinguishing genetic al- terations have been identified for
5.16 5 ORGAN SITE
Tumours of the nervous system Paul Kleihues Jill Barnholtz-Sloan Hiroko Ohgaki
Webster K. Cavenee (reviewer) Werner Paulus (reviewer)
other nervous system tumours, including oligodendroglioma and ependymoma.
• A range of targeted agents have been indicated, but establish- ment of optimal therapy has not been achieved. At present, the prognosis for most patients with brain cancer is poor.
Brain tumours account for less than 2% of the overall human cancer bur- den. However, they cause signifi- cant morbidity, and for gliomas, the most common histological type of central nervous system neoplasms, the prognosis is still poor. This ap- plies particularly to glioblastomas, the most common and most ma- lignant brain tumours in adults. Malignant embryonal tumours typi- cally manifest in children and oc- cur in the central nervous system in the case of medulloblastoma, and in the sympathetic nervous system and adrenal glands in the case of neuroblastoma. Meningiomas origi- nate in the meninges, the protective coverings of the brain, and are usu- ally benign and are more common in women. The brain is also a com- mon site of metastases, most often due to carcinomas of the breast and lung.
Etiology With the exception of brain tumours associated with inherited cancer syndromes and causation by thera- peutic irradiation, which accounts for very few cases, no causative envi- ronmental or lifestyle risk factor has been unequivocally identified.
Environmental factors Therapeutic irradiation of the head and neck regions is known to have caused brain cancer. In particular,
Fig. 5.16.1. Protocols for therapeutic irradiation of children now recognize the inherent risk such radiation presents and are designed to minimize any such risk.
512
Epidemiology
Nervous system tumours • Central nervous system tumours
represent the 17th most common cancer worldwide, with an esti- mated 256 000 new cases an- nually. Rates tend to be higher in more developed countries and therefore, combined with the rather high case fatality rate, these cancers are the 12th most frequent cause of cancer-related death worldwide.
• Comparisons of rates interna- tionally are hampered by under- reporting in many regions, and this may largely explain the low incidence and mortality rates observed in sub-Saharan Africa and some Asian countries.
• Incidence rates are highest in North America, Europe, and Australia, typically 4–8 new cas- es per 100 000 people annually, but are higher in some Nordic countries. In multi-ethnic com- munities, adults and children of African or Asian descent have an approximately 2-fold lower risk than people with European ancestry.
• The age distribution of brain tu- mours is bimodal, with a peak incidence in children and a sec- ond, larger peak in adults aged 45–70 years. Glioblastomas are more frequent in men, while benign meningiomas are signifi- cantly more common in women.
Map 5.16.1. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for tumours of the nervous system in men, 2012.
Map 5.16.2. Global distribution of estimated age-standardized (World) incidence rates (ASR) per 100 000, for tumours of the nervous system in women, 2012.
children treated with radiotherapy for acute myeloid leukaemia show a significantly elevated risk of develop- ing malignant gliomas, often within 10 years [1]. Radiation-induced meningiomas may follow low-dose irradiation for tinea capitis, a fungal infection of the scalp, and high-dose irradiation for primary brain tumours.
A recent systematic review of eight cohort studies assessing the rela- tionship between ionizing radiation and risk of brain tumours showed that the strength of the association varied between studies and that ion- izing radiation was generally more strongly associated with risk of me- ningioma than with risk of glioma [2].
Electromagnetic fields Exposure to occupational and resi- dential power-frequency electromag- netic fields has been a subject of public concern. A weak association of brain tumours with occupational exposure to magnetic fields has been observed in some studies [3]; others failed to reveal a significantly
For more details about the maps and charts presented in this chapter, see “A guide to the epidemiology data in World Cancer Report”.
Chapter 5.16 • Tumours of the nervous system 513
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
6
elevated risk [4]. The available evi- dence is insufficient to establish that electromagnetic fields cause brain tumours.
Mobile phones No consistent association has been found between use of mobile (cell) phones and brain tumours [5–9]. The Interphone study [8] recorded no overall elevated odds ratio 10 years or less after first mobile phone use. In the 10th (highest) decile of re- called cumulative call time, the odds ratio was 1.40 (95% confidence in- terval [CI], 1.03–1.89) for glioma and 1.15 (95% CI, 0.81–1.62) for menin- gioma. The odds ratios for glioma were somewhat higher in the tem- poral lobe. Analysis of United States data indicated possible consistency with the modest excess risks ob- served in highly exposed people in the Interphone study [9]. In 2011, an IARC Monographs evaluation
determined that the evidence for causation of glioma by mobile phone use was limited, resulting in a classification as Group 2B (pos- sibly carcinogenic to humans) [10]. The evidence for this association remains inconclusive because of, among other things, lack of data about long-term mobile phone use [11] (see “An IARC announcement that made waves” in Chapter 2.8).
Allergic conditions Several studies suggest that having a personal medical history of specific allergic conditions, such as asthma, eczema, and hay fever, has a protec- tive effect against development of glioma [12].
Environmental carcinogens Several occupational exposures and environmental carcinogens have been reported to be slightly associated
with gliomagenesis. However, many of these reports have not been vali- dated in independent studies, and none of these exposures has been unequivocally identified as causative [1]. Some studies suggest a causal relationship between parental occu- pational exposure to carcinogens and central nervous system tumours in their offspring, without identifying any specific environmental agent.
Diet N-nitroso compounds have been detected in nitrite-preserved food and can be formed in the stomach after uptake of their chemical pre- cursors, nitrate/nitrite and second- ary amines. Vitamins C and E inhibit the formation of nitroso compounds from these precursors. Since some N-nitroso compounds are potent neurocarcinogens in experimental animals, several epidemiological studies have addressed their role in
Chart 5.16.1. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for tumours of the nervous system in men, circa 1975–2012.
Chart 5.16.2. Age-standardized (World) incidence rates per 100 000 by year in selected populations, for tumours of the nervous system in women, circa 1975–2012.
514
the etiology of human brain tumours. A meta-analysis of nine published studies yielded an elevated risk of glioma among adults frequently consuming cured meat, bacon, and ham, but the data were insufficient to establish a dose–response rela- tionship [13].
Viral infections JC virus has neuro-oncogenic po- tential in experimental animals, but JC virus sequences are rarely de- tectable in human brain tumours [14]. The polyomavirus SV40 was ia- trogenically introduced into human populations in 1955 –1962 through SV40-contaminated polio vaccines. SV40 sequences have been identi- fied in a variety of human neoplasms, including brain tumours, raising the possibility of an etiological role [15]. SV40 sequences have been report- ed to be present at variable frequen- cies in brain tumour biopsies from countries that had used contami- nated polio vaccine [14]. However, assays are easily contaminated with plasmids containing SV40 sequences, and recent studies
failed to detect SV40 sequences in brain tumours [16].
Genetic factors Central nervous system tumours oc- cur in the setting of several inherited cancer syndromes (Table 5.16.1). However, their overall contribution amounts to less than 1% of brain tumour cases.
Genetic basis of non-syndromic familial aggregation There is a small but significant increased risk of primary brain tumours among relatives of patients with brain tumours. Cohort studies showed a standardized incidence ratio of approximately 2, i.e. a 2-fold increased risk of primary brain tu- mour or glioma among first-degree relatives of glioma probands, com- pared with population incidence data [17]. In a population-based study from Sweden and Norway, the standardized incidence ratio for glioma was 1.8 (95% CI, 1.5 –2.0) when a parent was a proband, and increased to 11.2 (95% CI, 5.7–19.5) in multiplex families in which a
parent and at least two siblings were affected by a nervous sys- tem tumour [18]. A pooled famil- ial aggregation analysis of glioma probands showed an increased risk of gliomas, sarcomas, and melano- mas [19]. Genome-wide associa- tion studies have identified several low-penetrance susceptibility al- leles for glioma [20,21].
Detection Signs and symptoms of brain tumours largely depend on the location of the neoplasm and include paresis, speech disturbances, and personality chang- es. Patients with oligodendroglioma often have a long history of epileptic seizures. Eventually, malignant brain tumours cause life-threatening intra- cranial pressure that ultimately leads to unconsciousness and respiratory arrest. Since the brain does not con- tain pain receptors, headache is typi- cally present if the tumour infiltrates the meninges. The presence of symp- toms usually leads to a detailed neuro- logical examination, using techniques such as computed tomography and magnetic resonance imaging.
Pathology and genetics The WHO Classification of Tumours of the Central Nervous System (4th edition) recognizes more than 50 clin- icopathological entities with a great variation in histology, biological be- haviour, response to therapy, and clinical outcome [22]. Clinical and ge- netic data on the most common brain tumours are detailed in Table 5.16.2.
Astrocytic tumours Tumours of astrocytic origin con- stitute the largest proportion of gliomas.
Pilocytic astrocytoma Pilocytic astrocytoma (WHO grade I), the most common central ner- vous system neoplasm in children, is predominantly located in the cerebellum and midline structures, including the optic tract, brain stem, and spinal cord. It infiltrates adjacent brain structures but grows
Fig. 5.16.2. Macroscopic features of frequently occurring brain tumours.
Chapter 5.16 • Tumours of the nervous system 515
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
6
slowly and usually has a favoura- ble prognosis, with 5-year survival rates of more than 85%. Some pi- locytic astrocytomas occur in the setting of neurofibromatosis type 1 (NF1), particularly those of the optic nerve (optic glioma). Pilocytic as- trocytomas typically show BRAF- KIAA1549 fusions [23].
Other types of astrocytoma (WHO grades II–IV) usually de- velop in the cerebral hemispheres of adults and diffusely infiltrate adjacent brain structures (diffuse astrocytomas).
Low-grade diffuse astrocytoma Low-grade diffuse astrocytomas (WHO grade II) occur in young adults and grow slowly. However, they in- filtrate adjacent brain structures and typically cannot be completely sur- gically resected. Morphologically, tumour cells resemble differentiated
astrocytes. IDH1 or IDH2 mutations (> 80%), TP53 mutations (~65%), and ATRX mutations (65%) are early and frequent genetic events [24,25].
Anaplastic astrocytoma Anaplastic astrocytomas (WHO grade III) often develop from low-grade astrocytomas, grow relatively quickly, and typically progress to glioblasto- ma within 2–3 years, accompanied by genetic alterations.
Glioblastoma Glioblastoma (WHO grade IV) is the most frequent and most ma- lignant nervous system tumour. Glioblastomas diffusely infiltrate the brain, including the opposite hemisphere, and show high cellu- larity and large areas of necrosis despite excessive vascular prolif- eration (Fig. 5.16.3).
Secondary glioblastomas develop by malignant progression from low- grade and anaplastic astrocytoma and are characterized by frequent IDH1 or IDH2 mutations, TP53 mu- tations, ATRX mutations, and loss of heterozygosity on chromosome 10q and 19p [26]. Primary glioblastomas are more frequent, accounting for approximately 90% of glioblasto- mas, and develop rapidly in older adults, with a short clinical history of usually less than 3 months (Fig. 5.16.4). Their genetic profile includes amplification and overexpression of the EGFR gene, PTEN mutations, TERT promoter mutations, and loss of heterozygosity on chromosome 10p and 10q [26,27].
Glioblastomas have also been classified on the basis of comple- mentary DNA expression profiles, with distinct proneural, neural, clas- sic, and mesenchymal patterns [28]. Secondary glioblastomas with
Table 5.16.1. Familial cancer syndromes predisposing to the development of tumours of the nervous system
Syndrome Gene (chromosome) Nervous system tumours Other lesions
Neurofibromatosis type 1 (MIM no. 162200)
NF1 (17q11.2) Neurofibroma, malignant peripheral nerve sheath tumour, optic nerve glioma, astrocytoma
Café au lait spots, axillary freckling, iris hamartomas, osseous lesions, phaeochromocytoma, leukaemia
Neurofibromatosis type 2 (MIM no. 101000)
NF2 (22q12) Schwannoma, meningioma, meningioangiomatosis, spinal ependymoma, astrocytoma
Posterior lens opacities, retinal hamartoma
Tuberous sclerosis TSC1 and TSC2 (MIM no. 191100, 191092)
TSC1 (9q34) TSC2 (16p13.3)
Subependymal giant cell astrocytoma
Cutaneous angiofibroma, peau chagrin, subungual fibroma, cardiac rhabdomyoma, cysts of lung and kidney, renal angiomyolipoma
Li–Fraumeni syndrome (MIM no. 151623, 191170)
TP53 (17p13) Astrocytoma, secondary glioblastoma, medulloblastoma
Adrenocortical carcinoma, soft tissue and bone sarcomas, leukaemia, breast carcinoma
Turcot syndrome type 1 (MIM no. 276300)
MLH1 (3p21) PMS2 (7p22) MSH2 (2p21)
Astrocytoma, glioblastoma Café au lait spots Colorectal polyps (< 100) Large polyps (> 3 cm), colorectal cancer
Turcot syndrome type 2 (MIM no. 276300)
APC (5q21) Medulloblastoma Small colorectal polyps (> 100) Colorectal cancer
Naevoid basal cell carcinoma syndrome (NBCCS)/Gorlin syndrome (MIM no. 109400)
PTCH (9q22) Medulloblastoma Basal cell carcinomas, palmar and plantar pits, jaw cysts, ovarian fibroma, skeletal abnormalities
Rhabdoid tumour predisposition syndrome (RTPS) (MIM no. 609322)
INI1 (22q11.2) Atypical teratoid/rhabdoid tumour (AT/RT)
Malignant rhabdoid tumour of the kidney
MIM, Mendelian Inheritance in Man (www.omim.org).
516
IDH1 mutation have a proneural ex- pression profile. Additional genetic changes typical of the proneural type include PDGFR amplification or PIK3CA/PIK3R1 mutations and loss or mutation of TP53 [29]. The neural subtype is related to the clas- sic expression pattern, which shows chromosome 7 amplification, focal CDKN2A deletion, chromosome 10 loss, and EGFR amplification or mu- tation. The mesenchymal subtype has frequent mutation or loss of NF1, TP53, and PTEN, and further chro- mosomal aberrations at CDK6, MET, PTEN, CDKN2A, and RB1 loci [29].
New drivers of glioblastoma have been identified [30]. LZTR1 and CTNND2 mutations and deletions, and EGFR-SEPT14 gene fusions are proposed to all enhance the self-re- newal and transformation of glioblas- toma stem cells, thereby contributing to glioblastoma progression.
Key genomic alterations in glio- blastoma suggest a range of targeted agents (Fig. 5.16.5). Particular path- ways may be identified as a basis for therapy (see “Glioma genomics and its implications in neuro-oncology”). The role of the alkylation-damage DNA repair gene MGMT in mediat- ing response of malignant glioma
to alkylating agents such as temo- zolomide has been clarified. Two randomized trials demonstrated that patients with an epigenetically silenced MGMT gene fared better when treated with temozolomide, whereas patients with an unmethyl- ated MGMT promoter had a longer survival with initial radiotherapy [31].
Oligodendroglial tumours These neoplasms are assumed to develop from myelin-producing oli- godendroglial cells or their precur- sors and are typically found in the cerebral hemispheres of adults, often including the basal ganglia. Histologically, they are isomorphic, with a typical honeycomb pattern and delicate tumour vessels (a “chicken wire” pattern). Anaplastic oligoden- drogliomas show features of ana- plasia and high mitotic activity and have a less favourable prognosis. Oligodendrogliomas share with dif- fuse astrocytomas common and fre- quent genetic alterations, i.e. IDH1 or IDH2 mutations, suggesting that as- trocytomas and oligodendrogliomas arise from the same precursor cells. Genetic hallmarks of oligodendro- gliomas are frequent IDH1 or IDH2 mutations (> 80%), co-deletion of
chromosomes 1p and 19q (~70%), CIC mutations (~40%), and FUBP1 mutations (~15%) [26].
Ependymomas These gliomas develop from the ependymal lining of the cerebral ventricles and the central canal of the spinal cord. They manifest preferentially in children and young adults. Histologically, they are cel- lular, with typical perivascular ro- settes. Spinal ependymomas show a high frequency of mutations in the neurofibromatosis gene NF2 [22].
Glioneuronal tumours This group of brain tumours is less frequent and generally has a favour- able prognosis. Some manifest pref- erentially in children (desmoplastic infantile astrocytoma/ganglioglioma, dysembryoplastic neuroepithelial tu- mour), others preferentially in ado- lescents and adults (gangliocytoma, central neurocytoma). They often cause a long-term history of epilep- tic seizures [22].
Embryonal tumours These neoplasms are derived from embryonal or fetal precursor cells,
Table 5.16.2. Clinical and genetic data on the most common tumours of the nervous system
Tumour Age (years)a 5-Year survival Genetic alterations
Pilocytic astrocytoma (WHO grade I)
< 20 > 85% BRAF-KIAA1549 fusion, NF1 mutation (neurofibromatosis cases)
Diffuse astrocytoma (WHO grade II)
20– 45 > 60% IDH mutation, TP53 mutation, ATRX mutation
Secondary glioblastoma (WHO grade IV)
~45 ~10% IDH mutation, TP53 mutation, ATRX mutation, LOH 10q, LOH 19q
Primary glioblastoma (WHO grade IV)
~60 < 3% EGFR amplification, PTEN mutation, LOH 10p, LOH 10q
Oligodendroglioma (WHO grade II/III)
> 20 > 50% IDH mutation, 1p and 19q co-deletion, CIC mutation, FUBP1 mutation
Ependymoma (WHO grade II)
< 45 < 30% Chromosome 22 alteration, NF2 mutation
Medulloblastoma (WHO grade IV)
< 20 > 50% Isochromosome 17, TP53 mutation, PTCH mutation, b-catenin mutation, DDX3X mutation, SMARCA4 mutation
Neuroblastoma (WHO grade IV)
< 10 > 90%; 20–50%b
LOH 1p, LOH 11q, MYCN amplification, trisomy 17q, ALK mutation
LOH, loss of heterozygosity. a Age range in which 60% or more of the respective tumour types are clinically manifested. b Patients < 1 year old; patients > 1 year old.
Chapter 5.16 • Tumours of the nervous system 517
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
6
typically manifest in children, and are highly malignant but often re- spond to radiotherapy or chemo- therapy. In the central nervous sys- tem, cerebellar medulloblastomas are the most common. The peak incidence age is 3– 6 years; these tumours are uncommon in adults. Expression array analysis and DNA sequencing have led to the identi- fication of four distinct subgroups, which differ greatly in clinical out- come [32,33]. The most benign course is seen in medulloblastomas with activation of the WNT pathway, as determined by nuclear β-catenin accumulation, and the presence of CTNNB1 mutations in most cases [34]. The WNT subtype shows clas- sic medulloblastoma histology and cannot be identified morphologi- cally. The SHH type is biologically characterized by activation of the sonic hedgehog (SHH) signalling pathway, which in about one third of cases is caused by a muta- tion in the PTCH gene [35]. This subtype tends to be associated with the desmoplastic phenotype [36]. Type 3 medulloblastomas are not associated with a distinct sig- nalling pathway, but they typically show NPR3 expression and MYC amplification. Type 4 medulloblas- tomas comprise about one third of all medulloblastomas and have a less favourable clinical outcome. Type 3 and 4 medulloblastomas are characterized by a strong male pre- dominance and a greater tendency to metastasize via cerebrospinal fluid pathways.
Tumours of peripheral nerves Most of these tumours develop from myelin-producing Schwann cells and are termed neurinomas or schwanno- mas. Bilateral acoustic schwannomas are diagnostic of inherited neurofi- bromatosis type 2 (NF2). Acoustic schwannomas are benign (WHO grade I) and rarely recur after surgical resection. Neurofibromas and malig- nant peripheral nerve sheath tumours represent typical manifestations of the NF1 syndrome [22].
Meningiomas These are generally benign, slowly growing intracranial tumours at- tached to the dura mater. They are composed of neoplastic menin- gothelial (arachnoidal) cells and typically occur in adults, with a marked female predominance, particularly for those located in the spine. Meningiomas do not infiltrate the brain but may cause symptoms of intracranial pressure due to compression of adjacent brain structures. Preferential sites
Fig. 5.16.3. Typical histological features of glioblastoma, with necrosis (NE) and vascular proliferation (VP).
Fig. 5.16.4. (A) Rapid evolution of a primary glioblastoma. Magnetic resonance image shows a small cortical lesion that within 68 days developed into a full-blown glioblastoma with perifocal oedema and central necrosis. This is in contrast to (B) the development of a secondary glioblastoma through progression from low-grade astrocytoma.
A
B
518
Glioma genomics and its implications in neuro-oncology Hai Yan
Mutations in the critical chromatin modifier ATRX and mutations in CIC and FUBP1, which are potent regulators of cell growth, have been discovered in specific subtypes of gliomas, the most common type of primary malignant brain tumour. However, the frequency of these mutations in many subtypes of glio- mas, and their association with the clinical features of patients, is poorly understood. These loci were ana- lysed in 363 brain tumours. ATRX is frequently mutated in grade II/III astrocytomas (71%), oligoastrocyto- mas (68%), and secondary glioblas- tomas (57%), and ATRX mutations are associated with IDH1 mutations and with the alternative lengthening
of telomeres phenotype. CIC and FUBP1 mutations occurred fre- quently in oligodendrogliomas (46% and 24%, respectively) but rarely in astrocytomas or oligoastrocytomas (< 10%). This analysis enabled the identification of two highly recurrent genetic signatures in gliomas: IDH1/ ATRX (I-A) and IDH1/CIC/FUBP1 (I-CF).
Patients with I-CF gliomas had a significantly longer median overall survival (96 months) com- pared with patients with I-A glio- mas (51 months) and patients with gliomas that did not harbour either signature (13 months). The genetic signatures distinguished clinically distinct groups of oligoastrocytoma
patients, who usually present a diag- nostic challenge, and were associ- ated with differences in clinical out- come even among individual tumour types. In addition to providing new clues about the genetic alterations underlying gliomas, the results have immediate clinical implications, pro- viding a tripartite genetic signature that can serve as a useful adjunct to conventional glioma classification and that may aid in prognosis, treat- ment selection, and therapeutic trial design.
Reference Duncan CG et al. (2012). Genome Res, 22:2339–2355. http://dx.doi.org/10.1101/gr.132 738.111 PMID:22899282
are the cerebral hemispheres [22]. Meningiomas can often be cured by surgical resection. Atypical or malignant meningiomas are less frequent (~5% of all meningiomas) (http://www.cbtrus.org) and may infiltrate the brain and often recur locally.
Neuroblastomas This tumour develops from the primitive neural crest and is the most frequent neoplasm in chil- dren younger than 1 year, with an incidence of approximately 1 per 100 000. It is typically located in the adrenal gland and sympathetic nervous system and is responsible for up to 15% of all cancer deaths in children. Neuroblastomas are heterogeneous with respect to the degree of maturation, ranging from highly malignant examples to be- nign ganglioneuroma.
More than 60% of neuroblas- tomas are metastatic, often carry- ing MYCN amplification or ATRX mutation, and/or ALK mutation. Neuroblastoma can evade T cells and natural killer cells. Anti-GD2 antibodies, when combined with
granulocyte-macrophage colony- stimulating factor with or without interleukin-2, offer strategies for
a curative approach to therapy. Neuroblastomas with single copy MYCN or similar indicators are
RAS PI3K
AKT
mTOR
MAPK pathway
S-phase genes
RB INK4A
Proliferation
Cell-cycle arrest
ARF/MDM2/p53
Apoptosis
MET, EGFR, PDGFR
NF1 PTEN
Fig. 5.16.5. Cell-cycle dysregulation in glioblastoma is in part a result of abnormal sig- nalling by several different receptor tyrosine kinases (RTKs), including EGFR, PDGFR, and MET. Genes shown in red may be targeted by agents, including RTK inhibitors. (For details, see [29].)
Chapter 5.16 • Tumours of the nervous system 519
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
6
highly curable with surgery alone or with surgery and low-dose chemo- therapy [37].
Prospects Although brain tumours do not oc- cur very frequently, they contribute significantly to morbidity, often affect
children, and have an overall poor prognosis. Due to marked resistance to radiation and chemotherapy, the prognosis for patients with glioblas- tomas, the most common type of malignant brain tumour in adults, is still very poor. Unfortunately no en- vironmental, lifestyle, or genetic risk
factor has been discovered that ac- counts for a significant proportion of newly diagnosed brain tumours. Many genetic alterations involved in the development of nervous system tumours have been identified and may lead to novel targeted therapeu- tic approaches.
Glial progenitor cells Glial progenitor cells Glial progenitor cells
Common precursor cells with IDH1/2 mutation
TP53 mutation (>65%) ATRX mutation (>65%)
Loss 1p/19q (>75%) CIC mutation ( ̴40%) FUBP1 mutation ( ̴15%)
Diffuse astrocytoma Oligodendroglioma
Anaplastic astrocytoma
Anaplastic oligodendroglioma
̴5 years
Primary glioblastoma
Secondary glioblastoma
LOH 19q ( 5̴0%) LOH 10q (>60%)
̴2 years
EGFR amp ( ̴35%) TP53 mutation ( ̴30%) PTEN mutation ( ̴25%) LOH 10p ( ̴50%) LOH 10q ( ̴70%)
3-6 months
Pilocytic astrocytoma
BRAF-KIAA1549 fusion (>60%)
WHO grade
I
II
III
IV
Fig. 5.16.6. Current concept of genetic pathways giving rise to primary and secondary gliomas.
520
References
1. Ohgaki H, Kleihues P (2005). Epidemiology and etiology of gliomas. Acta Neuropathol, 109:93–108. http://dx.doi.org/10.1007/ s00401-005-0991-y PMID:15685439
2. Braganza MZ, Kitahara CM, Berrington de González A et al. (2012). Ionizing radiation and the risk of brain and cen- tral nervous system tumors: a system- atic review. Neuro Oncol, 14:1316–1324. http://dx.doi.org/10.1093/neuonc/nos208 PMID:22952197
3. Villeneuve PJ, Agnew DA, Johnson KC, Mao Y; Canadian Cancer Registries Epidemiology Research Group (2002). Brain cancer and occupational exposure to magnetic fields among men: results from a Canadian population-based case- control study. Int J Epidemiol, 31:210– 217. http://dx.doi.org/10.1093/ije/31.1.210 PMID:11914323
4. Thériault G, Goldberg M, Miller AB et al. (1994). Cancer risks associated with oc- cupational exposure to magnetic fields among electric utility workers in Ontario and Quebec, Canada, and France: 1970– 1989. Am J Epidemiol, 139:550–572. PMID:8172168
5. Hardell L, Mild KH, Carlberg M (2002). Case-control study on the use of cel- lular and cordless phones and the risk for malignant brain tumours. Int J Radiat Biol, 78:931–936. http://dx.doi. org/10.1080/09553000210158038 PMID: 12465658
6. Inskip PD, Tarone RE, Hatch EE et al. (2001). Cellular-telephone use and brain tumors. N Engl J Med, 344:79–86. http://dx.doi.org/10.1056/NEJM2001011 13440201 PMID:11150357
7. Christensen HC, Schüz J, Kosteljanetz M et al. (2005). Cellular telephones and risk for brain tumors: a population- based, incident case-control study. Neurology, 64:1189–1195. http://dx.doi. org/10.1212/01.WNL.0000156351.72313. D3 PMID:15824345
8. INTERPHONE Study Group (2010). Brain tumour risk in relation to mobile telephone use: results of the INTERPHONE interna- tional case-control study. Int J Epidemiol, 39:675–694. http://dx.doi.org/10.1093/ije/ dyq079 PMID:20483835
9. Little MP, Rajaraman P, Curtis RE et al. (2012). Mobile phone use and glioma risk: comparison of epidemiological study results with incidence trends in the United States. BMJ, 344:e1147. http://dx.doi.org/10.1136/ bmj.e1147 PMID:22403263
10. Baan R, Grosse Y, Lauby-Secretan B et al.; WHO International Agency for Research on Cancer Monograph Working Group (2011). Carcinogenicity of radiofrequency electromagnetic fields. Lancet Oncol, 12:624–626. http://dx.doi.org/10.1016/S14 70-2045(11)70147-4 PMID:21845765
11. Ostrom QT, Barnholtz-Sloan JS (2011). Current state of our knowledge on brain tumor epidemiology. Curr Neurol Neurosci Rep, 11:329–335. http://dx.doi.org/10.1007/ s11910-011-0189-8 PMID:21336822
12. Chen C, Xu T, Chen J et al. (2011). Allergy and risk of glioma: a meta-analysis. Eur J Neurol, 18:387–395. http://dx.doi. org/10.1111/j.1468 -1331. 2010.03187.x PMID:20722711
13. Huncharek M, Kupelnick B, Wheeler L (2003). Dietary cured meat and the risk of adult glioma: a meta-analysis of nine obser- vational studies. J Environ Pathol Toxicol Oncol, 22:129–137. PMID:14533876
14. Huang H, Reis R, Yonekawa Y et al. (1999). Identification in human brain tumors of DNA sequences specific for SV40 large T antigen. Brain Pathol, 9:33– 42. http:// dx.doi.org/10.1111/j.1750-3639.1999.tb00 207.x PMID:9989448
15. Vilchez RA, Kozinetz CA, Arrington AS et al. (2003). Simian virus 40 in human can- cers. Am J Med, 114:675–684. http://dx.doi. org/10.1016/S0 0 02- 93 43(03) 0 0 0 87-1 PMID:12798456
16. Rollison DE, Utaipat U, Ryschkewitsch C et al. (2005). Investigation of human brain tumors for the presence of polyomavirus genome sequences by two independent laboratories. Int J Cancer, 113:769–774. http://dx.doi.org/10.1002/ijc.20641 PMID: 15499616
17. Malmer B, Adatto P, Armstrong G et al. (2007). GLIOGENE – an interna- tional consortium to understand familial glioma. Cancer Epidemiol Biomarkers Prev, 16:1730–1734. http://dx.doi.org/ 10.1158/1055-9965.EPI-07-0081 PMID: 17855690
18. Hemminki K, Tretli S, Sundquist J et al. (2009). Familial risks in nervous-system tumours: a histology-specific analysis from Sweden and Norway. Lancet Oncol, 10:481– 488. http://dx.doi.org/10.1016/S14 70-2045(09)70076-2 PMID:19356978
19. Scheurer ME, Etzel CJ, Liu M et al.; GLIOGENE Consortium (2010). Familial ag- gregation of glioma: a pooled analysis. Am J Epidemiol, 172:1099–1107. http://dx.doi. org/10.1093/aje/kwq261 PMID:20858744
20. Wrensch M, Jenkins RB, Chang JS et al. (2009). Variants in the CDKN2B and RTEL1 regions are associated with high- grade glioma susceptibility. Nat Genet, 41:905–908. http://dx.doi.org/10.1038/ng. 408 PMID:19578366
21. Liu Y, Shete S, Hosking FJ et al. (2010). New insights into susceptibility to gli- oma. Arch Neurol, 67:275–278. http:// dx.doi.org/10.10 01/archneurol. 2010.4 PMID:20212223
22. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds (2007). WHO Classification of Tumours of the Central Nervous System, 4th ed. Lyon: IARC.
23. Korshunov A, Meyer J, Capper D et al. (2009). Combined molecular analysis of BRAF and IDH1 distinguishes pilocytic astrocytoma from diffuse astrocytoma. Acta Neuropathol, 118:401– 405. http:// dx.doi.org/10.1007/s00401-009-0550-z PMID:19543740
24. Liu XY, Gerges N, Korshunov A et al. (2012). Frequent ATRX mutations and loss of expression in adult diffuse astrocytic tu- mors carrying IDH1/IDH2 and TP53 mu- tations. Acta Neuropathol, 124:615–625. http://dx.doi.org/10.10 07/s0 0 401- 012- 1031-3 PMID:22886134
25. Jiao Y, Killela PJ, Reitman ZJ et al. (2012). Frequent ATRX, CIC, FUBP1 and IDH1 mutations refine the classification of ma- lignant gliomas. Oncotarget, 3:709–722. PMID:22869205
26. Ohgaki H, Kleihues P (2013). The definition of primary and secondary glioblastoma. Clin Cancer Res, 19:764–772. http://dx.doi. org/10.1158/1078 - 0 432.CCR-12-30 02 PMID:23209033
27. Killela PJ, Reitman ZJ, Jiao Y et al. (2013). TERT promoter mutations occur frequently in gliomas and a subset of tumors derived from cells with low rates of self-renewal. Proc Natl Acad Sci U S A, 110:6021–6026. http://dx.doi.org/10.1073/pnas.1303607110 PMID:23530248
28. Phillips HS, Kharbanda S, Chen R et al. (2006). Molecular subclasses of high- grade glioma predict prognosis, delineate a pattern of disease progression, and re- semble stages in neurogenesis. Cancer Cell, 9:157–173. http://dx.doi.org/10.1016/j. ccr.2006.02.019 PMID:16530701
29. Kotliarova S, Fine HA (2012). SnapShot: glioblastoma multiforme. Cancer Cell, 21:710, e1. http://dx.doi.org/10.1016/j.ccr. 2012.04.031 PMID:22624719
30. Frattini V, Trifonov V, Chan JM et al. (2013). The integrated landscape of driver geno- mic alterations in glioblastoma. Nat Genet, 45:1141–1149. http://dx.doi.org/10.1038/ ng.2734 PMID:23917401
31. Stupp R, Hegi ME (2013). Brain cancer in 2012: molecular characterization leads the way. Nat Rev Clin Oncol, 10:69–70. http:// dx.doi.org/10.1038/nrclinonc.2012.240 PMID:23296110
Chapter 5.16 • Tumours of the nervous system 521
5 O
R G
A N
S IT
E C
H A
P T
E R
5 .1
6
32. Kool M, Korshunov A, Remke M et al. (2012). Molecular subgroups of me- dulloblastoma: an international meta- analysis of transcriptome, genetic aber- rations, and clinical data of WNT, SHH, Group 3, and Group 4 medulloblastomas. Acta Neuropathol, 123:473– 484. http:// dx.doi.org/10.1007/s00401- 012- 0958-8 PMID:22358457
33. Northcott PA, Korshunov A, Pfister SM, Taylor MD (2012). The clinical implica- tions of medulloblastoma subgroups. Nat Rev Neurol, 8:340–351. http://dx. doi.org/10.1038/nrneurol.2012.78 PMID: 22565209
34. Ellison DW, Onilude OE, Lindsey JC et al.; United Kingdom Children’s Cancer Study Group Brain Tumour Committee (2005). β-Catenin status predicts a favorable outcome in childhood medulloblastoma: the United Kingdom Children’s Cancer Study Group Brain Tumour Committee. J Clin Oncol, 23:7951–7957. http://dx. doi.org/10.1200/JCO.2005.01.5479 PMID: 16258095
35. Schwalbe EC, Lindsey JC, Straughton D et al. (2011). Rapid diagnosis of medulloblas- toma molecular subgroups. Clin Cancer Res, 17:1883–1894. http://dx.doi.org/ 10.1158/1078-0432.CCR-10-2210 PMID: 21325292
36. Pietsch T, Waha A, Koch A et al. (1997). Medulloblastomas of the desmoplastic var- iant carry mutations of the human homo- logue of Drosophila patched. Cancer Res, 57:2085–2088. PMID:9187099
37. Cheung NK, Dyer MA (2013). Neuroblastoma: developmental biology, cancer genomics and immunotherapy. Nat Rev Cancer, 13:397– 411. http://dx.doi. org/10.1038/nrc3526 PMID:23702928
Summary Cancer screening controversies have raged for a long time. However, there are paths forward that could modulate the content and tone of the discussions. They involve use of emerging molecular techniques to help us refine our assessment of screen-detected cancers, profes- sional education, and more nuanced information for the public. Time will tell if such strategies provide resolu- tions to the controversies.
Cancer screening is one of the most contentious areas in medicine. Debates over the balance of benefits and harms of screening tests play out at national conferences, in the medical literature, in the media, and occasionally in Congress. The intui- tive sense that detection of asymp- tomatic cancers must necessarily be of benefit is strong, sometimes triggering arguments when the ben- efits of screening are questioned or harms are reported.
The concept that early detection of cancer before clinically evident symptomatic presentation would lead to cure in the majority of cases has its origins more than a century ago. In 1907 Dr Charles Childe, in his book The Control of a Scourge,
Barnett “Barry” S. Kramer is the
director of the Division of Cancer
Prevention at the United States
National Cancer Institute (NCI)
and serves as editor-in-chief of
NCI’s Physician Data Query (PDQ)
Screening and Prevention Editorial
Board. Board-certified in internal
medicine and medical oncology,
Dr Kramer earned his medical degree
from the University of Maryland
Medical School and also holds a
master’s degree in public health from
Johns Hopkins University’s Bloomberg
School of Public Health. Dr Kramer
has extensive experience in primary
cancer prevention studies as well
as clinical screening trials of lung,
ovarian, breast, and prostate cancers.
From 1994 to 2012, he was editor-in-
chief of the Journal of the National
Cancer Institute. Dr Kramer’s long-
standing interest in the challenges of
reporting medical research led him
to create the Medicine in the Media
Workshop to help train journalists to
review medical findings critically and
to report them accurately.
Controversies in cancer screening and their resolution: a view from the United States “battleground”
or How Cancer is Curable, asserted that if people were to only heed the earliest, apparently trivial, symp- toms or signs of cancer, “it requires no stretch of the imagination… to say that the majority… would be cured” [1]. In 1924, Dr Joseph Colt Bloodgood of Johns Hopkins Hospital was quoted in the New York Times as stating that “deaths from cancer would be practically eliminated” by a careful search for “growths in any part of the body” [2]. (In retrospect, we now know how far off the mark these assertions were.) Nevertheless, the allure of the logic remains strong, and often inde- pendent of the actual evidence for several screening tests in common use. But the concept is easy to grasp by health professionals and the public alike. Public health messages have frequently been unequivocal, often at the expense of appropriate nuance [3]. Some cancer advocacy groups have focused on screening as the most important tool in the “war on cancer”, sometimes spawning health fairs dedicated to screening as many people as possible [4,5].
The screening message has taken hold, at least in the USA. Data from the United States National Health Interview Survey (NHIS) from 2005 and 2008 indicate that high rates of cancer screening persist even among people aged 80 years
Barnett S. Kramer
522
and older – a population unlikely to derive substantial benefit from screening but nevertheless at risk for all of the harms. More than 50% of the respondents aged 75 years and older reported that their physi- cian continued to actively recom- mend screening [6]. Perhaps even more strikingly, an analysis of the Surveillance, Epidemiology and End Results (SEER)-Medicare database showed that a sizeable proportion of patients aged 65 years and older already diagnosed with advanced incurable lung, colorectal, pancre- atic, gastrointestinal, or breast can- cers (with a median survival time of 4.3–16.2 months) were still be- ing screened with mammography, Pap tests, prostate-specific antigen (PSA) tests, and lower gastrointesti- nal endoscopy [7]. The assumption of benefit is so strong that the motives of anyone who raises the possibility of screening-associated harms are sometimes attacked. For example, a mammographer recently dismissed a study suggesting that the harms of breast cancer screening due to overdetection of non-life-threatening lesions (“overdiagnosis”) have been underestimated [8] as “malicious nonsense” driven simply by a desire to reduce health-care costs [9]. But the genesis of the disagreements goes beyond financial incentives on either side of the argument.
The numerator/denominator problem Public health is in some sense a sci- ence of the denominator (the general population), while clinical medicine is a science of the numerator (peo- ple plucked out of the denominator to become patients). Each of the two disciplines has separate train- ing programmes, and the respective trainees acquire distinctly differ- ent heuristics and mental shortcuts when interpreting and applying evi- dence [10]. The target population is usually the relatively healthy general population, but the testing gener- ally takes place in a clinical setting. Cancer screening therefore sits at the interface between these two
worlds, inviting a clash of philoso- phies that are anchored in totally dif- ferent worlds.
Health-care professionals are strongly influenced by personal ex- perience, and that is as it should be. Cumulative experience is a core element in the evolution of clinical judgement. However, it has been shown empirically that personal ex- perience can also distort diagnostic reasoning, termed “availability bias” [11]. Cancer specialists, who treat the numerator of cancer patients, witness the suffering of their patients on a daily basis and naturally em- brace strategies that could prevent that suffering, even if not based on the strongest achievable evidence. In fact, they may be impatient with public health-oriented professionals, who demand high-quality evidence before making recommendations that affect hundreds of thousands or even millions of healthy people. One of the core principles of cancer screening and prevention is that it is difficult to make healthy people bet- ter off than they already are – but not difficult to make them worse off. This numerator/denominator issue also probably accounts for the fact that clinical specialty societies frequently have more aggressive screening recommendations than do general practice specialties [12].
A similar phenomenon occurs in the public. Healthy people in- habit the denominator, not generally knowing which numerator(s) they will ultimately enter. However, the di- agnosis of cancer suddenly changes that perspective. Here, availability bias is particularly personal, bring- ing a desire to benefit others with the newly acquired perspective. Cancer advocacy groups often have more aggressive approaches to screen- ing for cancer than do broad-based health advocacy groups.
Direct experience may distort perceived outcomes of screening However, there is more than avail- ability bias at work in the clinical setting. Personal experience may
more directly affect a clinician’s perceptions of the value of screen- ing, both increasing the magni- tude of perceived benefits and de- creasing perceived harms [13,14]. Several well-known biases artefac- tually amplify the apparent benefit of a cancer screening test, and may even make a completely ineffective test appear highly effective in the eyes of an individual clinician. The ultimate benefit of a screening test is determined by its effect on over- all or disease-specific mortality, but individual clinicians cannot observe mortality changes in practice. They do, however, directly observe the survival of their patients after diag- nosis, and these observations can seriously mislead.
People self-select for screen- ing, introducing confounding fac- tors that are associated both with the propensity to be screened and with favourable health outcomes independent of the actual screen- ing – an effect known as healthy volunteer bias or healthy screenee bias. For example, the number of observed deaths from a variety of causes unrelated to the target can- cers was substantially less than expected in participants in a large randomized screening trial for pros- tate, lung, colorectal, and ovarian cancers [15]. This reduction even extended to deaths from accidents or poisoning, highly unlikely to be affected by the actual screening tests.
Lead-time bias occurs with every screening test because the date of diagnosis of screen-detected cancers is moved up, lengthening the apparent survival time even if the date and cause of death are unaltered. For example, if a can- cer killed all of its victims on the fourth anniversary of diagnosis, the 5-year survival would be zero. If a new (but ineffective) screening test advanced the date of diagnosis by 3 years without changing the risk of death, the 5-year survival rate would be 100%. Such a dramatic in- crease in survival time would make most clinicians true believers in the test.
Controversies in cancer screening and their resolution: a view from the United States “battleground” 523
Most, if not all, cancer screen- ing tests are also associated with length-biased sampling. This occurs because they are better at detecting slow-growing asymptomatic cancers than they are at picking up the most rapidly growing tumours that come to clinical attention between scheduled screenings, even when adjusted for clinical stage and histopathological phenotypic features [16]. An extreme form of length-biased sampling, over- diagnosis, is the detection of tumours that are so slow-growing that they would not have caused any harm during the lifespan of the patient. Without screening, the patient would have gone on to die of a competing cause of death without ever being la- belled as a cancer patient. Empirical evidence for detection-related over- diagnosis has been shown for a wide variety of cancers, including mela- noma and cancers of the thyroid, prostate, kidney, and breast [8,17]. Since cancer is primarily a disease of ageing, during a period of life in which competing causes of death in- crease in incidence, cancer screen- ing is particularly prone to overdiag- nosis. However, screening has even been shown to produce overdiagno- sis in the case of neuroblastoma, a disease of infancy (reviewed in [18]). The effect is not only to increase the survival rate but also to increase the cure rate. After all, it is easiest to “cure” patients who did not need to be treated in the first place.
All of these biases inflate surviv- al rates in association with screen- ing independent of the actual effect of a screening test on mortality. In fact, it has been shown empirically that increasing 5-year survival rates in the USA from 1950 to 1995 and changes in mortality rates for the same cancers had little or no rela- tionship (Pearson r = 0.00) [19]. In other words, survival is an unreli- able measure of success, whether judged at the individual or popula- tion level, when a screening test is involved.
Since survival time after diagno- sis, rather than mortality, is the only observation that a physician can di- rectly make, even astute clinicians
such as Drs Childe and Bloodgood may be misled by their own care- ful observations and experience. In fact, surveys show that most primary care physicians erroneously inter- pret improved survival in associa- tion with screening as evidence that screening saves lives [20].
Likewise, patients who undergo screening may interpret observa- tions from their own experience as evidence of benefit, whether or not the screening test is effective. A negative test provides reassurance. A false-positive test brings immense relief when cancer is ruled out. A true-positive test triggers gratitude towards the physician for order- ing the test and detecting the can- cer early (even if it is nothing more than a case of overdiagnosis). Even severe side-effects of therapy are judged to be worthwhile by both pa- tient and physician. If the patient still goes on to die of cancer, the physi- cian and family alike can feel reas- sured that everything possible was done. In essence, there is little or no negative feedback [21].
Little wonder that so much scep- ticism and vitriol is aimed at authors of research papers or media reports that question the net benefits of some screening tests. The reports seem to run counter to personal experience on the part of both the public and physicians. The result is cognitive dissonance, which breeds anxiety, a progenitor of anger.
Are there resolutions to the controversy? Potential resolutions map to the genesis of the controversies de- scribed above: (1) imprecise un- derstanding of the biology of many screen-detected cancers, (2) in- sufficient training to recognize the powerful biases that affect interpre- tation of personal experience, and (3) inadequate nuance and framing of messages about screening.
Towards a better understanding of biology Traditional staging and prognostic systems such as tumour–node–
metastasis (TNM) classification and histological grading are crude, providing insufficient guidance in discerning overdiagnosis at the individual level. Therefore, the pa- tient and physician may feel driven to treat all or most screen-detected cancers, even if that means unnec- essary treatment or overtreatment in many cases. Some of the con- troversies would be calmed if there were more reliable ways to dis- tinguish screen-detected cancers that are aggressive from those that would not cause harm, reserving treatment for the former. This ap- proach is implicit in the use of active surveillance for screen-detected prostate cancer and neuroblastoma [22–24]. However, prediction at the individual level for most cancers is too crude for comfort.
One strategy would be to use emerging molecular techniques to characterize screen-detected le- sions. This research strategy is un- der way within the Early Detection Research Network of the United States National Cancer Institute (http://edrn.nci.nih.gov/). An exam- ple of a prospective design would be to characterize tumours from patients with screen-detected pros- tate cancer who are undergoing se- rial biopsies as part of active sur- veillance. Cross-sectional designs may also provide initial insights if tumour specimens are accurately annotated with respect to method of diagnosis: screen-detected ver- sus symptomatic interval cancers in people being actively screened. Screen-detected cancers are, on average, less aggressive tumours compared with symptomatic inter- val cancers, and their molecular patterns can be compared. Finally, rapid autopsy studies to find sub- clinical cancers in people who died of causes unrelated to cancer could be used to characterize the reser- voir of cancers that could be overdi- agnosed by screening. With better molecular characterization of over- diagnosed cancers, any benefits of screening could be preserved while harms would be minimized.
524
Improvements in professional training Formal training that covers the strong screening-related biases that distort clinical observation (and non-randomized screening studies) could better prepare health profes- sionals to judge the value of screen- ing tests and to evaluate published studies, damping down the cogni- tive dissonance created by personal experience. Such training is best begun early in medical training, be- fore specialty-associated heuristics become imprinted. Additional pub- lic health and epidemiology training during medical school could alleviate some of the “numerator/denomina- tor” tension.
Better framing of screening messages The strength of screening mes- sages should match the strength
of evidence. We need to get be- yond cancer screening campaigns that oversimplify the complexities of screening [3]. Cancer screening is often a closer call than can be de- scribed in sound bites. Informed de- cision-making, rather than persua- sion, should be the goal. Likewise, the benefits of screening are most frequently described in relative terms. Empirical evidence, includ- ing randomized trials, has shown that the public develops a better grasp of the efficacy and harms of an intervention when they are pre- sented in terms of absolute rates or frequencies [25,26]. The statement that a screening test for lung cancer has been shown to decrease lung cancer mortality by 20% is roughly equivalent to the statement that if 1000 heavy smokers and former smokers were to be screened annu- ally for three screens, about 4 lung
cancer deaths would be avoided over the next 6–7 years (the results of the National Lung Screening Trial comparing low-dose helical computed tomography [CT] with chest radiograph) [27]. However, the “feel” of the two methods of ex- pressing the results is different, and the latter method is generally bet- ter incorporated into individual de- cision-making. An example of this method of presenting the results of the trial can be found at http://www. cancer.gov/newscenter/qa /20 02 / NLSTstudyGuidePatientsPhysicians.
Opinions expressed in this manuscript are those of the author and do not neces- sarily represent official positions of the United States federal government or of the Department of Health and Human Services.
Controversies in cancer screening and their resolution: a view from the United States “battleground” 525
References
1. Childe CP (1907). The Control of a Scourge, or How Cancer Is Curable. New York: Dutton.
2. [Anonymous]. Cure for cancer in prompt action; Dr. Bloodgood of Johns Hopkins declares elimination almost sure in early stage. New York Times, 8 June 1924, p. 25.
3. Woloshin S, Schwartz LM, Black WC, Kramer BS (2012). Cancer screening cam- paigns–getting past uninformative per- suasion. N Engl J Med, 367:1677–1679. http://dx.doi.org/10.1056/NEJMp1209407 PMID:23113476
4. Lerner BH (2001). Seek and ye shall find: mammography praised and scorned. In: The Breast Cancer Wars: Hope, Fear, and the Pursuit of a Cure in Twentieth- Century America. New York: Oxford University Press, pp. 196–222.
5. Brawley OW, Goldberg P (2011). From the health fair. In: How We Do Harm: A Doctor Breaks Ranks about Being Sick in America. New York: St. Martin’s Press, pp. 215–224.
6. Bellizzi KM, Breslau ES, Burness A, Waldron W (2011). Prevalence of cancer screening in older, racially diverse adults: still screening after all these years. Arch Intern Med, 171:2031–2037. http://dx.doi. o r g /10 .10 01/a r c h i n t e r n m e d . 2 011. 5 70 PMID:22158573
7. Sima CS, Panageas KS, Schrag D (2010). Cancer screening among patients with advanced cancer. JAMA, 304:1584–1591. http://dx.doi.org/10.1001/jama.2010.1449 PMID:20940384
8. Bleyer A, Welch HG (2012). Effect of three decades of screening mammog- raphy on breast-cancer incidence. N Engl J Med, 367:1998–2005. http:// d x . d o i . o r g /10 .10 5 6 / N E J M o a12 0 6 8 0 9 PMID:23171096
9. Morin M (2012). To screen or not to screen. Los Angeles Times, 21 November 2012.
10. Ferguson JH (1999). Curative and popu- lation medicine: bridging the great di- vide. Neuroepidemiology, 18:111–119. h t t p: //d x . d o i . o r g /10 .115 9 / 0 0 0 0 2 6 2 0 2 PMID:10202265
11. Mamede S, van Gog T, van den Berge K et al. (2010). Effect of availability bias and reflective reasoning on diagnos- tic accuracy among internal medicine residents. JAMA, 304:1198–1203. http:// d x . d o i . o r g / 1 0 .1 0 0 1 / j a m a . 2 0 1 0 .12 7 6 PMID:20841533
12. Hoffman RM, Barry MJ, Roberts RG, Sox HC (2012). Reconciling primary care and specialist perspectives on prostate can- cer screening. Ann Fam Med, 10:568– 571. http://dx.doi.org/10.1370/afm.1399 PMID:23149535
13. Kramer BS, Croswell JM (2009). Cancer screening: the clash of science and in- tuition. Annu Rev Med, 60:125–137. h t t p : / / d x . d o i . o r g / 1 0 .11 4 6 / a n n u r e v . med.60.101107.134802 PMID:18803476
14. Croswell JM, Ransohoff DF, Kramer BS (2010). Principles of cancer screen- ing: lessons from history and study de- sign issues. Semin Oncol, 37:202–215. h t t p : / / d x . d o i . o r g / 1 0 .1 0 5 3 / j . s e m i n o n col.2010.05.006 PMID:20709205
15. Pinsky PF, Miller A, Kramer BS et al. (2007). Evidence of a healthy volunteer effect in the prostate, lung, colorectal, and ovarian cancer screening trial. Am J Epidemiol, 165:874–881. http://dx.doi. org/10.1093/aje/kwk075 PMID:17244633
16. Domingo L, Blanch J, Servitja S et al. (2013). Aggressiveness features and outcomes of true interval cancers: comparison between screen-detected and symptom-detected cancers. Eur J Cancer Prev, 22:21–28. http://dx.doi. o r g /10.10 97/C E J. 0 b 013 e 3 2 8 3 5 4 d3 24 PMID:22584215
17. Welch HG, Black WC (2010). Overdiagnosis in cancer. J Natl Cancer Inst, 102:605– 613. http://dx.doi.org/10.1093/jnci/djq099 PMID:20413742
18 . N at i o n a l C a n c e r I n st i t ute. PD Q Neuroblastoma Screening. Bethesda, MD: National Cancer Institute. Date last modi- fied: 23 July 2010. Available at http://www. cancer.gov/cancertopics/pdq/screening/ neuroblastoma/HealthProfessional.
19. Welch HG, Schwartz LM, Woloshin S (2000). Are increasing 5-year sur- vival rates evidence of success against cancer? JAMA, 283:2975–2978. http:// dx.doi.o rg /10.10 01/ jama. 28 3 . 2 2 . 2 975 PMID:10865276
20. Wegwarth O, Schwartz LM, Woloshin S et al. (2012). Do physicians understand can- cer screening statistics? A national survey of primary care physicians in the United States. Ann Intern Med, 156:340–349. http://dx.doi.org/10.7326/0003-4819-156- 5-201203060-00005 PMID:22393129
21. Ransohoff DF, McNaughton Collins M, Fowler FJ (2002). Why is prostate cancer screening so common when the evi- dence is so uncertain? A system without negative feedback. Am J Med, 113:663– 6 67. ht t p: //dx.do i.o r g /10.1016 /S 0 0 0 2- 9343(02)01235-4 PMID:12505117
22. Ganz PA, Barry JM, Burke W et al. (2011). National Institutes of Health State-of-the- Science Conference Statement: Role of Active Surveillance in the Management of Men with Localized Prostate Cancer. NIH Consensus and State-of-the-Science Statements, Vol. 28, No. 1, December 5–7, 2011. Bethesda, MD: National Institutes of Health.
23. Nishihira H, Toyoda Y, Tanaka Y et al. (2000). Natural course of neuroblas- toma detected by mass screening: a 5-year prospective study at a single in- stitution. J Clin Oncol, 18:3012–3017. PMID:10944135
24. Yoneda A, Oue T, Imura K et al. (2001). Observation of untreated patients with neu- roblastoma detected by mass screening: a “wait and see” pilot study. Med Pediatr Oncol, 36:160–162. http://dx.doi.org/10.1002/1096- 911X(20010101)36:1<160::AID-MPO1039> 3.0.CO;2-G PMID:11464874
25. Fagerlin A, Zikmund-Fisher BJ, Ubel PA (2011). Helping patients decide: ten steps to better risk communication. J Natl Cancer Inst, 103:1436–1443. http://dx.doi. org/10.1093/jnci/djr318 PMID:21931068
26. Woloshin S, Schwartz LM (2011). Communicating data about the benefits and harms of treatment: a randomized trial. Ann Intern Med, 155:87–96. http:// dx.doi.org /10.10 5 9 / 0 0 03 - 4 819 -15 5 -2- 201107190-00004 PMID:21768582
27. Aberle DR, Adams AM, Berg CD et al.; National Lung Screening Trial Research Team (2011). Reduced lung-cancer mor- tality with low-dose computed tomograph- ic screening. N Engl J Med, 365:395–409. http://dx.doi.org/10.1056/NEJMoa1102873 PMID:21714641
526
Cancer control
6
Earlier sections of this Report have addressed the worldwide burden of cancer, its causation and preven- tion, and current understanding of cancer biology as un- derpinning new approaches to both therapy and preven- tion, with these various parameters explored in relation to each of the most frequent tumour types. Knowledge about all of these matters, specifically including inter- ventions to reduce cancer incidence and mortality, is
primarily dependent on research; research findings have determined the content of all previous chapters. Section 6 is not primarily a specification of research findings but more a description of the manner and extent to which knowledge of cancer control may be applied at the na- tional level and a consideration of the resources relevant to achieving this end. The focus shifts from the prospect of increased understanding to presently coordinated action predicated on methodologies already known to reduce the burden of cancer. Differences between coun- tries are evident, sometimes usefully expressed in rela- tion to national prosperity, but often extending to other parameters. Development and implementation of an ad- equately resourced national cancer control plan is now recognized as a fundamental element within the broad scope of population health and clinical services activ- ity. International collaboration provides an opportunity to minimize unnecessary evaluation and to optimize im- plementation for the benefit of national, or sometimes local, populations. In parallel with the implementation of cancer control measures, infrastructure for continued, locally relevant, implementation research may be adopt- ed and managed, thereby laying the foundation for even more effective cancer control measures consequent to such investigations.
528
Cancer control in Africa: options for a vulnerable continent Isaac F. Adewole
Cancer remains a leading noncom- municable disease in Africa, and it is also emerging as a great burden when compared with infections that are ravaging the continent. The tri- ad of ignorance, poverty, and poor health-seeking behaviour makes Africa vulnerable to the cancer bur- den in both male and female, and young and adult populations [1]. Of the 7.6 million cancer deaths that occur worldwide, 4.8 million occur in developing countries, translating to about 21 000 cancer deaths per day, and Africa shares the highest proportion globally. The rapidly in- creasing cancer burden is attribut- able to the transitional demographic profile of several countries in Africa, with increasing proportions of older people in the population and a shift towards lifestyles typical of indus- trialized countries. Only 17 African countries had operational cancer control policies in 2010, and virtually no African country presently imple- ments a well-funded national cancer control plan. This clearly suggests a lack of political will and commitment across the continent [2]. Addressing this is a continental imperative.
First, each country should pro- duce a blueprint for national cancer control using a framework that is socially and culturally sensitive.
Second, creating awareness about cancer as a complex group of diseas- es that often have identifiable triggers should be pursued in a systematic manner. This will involve population- wide health promotion strategies fo- cused on diet and exercise, lifestyle modifications, sexuality/family life education programmes, campaigns against cultural norms/practices, and other behavioural change commu- nication strategies that are linkable to cancer risk. In addition, legislative support that promotes healthy living within the household and in commu- nities is imperative. These initiatives are expected to foster a strong pri- mary prevention strategy in a con- tinent where about 33% of cancers are infection-related [3].
Credible evidence abounds that some cancers are preventable with vaccination, and African countries should actively participate in cur- rent global efforts using practical approaches that are feasible in their countries, without an over- dependence on donor support. Institutionalization of these vac- cines into the already existing and successful national immunization programmes provides a true oppor- tunity in the continent [4].
Screening for premalignant and early diseases is another control
strategy that has yet to be entrenched in Africa. Priorities should be the es- tablishment of cancer registries as well as centres of excellence, and the training of a critical mass of experts to offer multidisciplinary team care for cancer patients in Africa, through collaboration with cancer centres offering cutting-edge services, pro- fessional organizations, and pharma- ceutical companies. Tax incentives could be offered for interested foun- dations, multinational companies, and individuals that are ready to in- vest in cancer control in Africa.
Cancer control in Africa is fea- sible, but the focus should be on a control plan that is realistic, sustain- able, equitable, and part of a strong health-care system.
References 1. Morhason-Bello IO et al. (2013). Lancet
Oncol, 14:e142–e151. http://dx.doi.org/ 10.1016/S1470-2045(12)70482-5 PMID: 23561745
2. Stefan DC et al. (2013). Lancet Oncol, 14:e189–e195. http://dx.doi.org/10.1016/ S1470-2045(13)70100-1 PMID:23561751
3. de Martel C et al. (2012). Lancet Oncol, 13:607–615. http://dx.doi.org/10.1016/S14 70-2045(12)70137-7 PMID:22575588
4. Sylla BS, Wild CP (2012). Int J Cancer, 130:245–250. http://dx.doi.org/10.1002/ ijc.26333 PMID:21796634
Chapter 6.1 • National cancer control plans 529 6 C
O N
T R
O L
C H
A P
T E
R 6
.1
Summary • National cancer control plans
detail the strategy to address the population burden of cancer through interventions to reduce incidence, mortality, and mor- bidity and enhance the quality of life of those at risk of, or ex- periencing, cancer.
• The integrity of these plans is based on current and accurate determination of burden, real- istic targets for improvement, and continuous surveillance to document performance and outcomes.
• National cancer control plans define: the purpose of planned activity, the content of interven- tions to reduce the cancer bur- den, the context within which ac- tivities are adapted to prevailing circumstances, the relationships required between stakeholders, and the resources (internal and external) to execute the plan.
• Adoption and implementation of national cancer control plans has been particular to high- income countries, but progress is now occurring in middle- and low-income countries as strate- gies appropriate to circumstanc- es and resources are identified.
6.1 6 CONTROL
National cancer control plans Simon B. Sutcliffe Raul Hernando Murillo Moreno
(reviewer) Edward L. Trimble (reviewer)
• Resistance, barriers to imple- mentation, and susceptibil- ity to unplanned modification or revised priorities are an im- mediate and ever-present real- ity. Determination, commitment, resolve, and collaboration are mandatory requirements to real- ize the future gains of population- based interventions to control cancer and other noncommuni- cable diseases.
A national cancer control plan is a public health programme designed to reduce the number of new cancer cases and cancer-related deaths and to improve the quality of life of cancer patients, through systematic and equitable implementation of evidence-based strategies for pre- vention, early detection, diagnosis, treatment, and palliation, making the best use of available resources [1].
The population burden of cancer is considerable, with 32.5 million people globally living with cancer [2] and 169.3 million years of healthy life (disability-adjusted life years) lost per year due to cancer [3]. The global burden will only continue to increase in the coming decades due to popu- lation growth and ageing. In 2012, there were an estimated 14.1 million new cases of cancer and 8.2 million cancer-related deaths worldwide [2].
Deaths due to cancer are projected to increase to 13 million in 2030. Although incidence rates are lower in lower-resource countries, mortality rates are higher; patterns of disease distribution and survival reflect vary- ing levels of socioeconomic develop- ment [2,4,5]. Mortality-to-incidence ratios for cancer vary from less than 0.50 to more than 0.90 across world regions (see Chapter 1.1). This variation reflects less a lack of knowledge of what should be done to control cancer than the level of commitment to implementing ef- fective cancer control interventions population-wide.
The cost of cancer is substantial. In 2009, new cancer cases globally were estimated to cost US$ 286 bil- lion, both in direct – i.e. medical – costs and in indirect costs, prin- cipally including lost productivity. Approximately 94% of these costs were incurred in high-resource countries; only 6% were incurred in upper-middle, lower-middle, and low-income countries [6]. Other estimates of the economic cost of cancer are far higher (US$ 1.16 tril- lion) when taking account of costs of prevention and treatment, and the annual economic cost of disability- adjusted life years (see Chapter 6.7). As the burden of cancer grows, it will unequally and inequitably affect
530
those with the least ability and capa- bility to respond.
To address this growing burden of cancer, population-based cancer control must be recognized as far more than a medical response to cancer. The cancer burden is driven by a complex interaction of changing factors, including social determi- nants such as poverty, education, the built environment, gender equity, and so on, and population health, in- cluding primary prevention/risk fac- tor control and early detection and management of disease and illness. Accordingly, improving cancer con- trol at a population level is a social imperative, which requires a societal response that may be challenging to prevailing social, business, and economic interests.
Addressing the increasing bur- den of cancer requires a country or population cancer control plan. The WHO strategic definition of a na- tional cancer control plan (NCCP), as specified above, is predicated on evidence-based interventions that need to be implemented for
prevention, early detection, diagno- sis, treatment, and palliation. This definition emphasizes the scientific and medical content of a plan, which is the focus of much of the discourse around NCCPs. The success of the plan, however, is dependent not only on its content – what needs to be done – but also on its context, which includes: cultural and circum- stance-specific factors influencing implementation; the key actors and stakeholders, and their responsibili- ties and accountabilities; available resources, including human, tech- nology, facilities, and financing per- spectives that influence how the plan will be operationalized; and evalu- ation of the plan and its impact on cancer control outcomes and the achievement of defined targets. If these elements are not addressed in the NCCP, implementation and sus- tainability will be difficult to achieve and the impact of the plan on cancer control and the health of the popula- tion will be jeopardized regardless of the effectiveness of the interventions outlined within it.
The fundamental role of data The first step in establishing an NCCP is to understand the scope of the region-specific cancer burden, as patterns of cancer vary by geog- raphy. This variation can arise for many reasons, including heteroge- neities in etiological risk factors such as tobacco use and chronic infec- tion; geography, as exemplified by climate; social determinants, such as poverty; and health systems, specifi- cally including access to diagnostic, treatment, and care services. These heterogeneities can contribute to variation in pathology and stage, survival, functionality, and quality of life [7]. Given this wide variation, no single NCCP will satisfy the needs of every country or population. Instead, NCCPs must be adapted to national and regional needs and capabilities.
Population data are the founda- tion for understanding the burden and pattern of cancer. These data can also be used to synthesize and prioritize planned interventions, establish system capacity require- ments for care, evaluate population- based cancer control activities, and justify continued investment of re- sources according to performance and outcomes of plans. Projection
CANCER CONTROL
PROGRAMME
Early detection
Treatment
Prevention Palliative
care
Fig. 6.1.1. National cancer control programme: a systemic and comprehensive approach.
Fig. 6.1.2. A WHO poster for World No Tobacco Day 2010 focuses on the mar- keting of tobacco to women.
Chapter 6.1 • National cancer control plans 531 6 C
O N
T R
O L
C H
A P
T E
R 6
.1
of data can also define future bur- den, needs, capacity, impact, and required investment.
Cancer registries are the primary source of data for NCCPs. Registries present the population burden of cancer and can do so by time, key population attributes (including age, gender, and recognition of paedi- atric and adolescent/young adult cases), cancer site/type, impact of interventions as affecting stage dis- tribution, 5-year survival, disability, the presence or absence of health insurance [8], and peer comparison with reference, for example, to geo- graphical, political, economic, eth- nic, and heritage status [2,7,9]. In the United Kingdom and the European Union, comparative presentation of data has provided the stimulus for NCCP development, renewal, and implementation. Indeed, increasing recognition of the global variation in incidence, mortality, and 5-year survival, and reasons for this varia- tion – which include access to care, diagnostic and treatment services, and investment in health resources – have provided an important stimulus for viewing NCCPs as a means of improving cancer burden, mitigating variation, and addressing disparities.
However, as fundamental as cancer registries are for cancer con- trol planning, many lower-resource countries are challenged by having neither registries nor a systematic ability to collect data. These chal- lenges result from a lack of medi- cal facilities, low cancer awareness, poor follow-up, poorly maintained records, lack of trained staff, and/ or lack of fiscal, medical, and po- litical resolve. Potential solutions include establishing the culture of evidence, supported by data, be- tween clinicians and data registrars; educational meetings to raise aware- ness; household visits; collection of data on admissions, procedures, and mortality from urban reference centres; and a focus on good data collection from a few, more estab- lished centres rather than an entire population survey. Ultimately, ef- forts to establish reliable population data for programme planning need
to consider comprehensiveness across the disease control spectrum, completeness, accuracy, timeliness, and coverage. Notable examples of approaches by lower-resource countries exist in South and Central America and India.
Once established, registries should not be viewed as static re- positories of data. They should be enhanced over time to incorporate more detailed data on diagnosis and treatment. Along with projec- tion and modelling methodologies, these continually updated data will help refine needs, capacity, resourc- es, programme evaluation, and in- vestments, thereby allowing health systems to maintain optimal cancer control outcomes.
The coverage and quality of registry data can vary considerably worldwide. Estimates of the propor- tion of the population covered by cancer registries range from more than 80% in North America, Europe, and Australia to approximately 30% in the Russian Federation, less than 10% in South America and South- East Asia, and only a few percent in Africa. Regional registries also vary in their comprehensiveness, completeness of information cap- ture, timeliness, accuracy and con- sistency of terminology, and use of international classifications and recording standards. Irrespective of their coverage and quality, registries are nevertheless useful for informing and refining NCCPs. Thus, even in regions without functioning regis- tries, at least minimal cancer-rele- vant health surveillance should be developed concurrent with NCCPs to rationalize activities, investments, and performance of cancer control plans [10].
The content of NCCPs An NCCP is a strategic plan en- compassing the spectrum of cancer control measures and is evidence- based, scientifically accurate, and current. Establishing the content of this plan is relatively straightforward as it is based on best practices, the scientific and clinical literature, and
Box 6.1.1. Guiding principles for developing a national cancer control plan.
1. Comprehensiveness: The plan should address all members of the population, with attention to addressing inequities and disparities.
2. Scope: The plan should address cancer control from the perspec- tives of human development, risk factor control, and health and dis- ease management.
3. Evidence base: The plan should be based on evidence or best practices and should incorporate indicators and metrics of perfor- mance, output, and outcomes.
4. Standards of practice: The cancer control plan should take into account measures to define standards and ensure consis- tent application, such as access, timeliness, quality of care, and safety.
5. Minimization of inappropriate variation of practice and fac- tors influencing compliance with therapy: The plan should use guidelines or best practices and measures to support adher- ence and compliance.
6. Integration and continuity: The plan should strive for conti- nuity across states of health and illness, and across home, com- munity, and tertiary or specialist environments.
7. Inclusiveness: The cancer con- trol plan should be developed with input and support from the public, patients, providers, poli- cy-makers, and payers.
8. Governance: The plan should align with population health priorities and outline respon- sibilities, accountabilities, and reporting.
9. Sustainability: The plan should be developed with the inten- tion of being self-sufficient and sustainable.
532
training and planning manuals [11– 13]. Published plans are accessible [14–16]. As NCCPs are about con- trol of disease within the context of a country and the health of its popula- tion, each country must adapt the content of its plan to its population and circumstances. A set of core principles have been recognized as guiding acceptance and imple- mentation (Box 6.1.1). The plan must describe how beneficial change in cancer control will occur; address purpose, content, context, relation- ships, and resources; present a busi- ness plan to determine how resource inputs will be acquired, deployed, and accounted for; and include an operational plan to determine how implementation is going to roll out
over time in a coordinated, inte- grated manner that leads to perfor- mance and achievement of results (Box 6.1.2).
Surveys of WHO Member States indicate that elements of functional cancer control plans are present in 35% of low-income countries, 35 – 60% of countries in the lower-middle and upper-middle income groups, and 75% of high- income countries. However, sub- stantial variation exists within re- source settings with respect to the specific components of the plans (see Chapter 6.2).
The context of NCCPs Critical to the success of the NCCP is its adaptation to its country or
population context. Contextual fac- tors include resources and priorities.
For any country, the level of resources available in the health system is the major determinant for decision-making in health plan- ning and its limitations. Resources include human resources, and specifically qualified personnel; technology, which includes infor- mation technology; drugs, and diag- nostic and therapeutic equipment; facilities, with reference, for exam- ple, to access to transportation and affordable in-patient and ambula- tory accommodation; and financial resources, including current and pro- jected annual growth. As interven- tions move from a population level to a person, tissue, cell, and genome,
Box 6.1.2. Elements to be included in a national cancer control plan.
1. Why a national cancer control plan is necessary: • What is the burden of cancer?
• What is the capability and capacity for control?
• What is the imperative for change?
• How important is cancer relative to other health challenges?
2. What interventions are required: • What interventions for cancer are most important: risk factor control, early detection, diagnosis, treatment,
and care?
• How will these interventions be prioritized in relation to other elements of cancer and noncommunicable disease control?
3. How interventions will be implemented, and how the process of implementation will be monitored and evaluated:
• How will human, technology, facilities, and organizational resources be aligned to the implementation?
4. Who will be involved: • How will collaboration be fostered and managed?
• How will roles and resources be shared?
• How will responsibility and accountability be established?
• How will “buy-in”, progress, and performance be communicated and reported?
5. When activities will be undertaken according to priority, what is the capability and the resource avail- ability (the operational plan).
6. How the national cancer control plan and its implementation will be financed (the business plan): • What are the contributions of foreign aid, in-kind services, and local financing?
• How are self-sufficiency and sustainability being addressed?
7. How the plan will be evaluated: • How will outputs and outcomes be measured (system performance), linked to a changing burden of cancer,
and aligned with resource allocation to determine value?
Chapter 6.1 • National cancer control plans 533 6 C
O N
T R
O L
C H
A P
T E
R 6
.1
the requirements for level of qualifi- cations of personnel, technology, in- frastructure, time, and cost increase, while population access commonly decreases. Interventions directed at behaviour change at a community or individual level, while sometimes difficult to achieve, require modest technical support, whereas interven- tions involving clinical care – such as surgery and provision of radiation and drug-based therapy – require a proficient level of specialist skill and technological capacity.
Recognizing the resource chal- lenges of lower-resource countries, WHO has presented a core set of “best buy” lower-cost intervention strategies for four noncommunica- ble diseases: cardiovascular dis- ease, chronic respiratory diseases, diabetes, and cancer [17]. These strategies address factors common to more than one such disease, such as tobacco use, alcohol con- sumption, unhealthy diet, and lack of physical activity. Interventions in- clude tax increases and legislation, as illustrated by measures relating
to smoke-free environments, limited retail access, and advertising bans; hepatitis B vaccination and cervi- cal cancer screening programmes; and recommendations for reducing salt intake, substituting trans fats, and using aspirin for myocardial ischaemia.
The annual cost of implement- ing these interventions for all people in lower-middle-income countries is estimated to be US$ 11.4 billion – an investment of US$ 1 per person in low-income countries to US$ 3 per person in upper-middle-income countries. This contrasts with a cur- rent “business as usual” scenario whereby economic losses of US$ 500 billion per year are incurred for the four noncommunicable diseases – a loss equivalent to 4% of annual output of low-income countries and lower-middle-income countries, or US$ 25–50 per person per year. Initial investments, which may be quite modest for many aspects of population health, have a personal, family, community, and economic return on investment based on in- creased longevity, productivity, so- cietal contribution, and avoidance of health-care costs.
Most of the above-mentioned “best buy” interventions relate to prevention, intuitively considered to be more cost-effective than interven- tions for early or established dis- ease. A further distinction of disease prevention interventions relates to whether they are environmental (involving measures that affect the whole population) or personal (di- rected towards changing individual behaviour) and whether they are clinical (delivered in a health-care facility) or non-clinical (delivered outside of the health-care setting). Environmental interventions are generally the most cost-effective as they affect a large proportion of the population. Typically such measures are achieved through policy and reg- ulation – examples are smoke-free laws or food regulations addressing fortification or control of unhealthy food components such as trans fats. In contrast, interventions directed at lifestyle or individual choice, or
delivered in clinical settings, are less cost-effective, although not mutually exclusive of environmental interven- tions [18].
Even though “best buy” inter- ventions are clearly cost-effective, economics is not the sole criterion for NCCP development within limited resources. Other factors, including societal and circumstance-specific factors [19], alternative models of service delivery [20], and priorities within the disease, health, and other sectors will determine ultimate de- cisions. In low-resource settings, NCCPs may not be feasible as stand-alone plans. The principles underlying NCCPs, however, are also broadly applicable to other non- communicable diseases, as noted above, even though treatment op- tions may be disease-specific [21].
Enhancing NCCPs NCCPs commonly define strategies to control cancer. Cancer control measures, however, will be achieved only if the NCCP is implemented within the health system through the
Fig. 6.1.3. Poster from the National Research and Safety Institute (INRS), France, advocating the wearing of protec- tive masks by workers exposed to asbes- tos: “When working on construction sites, I never wore a protective mask against asbestos. Now, I wear a mask every day. With asbestos, don’t take chances. Protect yourself!”
Fig. 6.1.4. Poster from the Iowa De- partment of Public Health, USA, advo- cating colorectal cancer screening for Vietnamese Americans: “Life is good, and I’m not letting anything take me away from my family. Asian Americans are at higher risk of colorectal cancer. Treatment is easy if it is detected early through testing. If you are 50 or older, get tested today! ”
534
“push” of science and medicine and the “pull” of patients and the public (sociopolitical imperative) [22].
Such an approach involves social mobilization, causing the challenge to be embraced and championed by key constituencies influencing so- cial and political change. There are roles for advocate and stakeholder participation, together with the en- gagement of relevant networks and coalitions.
Collaboration across networks and coalitions will increase their impact and may extend from infor- mation exchange to agreed coor- dination and cooperation. The lat- ter requires trust, mutual respect, a sensitivity to issues of authority, and understandings about roles, responsibilities, and accountabili- ties. Examples of such networks include national organizations, such as the Canadian Partnership Against Cancer; regional entities for cancer control, such as the African Organisation for Research and Training in Cancer and the Latin American Network of National Cancer Institutes; and international entities, such as the International Network for Cancer Treatment and Research and the International Cancer Control Congresses. Also key to effective collaboration are well-defined and secure financial commitments through, for example, a business plan, a detailed imple- mentation plan, and an understand- ing that the NCCP is an integral component of the population health system.
The value of NCCPs Proof of effectiveness of cancer control measures is hindered by the long time gap between implemen- tation of interventions and changes in incidence and mortality. Other limitations stem from the paucity of data linking intervention and effect directly, the continuous introduc- tion of new technologies and thera- peutic advances, and the inability to undertake a controlled study of NCCP versus no-NCCP within a single population over a given time.
Thus, causality can be approximated and can be discussed; it cannot be assumed [23]. For NCCPs to dem- onstrate value, they must be imple- mented and audited.
Although direct evidence of ef- fectiveness is not available, sup- portive evidence for the benefit of NCCPs exists. This evidence de- rives from several sources.
The first is programmes address- ing specific elements of cancer con- trol. Changes in mortality are evident after the implementation of tobacco control programmes (in Australia, Canada, and Brazil), organized pre- vention through vaccination (against human papillomavirus in Costa Rica and against hepatitis B virus in Taiwan, China), sun protection cam- paigns and screening for melanoma (in Australia), and organized early detection programmes for cervical cancer (in Colombia, Ecuador, Chile, the United Kingdom, and Canada).
The second is control of multi- ple risk factors for noncommunicable diseases, including cancer. In North Karelia, Finland, substantial reduc- tions in cardiovascular disease and cancer – and particularly lung can- cer – mortality have been achieved through control of targeted risk
factors [24]. In the USA, decreases in coronary heart disease mortal- ity over the past four decades have been achieved through improve- ments in risk factor control, exem- plified by smoking, hypercholesterol- aemia, and hypertension, and, more recently, through the introduction and adoption of treatment innova- tions [25].
The third is treatment improve- ments. Increased clinical trial en- rolment and rigorous definition of benefit and measures to encour- age adoption of evidence-based care have led to reduced paediat- ric cancer mortality in high-income countries and in lower-resource countries such as Argentina. The adoption of professional practice standards to reduce inconsisten- cies, deficiencies, and variability in clinical practice – including, for ex- ample, surgery, pathology assess- ment, and therapy allocation – has led to decreased mortality for resect- able colorectal cancer. Compliance with evidence-based/informed clini- cal practice guidelines has led to increased breast cancer survival, and implementation and application of evidence-based new therapies have led to reduced mortality for
Fig. 6.1.5. The “Nobody’s Immune to Breast Cancer” campaign of Associação da Luta Contra o Cancro (Association for the Fight Against Cancer), Mozambique: “When we talk about breast cancer, there are no women or superwomen. Everybody has to do a monthly self-examination. Fight with us against this enemy and, when in doubt, talk with your doctor.”
Chapter 6.1 • National cancer control plans 535 6 C
O N
T R
O L
C H
A P
T E
R 6
.1
B-cell non-Hodgkin lymphoma. The achievement of surrogate targets of improved health outcomes is illus- trated by access to care and treat- ment, screening rates, and quality and timeliness of diagnostic services [26].
Finally, comparisons have been made of trends in mortality rates in comparable populations with and without NCCPs, as indicated, for ex- ample, by survival trends in England versus Wales up to 2007 [27].
Conclusions Cancer is a major and increasing public health problem and popula- tion burden for all countries. An
NCCP presents an opportunity to implement effective change at a population level on an ongoing, adaptable basis, using a variety of interventional methods that can be accommodated within the contex- tual, social, and financial means of even the poorest nations.
The challenge for NCCPs is less the derivation of content than the ability to apply and achieve sustainable change at a population level. If conditions are favourable (Box 6.1.3), with the support and engagement of all key actors and a clear and common vision for cancer control, the plan is more likely to suc- ceed in improving the health of the population.
NCCPs identify why action is nec- essary and what action is required. To change the population impact of cancer, however, it must be clear how NCCPs will be implemented, adopted, evaluated, and sustained. An NCCP, as defined by WHO, has the potential to change population incidence, suffering, and mortality from cancer through implementa- tion in a contextually aligned man- ner when supported by all relevant stakeholders in an inclusive and inter- disciplinary partnership, and financed and governed to achieve long-term, sustainable improvements in cancer control.
Box 6.1.3. Conditions underlying the probability of a successful national cancer control plan.
1. Political and professional consistency and resolve to address the population cancer burden. 2. Use of data and a commitment to support and maintain cancer registration and surveillance. 3. Clearly defined vision for cancer control, commonly agreed and supported. 4. Contextual relevance, defined priorities, achievable implementation, and an appropriate time frame to
achieve goals.
5. Defined, secure funding commitment. 6. Trust, mutual respect, and willingness to achieve commonly defined goals through collaboration by all key
actors.
7. Scalability, incorporation into the health system, self-sufficiency, and sustainability. 8. Applicability to, and coordination and cooperation with, other population disease control plans. 9. Sound governance, evaluation, communication, and ongoing adaptation to meet future needs.
536
References
1. WHO (2012). National Cancer Control Programmes: Planning. Geneva: WHO. Available at http://www.who.int/cancer/ nccp/planning/en/.
2. Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
3. Soerjomataram I, Lortet-Tieulent J, Parkin DM et al. (2012). Global burden of cancer in 2008: a systematic analysis of disability-ad- justed life years in 12 world regions. Lancet, 380:1840–1850. http://dx.doi.org/10.1016/ S0140-6736(12)60919-2 PMID:23079588
4. Bray F, Jemal A, Grey N et al. (2012). Global cancer transitions according to the Human Development Index (2008–2030): a population-based study. Lancet Oncol, 13:790–801. http://dx.doi.org/10.1016/S14 70-2045(12)70211-5 PMID:22658655
5. Bray F, Ren JS, Masuyer E, Ferlay J (2013). Global estimates of cancer prevalence for 27 sites in the adult population in 2008. Int J Cancer, 132:1133–1145. http://dx.doi. org/10.1002/ijc.27711 PMID:22752881
6. Economist Intelligence Unit (2009). Breakaway: The Global Burden of Cancer – Challenges and Opportunities. London: The Economist Group. Available at http://www. livestrong.org/pdfs/GlobalEconomicImpact.
7. Coleman MP, Quaresma M, Berrino F et al.; CONCORD Working Group (2008). Cancer survival in five continents: a worldwide pop- ulation-based study (CONCORD). Lancet Oncol, 9:730–756. http://dx.doi.org/10.1016/ S1470-2045(08)70179-7 PMID:18639491
8. McDavid K, Tucker TC, Sloggett A, Coleman MP (2003). Cancer survival in Kentucky and health insurance coverage. Arch Intern Med, 163:2135–2144. http:// dx.doi.org/10.1001/archinte.163.18.2135 PMID:14557210
9. Capocaccia R, Gavin A, Hakulinen T et al.; the EUROCARE Working Group (2009). Survival of cancer patients in Europe, 1995–2002: the Eurocare 4 study. Eur J Cancer, 45:901–1094. PMID:19217771
10. Global Initiative for Cancer Registry Development in Low- and Middle-Income Countries (2012). Available at http://gicr. iarc.fr.
11. Planning WHO (2006). Cancer Control: Knowledge into Action. WHO Guide for Effective Programmes. Module 1: Planning. Geneva: WHO. Available at http://www.who.int/cancer/publications/ cancer_control_planning/en/index.html.
12. Union for International Cancer Control (2012). Supporting National Cancer Control Planning: A Toolkit for Civil Society Organizations (CSOs). Available at http://www.uicc.org/advocacy/advocacy- resources/nccp-toolkit.
13. WHO, International Atomic Energy Agency (2011). National Cancer Control Programmes Core Capacity Self- Assessment Tool (NCCP core self-as- sessment tool). Geneva: WHO. Available at http://www.who.int/cancer/publications/ nccp_tool2011/en/.
14. France: le Plan Cancer 2009–2013. Available at http://www.plan-cancer.gouv.fr/ le-plan-cancer/presentation.html.
15. The National Health Service Cancer Plan: a plan for investment, a plan for reform. Available at http://www.dh.gov/ uk/en/publicationsandstatistics/publica tions/publicationspolicyandguidance/dh_ 4009609.
16. Uruguay: Programa Nacional de Control del Cáncer. Available at http://www.bvson cologia.org.uy/pdfs/destacados/prona ccan_2005-2010.pdf.
17. WHO (2011). From Burden to “Best Buys”: Reducing the Economic Impact of Non- Communicable Diseases in Low- and Middle-Income Countries. Geneva: WHO. Available at http://www.who.int/nmh/publi cations/best_buys_summary/en/.
18. Chokshi DA, Farley TA (2012). The cost- effectiveness of environmental approach- es to disease prevention. N Engl J Med, 367:295–297. http://dx.doi.org/10.1056/ NEJMp1206268 PMID:22830461
19. WHO (2008). Task Shifting: Global Recom- mendations and Guidelines. Available at http://www.who.int/healthsystems/TTR- TaskShifting.pdf.
20. Wools-Kaloustian K, Kimaiyo S (2006). Extending HIV care in resource-limited set- tings. Curr HIV/AIDS Rep, 3:182–186. http:// dx.doi.org/10.1007/s11904- 006 - 0014-1 PMID:17032578
21. United Nations (2011). Political Declaration of the High-Level Meeting of the General Assembly on the Prevention and Control of Non-communicable Diseases. New York: United Nations. Available at www. who.int/nmh/events/un_ncd_summit2011/ political_declaration_en.pdf.
22. Trubek LG, Oliver TR, Liang C-M, Mokrohisky M (2011). Improving cancer con- trol outcomes through strong networks and regulatory frameworks: lessons from the United States and European Union. J Health Care Law Policy, 14:119–151. Available at http://digitalcommons.law.umaryland. edu/jhclp/vol14/iss1/5.
23. Rochester P, Chapel T, Black B et al. (2005). The evaluation of comprehen- sive cancer control efforts: useful tech- niques and unique requirements. Cancer Causes Control, 16 Suppl 1:69–78. http:// dx.doi.org/10.1007/s10552- 005 - 0510- 4 PMID:16208576
24. Puska P (2002). Successful prevention of non-communicable diseases: 25 years experience with North Karelia project in Finland. Public Health Med, 4:5–7. Available at http://www.who.int/entity/chp/ media/en/north_karelia_successful_ncd_ prevention.pdf.
25. Ford ES, Capewell S (2011). Proportion of the decline in cardiovascular mortality disease due to prevention versus treat- ment: public health versus clinical care. Annu Rev Public Health, 32:5–22. http:// dx.doi.org/10.1146/annurev-publhealth- 031210-101211 PMID:21417752
26. Canadian Partnership Against Cancer (2011). The 2011 Cancer System Performance Report. Available at http:// www.cancer view.ca/idc/groups/public/ documents/webcontent /2011_system_ performance_rep.pdf.
27. Rachet B, Maringe C, Nur U et al. (2009). Population-based cancer survival trends in England and Wales up to 2007: an assess- ment of the NHS cancer plan for England. Lancet Oncol, 10:351–369. http://dx.doi. org /10.1016/S1470 -20 45(0 9)70 028 -2 PMID:19303813
Cancer control experiences 537 6 C
O N
T R
O L
Cancer control in Canada: challenges and strategies in a high-income country Heather Bryant and Lee Fairclough
In 2006, the federal government of Canada created the Canadian Partnership Against Cancer, a fed- erally funded nongovernmental organization charged with the im- plementation of the national cancer control strategy [1,2]. While the de- velopment of the strategy was built on many years of collaborative work, there were challenges that needed to be met and lessons to be learned in its implementation.
Many of the initial challenges concerned creating the appropriate measures needed to stimulate and assess the impact of such a strat- egy. Canada had the same excellent starting point that many high-income countries have: the presence of pop- ulation-based cancer registries with high-quality data throughout most of the country. However, there were gaps that had not been addressed. The first related to high-quality staging data, which was absent at a pan-Canadian level. Thus, one of the first initiatives of the Canadian Partnership Against Cancer was to work with the provincial registries and clinicians (particularly patholo- gists) to identify an approach to address this issue. The result was the National Staging Initiative, a large-scale programme designed
to catalyse coordinated efforts in all jurisdictions with investments in infrastructure and implementation of standards. The National Staging Initiative has successfully achieved the goal of making collaborative stage information available in a sus- tainable way for more than 90% of the four major cancers (lung, colo- rectal, breast, and prostate) starting in the 2010 coding year.
For many years, routine inter- provincial comparisons had been done on incidence, mortality, and, to a lesser extent, prevalence and survival. However, none of these measures were able to shine a light on what occurred between the two key time points: diagnosis and death. Some provinces had initiated reporting in this area where limited data existed, but the comparison of this across the country had not occurred. Thus, provincial cancer agencies and health departments were engaged to come to consen- sus on indicators that would be most useful as initial assessments of con- cordance of treatment with high-lev- el recommendations. Initially, only a few provinces were able to pro- duce the necessary data. However, over the past few years, the Cancer System Performance Report [3] has
evolved from having “proxy” indica- tors for many variables to a report with an ever-increasing insight into possible areas for improvement in each jurisdiction. These measures are placed with others to provide a view across the cancer control continuum, including population measures of risk factors, of screen- ing, and of patient experience. The report continues to develop, with the joint leadership of senior representa- tives of each provincial cancer agen- cy and the Canadian Partnership Against Cancer, and has been used by several provinces to target quality initiatives in their own jurisdictions.
References 1. Canadian Partnership Against Cancer
(2012). Sustaining Action Toward a Shared Vision: 2012–2017 Strategic Plan. Available at http://www.partnership againstcancer.ca/wp-content/uploads/ Sustaining-Action-Toward-a- Shared- Vision-Full-Document.pdf.
2. Fairclough L et al. (2012). Curr Oncol, 19:70–77. http://dx.doi.org/10.3747/co.19. 1019 PMID:22514493
3. Canadian Partnership Against Cancer (2012). The 2012 Cancer System Performance Report. Available at http:// www.cancerview.ca/systemperformance report.
538
Summary • The national capacity of a
country for cancer prevention and control may be assessed based on the WHO 2013 Noncommunicable Disease Country Capacity Survey by analysing cancer-specific data. Information is available from 176 countries.
• Although 85% of the participat- ing countries have cancer plans or policies, many of those plans are not operational. Only 38% of countries in Africa and 45% of low-income countries have operational plans with a budget to support implementation.
• Population-based cancer reg- istries were reported to exist in 72% of countries.
• From survey information re- ceived by WHO, major gaps in health care for cancer control were identified. Many countries, especially low-income coun- tries, lack a national cancer con- trol policy or plan. Where cancer policies have been developed, such policies are not always comprehensive.
• Because the cancer burden is high and is increasing globally, adoption and implementation,
6.2 6 CONTROL
Current global national cancer control capacity Andreas Ullrich Leanne Riley
Robert C. Burton (reviewer) Folakemi T. Odedina (reviewer)
by each country, of a national cancer policy and plan is an im- perative. Political, financial, and technical support is required. Progress could be furthered by making existing cancer plans operational and, in particular, by increasing resources for early detection and treatment capacity.
With increasing awareness by pol- icy-makers of the burden that non- communicable diseases (NCDs) – including cardiovascular disease, diabetes, and cancer – impose on national health systems and econo- mies, WHO has been asked to pro- vide data indicating the prepared- ness of national health systems to prevent and control NCDs. Together with monitoring of the cancer burden by cancer registries, cancer capac- ity assessment is a fundamental re- quirement for rational cancer control planning. The WHO Global Action Plan for the Prevention and Control of NCDs 2013–2020, endorsed by the 66th World Health Assembly (resolution WHA66.10) in 2013, ad- dresses an urgent requirement to provide the information relevant to national NCD planning. National capacity data from the WHO 2013 Noncommunicable Disease Country Capacity Survey [1] indicate the
current preparedness of countries to undertake cancer control.
A WHO survey of Member States The WHO 2013 Noncommunicable Disease Country Capacity Survey was a further development of a regular survey conducted periodi- cally to assess responses by indi- vidual countries to NCD challenges. Cancer-control-specific items are an integral part of the survey. Those re- sults that are common to all NCDs are published in country profiles [2], which provide an overview of nation- al priority setting for NCDs and the availability of corresponding inter- ventions. The NCD country profiles complement the WHO global report on the burden of these diseases and respective recommended interven- tions [3].
National NCD capacity is as- sessed by the dimensions of public health infrastructure and partner- ships focused on NCDs, taking account of relevant strategies, pro- grammes, and action plans, health information systems, and health- care capacity. The questionnaire used in the 2013 Noncommunicable Disease Country Capacity Survey included these dimensions in four modules that captured information on cancer. Cancer control capacity is defined as the availability of plans,
Chapter 6.2 • Current global national cancer control capacity 539 6 C
O N
T R
O L
C H
A P
T E
R 6
.2
programmes, and services covering the essential components of a na- tional cancer control programme [4], including prevention, early detection, treatment, and palliative care.
Information to assess the na- tional capacity for prevention and control of cancer was extracted from the 2013 Noncommunicable Disease Country Capacity Survey by using methods that have been previously used and described in detail [1]. The relevant questionnaire was sent in 2012 to appropriate authorities within the ministries of health of all 193 WHO Member States. Detailed information about the situation with respect to NCD control was received from 176 countries.
National cancer plans and their implementation Ministries of health in the majority of participating countries, specifically 149 of 176, reported having a national cancer policy, strategy, or action plan, hereafter referred to as a “cancer plan”. There are, however, regional differences. The European Region
has the highest coverage of national plans, with 49 of the 50 countries hav- ing plans, while the South-East Asia Region has the lowest proportion of national cancer plans, with 7 of the 10 countries so identified. In the African Region, 27 of 37 countries reported having a cancer plan. Overall, cancer control is formally included in nation- al health planning in the majority of countries.
To gain more detailed insight into national cancer control planning processes, several criteria were as- sessed. These included whether the plan was operational as well as the availability of dedicated funding for the plan. The inclusion of these additional criteria enabled a more differentiated assessment of the status of national cancer plans and policies. In the African and Eastern Mediterranean Regions, in contrast to the high percentage of countries that reported having a formal can- cer plan, only 36% of the countries
Fig. 6.2.1. Mothers and children at a health-care clinic in Sarlahi District, Nepal. In low- income countries generally, the burden associated with cervical cancer may be mark- edly reduced by the incorporation of appropriate screening into routine health services.
Table 6.2.1. Results from the WHO 2013 Noncommunicable Disease Country Capacity Survey: percentage of countries with cancer policies, strategies, or action plans and their functionality
WHO regiona Number of countries
Percentage of countries with cancer policies, strategies, or plans
Existing policy
Operational policy with funding
African 37 73% 38%
Americas 31 90% 65%
Eastern Mediterranean 21 76% 43%
European 50 98% 84%
South-East Asia 10 70% 70%
Western Pacific 27 85% 78%
World Bank income groupb
Low 29 79% 45%
Lower-middle 43 84% 58%
Upper-middle 49 84% 65%
High 55 91% 78%
Total 176 85% 64%
a For a list of countries corresponding to each WHO region, see http://www.who.int/about/regions/en/ index.html.
b For definitions of World Bank income groups, see http://data.worldbank.org/about/country- classifications.
540
reported that the plan was operation- al and only 43% had dedicated fund- ing available for the implementation of the plan. Similar, but less marked, differences were evident in the other WHO regions, where allocation of re- sources for implementation of many national plans is lacking. In addition, the specification of a plan does not guarantee that its implementation is monitored or that there is a clear definition of measurable anticipated outcomes.
Approximately 15% of all partici- pating countries did not have a cancer plan. Overall, only 64% of countries reported having cancer control plans that were operational and funded. Analysis of the available findings by World Bank income group revealed a consistent gradient from low-income countries to high-income countries with respect to the specification and completeness of national cancer plans. High-income countries are more likely to have plans that are also funded and monitored with regard to their defined outcomes (Table 6.2.1). However, even within countries in the highest income group, characterized by 90% of countries having a national cancer plan, only a proportion (78%), rather than all countries, confirmed that these plans were operational and funded.
Monitoring the cancer burden The availability of data establishing the national cancer burden is the ba- sis for any rational cancer planning and the ordering of priorities. Cancer registries therefore play a key role in national cancer planning. In the ma- jority of countries, i.e. in 80% of the
176 countries responding, a registry for cancer was reported, regardless of the type of the registration under- taken (Table 6.2.2). The European Region exhibited the highest propor- tion of countries with cancer regis- tries (94%), and the South-East Asia Region had the lowest proportion of countries (60%). In 25 of 37 coun- tries in the African Region and in 24 of 27 countries in the Western Pacific Region, there was a moni- toring system for cancer.
The various cancer registries were further differentiated from each other by type of registry – some of the key differences are whether the registry is popula- tion-based or hospital-based, and whether the population covered is national or is identified at a less comprehensive level. In almost all regions, population-based reg- istries are the predominant type, accounting for 69% in the Eastern Mediterranean Region and 81% in the European Region. The region in which population-based cancer registries are the least common is the South-East Asia Region (50%). There is a consistent link between cancer registration and national
Fig. 6.2.2. The Children’s Vaccine Pro- gram is working with the Cambodian government and other partners to make hepatitis B vaccines available to all chil- dren in the country. Here, health-related brochures are distributed to villagers by the mobile health team in the village of Choumpou in Kampong Cham Province.
Table 6.2.2. Results from the WHO 2013 Noncommunicable Disease Country Capacity Survey: percentage of countries with cancer registries and type of registry
WHO regiona Number of countries
Percentage (number) of countries with cancer registries
Existing cancer registriesb
National registriesc,d
Subnational registriesc,d
Population-based registriesd
Hospital-based registriesd
African 37 68% (25) 32% (8) 68% (17) 64% (16) 36% (9)
Americas 31 71% (22) 64% 36% 77% (17) 18% (4) (missing 1)
Eastern Mediterranean 21 76% (16) 69% 31% 69% (11) 25% (4) (missing 1)
European 50 94% (47) 81% 19% 81% (38) 15% (7) (missing 2)
South-East Asia 10 60% (6) 17% 83% 50% (3) 50% (3)
Western Pacific 27 89% (24) 75% 25% 58% (14) 42% (10)
World Bank income groupe
Low 29 59% (17) 35% 65% 47% (8) 47%
Lower-middle 43 72% (31) 35% 65% 48% (15) 52%
Upper-middle 49 84% (41) 68% 32% 71% (29) 22%
High 55 93% (51) 88% 12% 92% (47) 8%
Total 176 80% (140) 64% 36% 71% (99) 26% a For a list of countries corresponding to each WHO region, see http://www.who.int/about/regions/en/index.html. b Data reporting on the answer “yes” to the overall question “Does your country have a cancer registry (of any kind)?” c Data reporting on the answer to the question “Does your country have a national cancer registry vs a regional registry?” The positive response to this
question does not imply that there is national coverage. d Percentage of countries out of the total responding “yes” to the overall question “Does your country have a cancer registry (of any kind)?” e For definitions of World Bank income groups, see http://data.worldbank.org/about/country-classifications.
Chapter 6.2 • Current global national cancer control capacity 541 6 C
O N
T R
O L
C H
A P
T E
R 6
.2
Cervical cancer prevention in six African countries Nathalie Broutet
A WHO demonstration programme for cervical cancer prevention began in September 2005 and was com- pleted in May 2009. Training of proj- ect coordinators was organized at the University of Zimbabwe, Harare. Training in data management was undertaken at all project sites, with the African Population and Health Research Center (APHRC) and IARC. The project created aware- ness in communities about cervi- cal cancer and its prevention in six African countries: Madagascar, Malawi, Nigeria, Uganda, the United Republic of Tanzania, and Zambia.
Women were counselled and offered screening using visual in- spection with acetic acid (VIA), and patients with a positive screening test were treated using cryothera- py. Patients who were not eligible for cryotherapy were referred to a higher level of health care for fur- ther evaluation and treatment (Fig. B6.2.1). Continuous monitoring and
evaluation of the project was carried out by IARC and APHRC, to gener- ate evidence about the acceptability and feasibility in a primary health- care setting, referral site, or district hospital. The project targeted all women resident in the catchment area and aged between 30 and 50 years.
Between September 2005 and May 2009, a total of 19 579 wom- en were screened in the six coun- tries. Overall, 10.1% of screening with VIA results were positive, and 1.7% of women had lesions suspi- cious of cancer on inspection. A total of 87.7% of all VIA-positive cases were eligible for cryotherapy. The majority of women (63.4%) received cryotherapy within one week of initial screening. The sin- gle-visit approach enabled 39.1% of women to be screened and treated on the same day. However, more than 39.1% of all women eligible for cryotherapy did not receive
treatment, for various reasons, including equipment failure and women needing to obtain consent from their spouses before cryo- therapy could be done. The VIA and cryotherapy procedures were well tolerated by the women, and almost all of those who underwent these procedures said they would recommend them to other women.
This programme has shown that the “screen and treat” approach can be introduced into existing reproduc- tive health services in low-resource countries. Screening for precancer- ous lesions using VIA and treatment with cryotherapy is acceptable and feasible at low-level health facili- ties. VIA is an attractive alternative to cytology-based screening in low- resource settings. The alternative simple and safe cervical cancer prevention techniques simplify the process and render it feasible and acceptable to women and service providers in low-resource settings.
At the final meeting of the proj- ect, country teams presented plans on how best to scale up cervical cancer prevention services using the “screen and treat” approach. Training of an adequate number of providers, sustainable supervi- sion, and supply and maintenance of equipment and consumables are key elements of the programme. The country teams noted that fund- ing shortages and limited human re- sources are some of the factors that may prevent the ministries of health in the six countries from sustain- ing and scaling up the programme. However, at least four of these six countries now have national cervi- cal cancer prevention programmes.
Fig. B6.2.1. Programme for strengthening cervical cancer prevention: operational framework. VIA, visual inspection with acetic acid.
income. In the highest income cate- gory, 93% of countries have a moni- toring system for cancer, and this proportion falls to 59% for countries in the lowest income group. This gradient in cancer monitoring with the level of prosperity is particularly
dramatic for population-based reg- istries, which are available in only 8 of the 29 countries in the lowest in- come group compared with 47 of the 55 countries in the highest. Overall, 36 countries reported that they have no cancer-specific registration.
Prevention and early detection policies Prevention Certain declared policies to prevent cancer are also amenable to com- parison and evaluation. Interventions based on exposure to relevant
542
infectious agents and environmental carcinogens, apart from those en- countered through lifestyle choice, were not addressed by the survey. Of the 176 responding countries, 92% report having a national policy
for tobacco control, 77% have policies on harmful use of alcohol, 84% address unhealthy diet, 77% act on overweight or obesity, and 81% have a policy that focuses on improv- ing physical activity. Although more
countries have a policy for tobacco use than have policies in relation to any other NCD-based prevention strategy, only 69% of the responding countries have a tobacco policy that was reported as operational and as having a dedicated budget.
Early detection Several cancer types of major public health relevance have a high prob- ability of being cured if they are detected early in their clinical course and relevant cases are then referred, in a timely fashion, for definitive diag- nosis and treatment at secondary or tertiary levels of care.
Population-based screening for certain cancers is established (see Chapter 4.7) and subject to national survey. Cytology (Pap smear) testing for cervical cancer is the most widely applied detection procedure but re- quires specialized laboratories and a recall system for test-positive cases. Visual methods, such as inspection of the cervix after acetic acid staining, are options for low-income settings because no sophisticated technol- ogy is required (see “Cervical cancer
Fig. 6.2.3. The World Health Assembly in Geneva, 2011. WHO initiatives, in relation to the global burden of noncommunicable diseases generally, have permitted assessment of national policies for cancer control and their implementation. The data indicate that low-income countries face challenges in implementing proven means to reduce cancer incidence and mortality.
Table 6.2.3. Results from the WHO 2013 Noncommunicable Disease Country Capacity Survey: percentage of countries where tests and procedures for the early detection of cancer are available
WHO regiona Number of countries
Percentage of countries where available
Pap smear cervical cytology
VIA FOBT DRE Breast palpation Mammography
African 37 38% 24% 14% 16% 57% 30%
Americas 31 97% 45% 81% 65% 94% 61%
Eastern Mediterranean 21 38% 24% 38% 24% 81% 43%
European 50 80% 56% 80% 68% 86% 70%
South-East Asia 10 20% 10% 0% 10% 60% 0%
Western Pacific 27 67% 30% 63% 56% 89% 56%
World Bank income groupb
Low 29 21% 14% 3% 7% 52% 17%
Lower-middle 43 42% 28% 26% 23% 72% 23%
Upper-middle 49 78% 31% 65% 51% 90% 55%
High 55 91% 62% 93% 80% 91% 85%
Total 176 64% 37% 54% 46% 80% 51%
DRE, digital rectal examination; FOBT, faecal occult blood testing; VIA, visual inspection with acetic acid. a For a list of countries corresponding to each WHO region, see http://www.who.int/about/regions/en/index.html. b For definitions of World Bank income groups, see http://data.worldbank.org/about/country-classifications.
Chapter 6.2 • Current global national cancer control capacity 543 6 C
O N
T R
O L
C H
A P
T E
R 6
.2
prevention in six African countries”). Mammography is used to detect ear- ly-stage breast cancer. Clinical breast examination by palpation performed by health professionals is a low-cost approach to detect early breast can- cer. Rectal cancer can be detected at an early stage by clinical examination
(digital rectal examination). Faecal occult blood testing is in use for early detection of colorectal cancer.
Ministry of health officials were asked whether tests for early detec- tion of cervical, breast, and colorectal cancers were available in the national public system at primary health-care
level. Responses received are sum- marized in Table 6.2.3. A more de- tailed assessment of the presence of a specific detection programme (screening programme) was not in- cluded in the survey.
Concerning detection of cervical cancer and precancer, 64% of the
Traditional household cooking: risks and prevention Majid Ezzati
More than 2 billion people world- wide use biomass fuels (firewood, charcoal, animal dung, and crop residues) and coal for cooking and heating. Biomass use is currently most common in sub-Saharan Africa [1], followed by South Asia. In devel- oping countries, these fuels, which together are called solid fuels, are often burned in open fires or tradi- tional stoves with poor combustion efficiency. Solid-fuel smoke contains many solid (particle) and gaseous pollutants, including known human carcinogens. As a result, people liv- ing in homes that use solid fuels, es- pecially women and young children who may spend hours near cooking fires, have high exposures to parti- cles and other pollutants.
There is increasing epidemiolog- ical evidence that exposure to bio- mass and coal smoke is a risk factor for childhood pneumonia, chronic obstructive pulmonary disease, lung cancer, and cataracts, with some evidence of a dose–response rela- tionship. Evidence is also emerging for effects on tuberculosis, low birth weight, and cardiovascular out-
comes. Exposure to air pollution from solid fuels is responsible for an estimated 3.5 million deaths per year and more than 4% of the global disease burden. Importantly, the ma- jority of the burden is now from non- communicable diseases. The health effects of household air pollution are largest in low-income regions and in poor households.
The large disease burden in poor communities and households has helped make solid fuel use a major global health concern. For ex- ample, solid fuel use is an indicator for Goal 7 (ensuring environmental sustainability) of the United Nations Millennium Development Goals. Yet there are several challenges in implementing interventions that re- duce exposure by a large amount. Alternative stove technologies, de- spite having gained support among activists in developed countries, have proven not to be a viable option for intervention in most places be- cause user behaviours modify their
performance and because they re- quire regular maintenance. Clean fuels such as electricity and natural gas will certainly reduce exposure substantially and provide substantial health gains [1,2]. However, wide- spread use of these clean fuels is restricted by their cost, uncertainty about prices, and the absence of an energy infrastructure that allows households to buy fuels when they need them and without having to travel long distances [3]. In addition to clean fuels, it may be possible to “pre-process” biomass and coal to burn more cleanly, by converting it to gaseous or liquid fuels, charcoal, or briquettes.
In summary, research over the past two decades has helped to establish air pollution from house- hold solid fuel use as an important global health risk, especially among the poor [4]. The increased visibility must be sustained but needs to be accompanied with efforts to reduce the burden of disease, including cancers, caused by emissions from indoor combustion.
References 1. Bailis R et al. (2005). Science, 308:98–103.
http://dx.doi.org/10.1126/science.1106881 PMID:15802601
2. Lin HH et al. (2008). Lancet, 372:1473– 1483. http://dx.doi.org/10.1016/S0140-6736 (08)61345-8 PMID:18835640
3. Ezzati M DM et al. (2004). Annu Rev Environ Resour, 29:383–420. http:// d x.d o i.o r g /10.114 6 /annurev.ene r gy. 29.062103.121246
4. IARC (2010). IARC Monogr Eval Carcinog Risks Hum, 95:1–430. PMID:20701241
Fig. B6.2.2. A traditional three-stone open fire, as used in many households in developing countries.
Fig. B6.2.3. The traditional open fire used in many developing counties is associated with high concentrations of pollutants, as seen from the smoke rising from the roof of a house in central Kenya.
544
176 countries reported the availability of a cytology-based screening test at primary care level, a major increase over the corresponding data reported in 2010 (26%) (Box 6.2.1). The pres- ence of cytology services was still low in the African Region (38%) and the South-East Asia Region (20%), as was the availability of visual in- spection after acetic acid staining, with 24% of countries in the African Region and 10% in the South-East Asia Region reporting positively. Compared with cytology tests, visual inspection was generally less avail- able in all regions. The availabil- ity of breast palpation by a health worker as routine in primary health care was low in the African Region, where 57% of countries reported availability, and in the South-East Asia Region, where the procedure was available in 60% of countries. In Europe, the Eastern Mediterranean, and the Americas, breast examina- tion was reported as generally avail- able in primary health-care facilities in 86%, 81%, and 94% of countries, respectively.
The situations concerning avail- ability of radiotherapy and chemo- therapy for cancer treatment in the public health sector, as well as the
availability of palliative care servic- es, were also subject to report in the survey but are outside the scope of the present Report.
Imperatives for improvement of cancer control Analysis of the data collected by the Noncommunicable Disease Country Capacity Survey provides an over- view of the extent of national capac- ity to address the cancer burden in almost all WHO Member States. Although the survey was designed to assess national capacity for NCD control from the perspective of an integrated approach, it is practica- ble to develop a comparable global picture specifically for cancer control along the continuum from prevention to palliative care. The summation in this chapter has addressed cancer control apart from clinical care.
The recognized limitations of such an analysis arise from the fact that the sources of information were NCD ref- erence points in ministries of health not necessarily closely connected with the cancer control situation. The design of the survey allowed the assessment of key areas of cancer control to inform a global rating, but without any details about, for
example, the proportion of the popu- lation covered by defined interven- tions (see “Traditional household cooking: risks and prevention”). The question and response structure of the survey may not have allowed re- spondents to reflect in detail the na-
Box 6.2.1. Trends in national cancer control capacity.
Of the 176 countries that participated in the WHO 2013 Noncommunicable Disease Country Capacity Survey, responses from those with prior data from a 2010 study were assessed for trends.
Overall, the percentage of countries with a cancer plan increased, from 81% in 2010 to 86% in 2013, as did the percentage with an operational cancer plan (from 61% to 65%). The largest gain in national cancer plans was in low-income countries (from 54% to 79%). However, increases in an operational cancer plan followed an income gradient: high-income countries had the most advance- ment from 2010 to 2013 (from 74% to 81%), whereas there was little or no gain in lower-middle and low-
income countries (remaining at about 56% and 43%, respectively).
The percentage of countries with cancer registries, regardless of their type, remained largely un- changed (80%). Further differen- tiation found an increase in popu- lation-based registries rather than hospital-based registries, from 46% in 2010 to 59% in 2013. While na- tional-level registries among high- income countries remained at a steady 85%, overall this percentage decreased from 64% to 49%, a drop distributed across all other income groups and regions.
The availability of cervical cancer screening by visual inspection af- ter acetic acid staining increased in European countries (from 27%
to 56%) and in African countries (from 14% to 24%). The availabil- ity of cervical cytology and breast cancer screening by palpation was steady across all countries. Only regional differences were seen: cervical cytology access increased in Africa (from 27% to 38%) and in the Western Pacific Region (from 56% to 67%) while decreasing significantly in Eastern Mediterranean countries (from 57% to 38%). Notably, the South-East Asia Region showed a consistent 20% drop in all survey indicators, including the presence of a cancer plan, an operational cancer plan, cancer registries, and access to cervical cytology and breast cancer screening by palpation.
Fig. 6.2.4. Cover of WHO Guide for Effec- tive Programmes on cancer control. This publication is a series of six modules on cancer control: Planning, Prevention, Early detection, Diagnosis and treatment, Palliative care, and Policy and advocacy.
Chapter 6.2 • Current global national cancer control capacity 545 6 C
O N
T R
O L
C H
A P
T E
R 6
.2
tional particularities of health-care systems and their organizational structures. Finally, although the ques- tionnaire and its instructions were available in almost all WHO official languages (English, French, Spanish, and Russian), there may have been language barriers with regard to the definitions of technical terms and their meaning.
Despite such limitations, the key results available provide evidence of major gaps between what is known about effective cancer control and the implementation of measures to reduce the cancer burden. • The formal specification of a na-
tional cancer control policy or plan does not guarantee implementa- tion. Government officials were very clear in reporting major dis- crepancies between the endorse- ment of a formal government policy or plan and its operational status. In only two thirds of all respond- ing countries did a cancer plan or
policy have a specific budget that would allow the necessary organi- zational arrangements as recom- mended by WHO (see Chapter 6.1).
• Established cancer prevention policies were far from being im- plemented universally. Although there is definitive knowledge about the behavioural aspects of cancer causation, such as tobacco use, and there are effective strategies to reduce such use, as outlined in this Report (see Chapter 4.1), na- tional implementation capacity to prevent cancer caused by tobacco and alcohol use, unhealthy diet, obesity, and lack of physical activ- ity is markedly limited.
• Early detection of cancer was gen- erally absent from the public care delivery systems in most countries in the low and lower-middle income groups. Because cervical cancer is of major public health relevance in most of these countries, increasing access to early detection through
organized cervical cancer screen- ing will have a major impact on the cancer burden.
The survey identified major gaps and challenges for national health- care systems to achieve better can- cer control. The majority of countries reported the existence of a national policy, strategy, or action plan to ad- dress cancer. Even among countries in the lowest income group, where communicable diseases such as HIV/ AIDS, tuberculosis, and malaria are the major focus of national health planning and international health aid, national cancer plans are at least formally specified in more than half. However, to be effective, these policies and plans must be further developed and funded. Every country will have to determine national priori- ties for cancer control in the context of other NCDs and the opportunities that health system reforms will offer.
References
1. WHO (2012). Assessing National Capacity for the Prevention and Control of Noncommunicable Diseases: Report of the 2010 Global Survey. Geneva: WHO. Available at http://www.who.int/chp/know ledge/national_prevention_ncds/en/.
2. WHO (2011). Noncommunicable Diseases Country Profiles 2011. Geneva: WHO. Available at http://www.who.int/nmh/publi cations/ncd_profiles2011/en/.
3. WHO (2011). Global Status Report on Noncommunicable Diseases 2010. Geneva: WHO. Available at http://www.who.int/chp/ ncd_global_status_report/en/.
4. WHO (2002). National Cancer Control Programmes: Policies and Managerial Guidelines, 2nd ed. Available at http://www. who.int/cancer/publications/nccp2002/en/ index.html.
Websites WHO Global Action Plan for the Prevention and Control of NCDs 2013–2020: http://www.who.int/nmh/events/ncd_action_ plan/en/
WHO Guide for Effective Programmes – Module 1: Planning: http://www.who.int/cancer/publications/ cancer_control_planning/en/index.html
546
Cancer control in China: preventive policies and accessible health care Guiqi Wang
In 2010, cancer accounted for 23.82% of overall deaths and be- came one of the leading causes of death in China [1]. Most of the cases were diagnosed at a medium or late stage, so the efficacy of treatment was unsatisfactory. The health-care cost of these diseases created a severe economic burden on many individuals and on society, and be- came a major cause of poverty or returning to poverty for the patients and their families. The Chinese gov- ernment places great importance on cancer control and carries out ac- tive preventive strategies and cancer control projects in collaboration with all sectors of society.
In 2005, China initiated a nation- al programme of early detection and treatment for oesophageal cancer, gastric cancer, colorectal cancer, liv- er cancer, nasopharyngeal cancer, and lung cancer in high-incidence areas, and in 2009, China started a national screening programme for cervical cancer and breast cancer in rural areas. In both programmes, early detection and treatment has be- come an effective strategy for cancer control. For example, in some local areas of China, people have had very
high rates of oesophageal cancer for a long time. The early detection and treatment programme screens residents in those areas aged 40– 69 years by endoscopy with iodine staining and biopsy of early lesions. Patients with early-stage neoplasia, including severe squamous dyspla- sia, carcinoma in situ, and intramu- cosal carcinoma, can receive early treatments in a timely manner.
To accumulate experience and optimize techniques for this plan, the national programme first chose eight high-incidence areas as demon- stration sites and screened about 13 000 high-risk adults each year. Now, after a step-by-step expansion, 88 high-incidence areas in 26 prov- inces are participating in this early detection and treatment programme for oesophageal cancer [2]. From 2006 to 2012, 412 641 adults from high-risk areas were screened by endoscopy, and 4011 patients were diagnosed with severe precancer- ous lesions or early-stage cancer [2]. Most of these patients received timely treatment, with great benefit to health, and an economic analysis of this programme has shown that it is cost-effective [3].
In addition to such early detec- tion and treatment programmes, the Chinese government promotes widespread public education about cancer prevention and treatment, and has conducted interventions to control cancer occurrence, including neonatal vaccination against hepa- titis B virus, programmes promoting better nutrition, and targeted pro- grammes to improve occupational safety.
The China National Central Cancer Registry has also improved the national cancer registry sys- tem, and since 2008 has reported national cancer registry data annu- ally. In 2012, there were 222 cancer registry sites, covering 200 million people nationwide.
References 1. Yu W (2012). National Disease Monitoring
System 2010: Data Set of Death Causes. Beijing: Military Medical Science Press.
2. Diseases Prevention and Control Bureau of Ministry of Health, Cancer Foundation of China, Committee of Experts of the Cancer Early Detection and Early Treatment Project (2012). Report of the Cancer Early Detection and Early Treatment Project 2011/2012.
3. Dong Z et al. (2012). Chin J Oncol, 34: 637–640.
Chapter 6.3 • Health systems strengthening for cancer control 547 6 C
O N
T R
O L
C H
A P
T E
R 6
.3
Summary • Developing countries need
stronger health systems to sup- port essential cancer control efforts. The experience in high- income countries has dem- onstrated that cancer control cannot succeed without well- functioning and flexible health systems.
• In most developing countries, health-related infrastructure and trained human resources are lacking or poorly developed. However, when an infrastructure is established, the availability of treatment services indicates the likelihood of further positive de- velopments within the context of a national cancer control plan.
• Since 2005, WHO and the International Atomic Energy Agency have engaged in dem- onstration projects using exist- ing radiation medicine capacity in developing countries to initiate multidisciplinary cancer capac- ity building programmes. These programmes complement and enhance the clinical and public health impact of treatments by concurrently building capacity for all aspects of cancer control, including advocacy, epidemiol- ogy, prevention, early detection,
6.3 6 CONTROL
Health systems strengthening for cancer control Massoud Samiei Benjamin O. Anderson (reviewer)
Eduardo L. Cazap (reviewer) Nobuo Koinuma (reviewer)
diagnosis and treatment, pallia- tive care, and society building.
• Basic strength in the health system, exemplified by a radio- therapy service with adequate personnel and diagnostic facili- ties, has led to the initiation of a more robust cancer control plan, which in turn tends to further strengthen the health system.
With 8.2 million deaths due to cancer in 2012, cancer remains a leading cause of death and disability world- wide [1]. As highlighted throughout this Report, the pattern of disease has shifted so that an increasing burden now falls on low- and mid- dle-income countries, where health- care systems are not designed, or prepared, to tackle chronic noncom- municable diseases in general. As population age distributions trend higher [2] and unhealthy lifestyles are increasingly adopted, popula- tions in low- and middle-income countries face an expected rise in annual cancer incidence of nearly 70% by 2030 relative to the 2010 rates [3]. In 2010, there were more than 7.5 million new cases of cancer in low- and middle-income countries, and less than 30% of those patients had access to any reasonable treat- ment services [4].
For particular malignancies, specific modes of therapy, health technologies, and relevant skills and experience are recognized as the basis for treatment and, in some cases, cure.
All people, including those in low- and middle-income countries, are entitled to means of cancer prevention and appropriate care if they are diagnosed with cancer.
Fig. 6.3.1. In Sri Lanka, a patient with brain cancer looks at his scan. Sri Lanka is one of the WHO-IAEA Programme of Action for Cancer Therapy (PACT) Model Demonstration Sites. The country has im- plemented a successful comprehensive cancer control plan.
548
Strengthening national health sys- tems, by providing equitable and affordable access to cancer care for all who require it and making the essential clinical services, such as diagnostic imaging and radiother- apy, available, is an increasingly high priority for most governments in low- and middle-income coun- tries. This is also a priority for WHO and other United Nations agencies, such as IARC and the International Atomic Energy Agency (IAEA), and active nongovernmental or- ganizations such as the Union for International Cancer Control [5] and the International Network for Cancer Treatment and Research [2]. This imperative has been given new em- phasis after the resolution approved by all United Nations Member States in September 2011 on the preven- tion and control of noncommunica- ble diseases.
In 2012, the World Health As- sembly set a global target of a 25% reduction in premature mortality from noncommunicable diseases by 2025, among eight other volun- tary targets [6]. As highlighted in the discussion of national cancer control plans (Chapter 6.1), the prevention and control of cancer should be in- tegrated into all major actions for other noncommunicable diseases. Various aspects of cancer control plans – from prevention to manage- ment of cancer and related research priorities – should be part of the na- tional initiatives for strengthening each country’s health system [7].
Treatment: an essential element of cancer control As a growing number of cancer pa- tients seek relief from pain and suf- fering in most developing countries, the traditional focus of attention by health authorities on treatment as the first priority is understandable. Indeed, even with the optimal appli- cation of cancer prevention strate- gies, benefits are evident only after the passage of about 20–30 years, and are not relevant to millions of new cancer patients diagnosed in the interim, specifically in developing
countries. However, early diagnosis and/or screening combined with ad- equate treatment – often designated as secondary prevention – of certain common cancers has the potential to improve cure rates and to reduce mortality from these cancers in a time frame of 5–10 years. If there is a pros- pect of effective treatment, people will be motivated to seek screening [8]. Moreover, availability of treatment involves families and different sec- tors of the community in a common effort to promote health care. The availability of effective treatment also builds public support and motivates a core of professionals who may then campaign for a higher national prior- ity to be accorded to comprehensive cancer control.
Over the past 30 years, a com- bination of early detection and new treatment modalities has contributed to a modest, but constant, decline in cancer mortality rates in high-income countries [9]. This decline is not only due to the availability of more effec- tive curative drugs, but is also the result of effective national cancer control plans, which have led to bet- ter public education and community
access to prompt diagnosis and treat- ment. Such programmes now result in increased health awareness and prevention, improved cure rates, and improved quality of life for cancer pa- tients in high-income countries [10].
The experience in high-income countries has demonstrated that cancer control cannot succeed with- out well-functioning, robust, and flex- ible health systems. More specifically, strengthening health systems to de- liver life-saving treatments requires well-developed technical infrastruc- ture and a sufficient number of clini- cians trained in treatment delivery. Among other services, pathology services to provide accurate diag- nosis and staging of cancers and laboratory facilities, equipment, and technicians to run screening tests and deliver radiotherapy treatments must also be available.
Unfortunately, in most low-income countries and some other countries, such infrastructure and trained hu- man resources are not available or are poorly developed. However, emerging evidence from many de- veloping countries establishes that availability of treatment services can
Fig. 6.3.2. The children’s cancer ward at Ocean Road Cancer Institute in Dar es Salaam, Tanzania. Despite the crowded conditions, these children are among those fortunate enough to receive treatment.
Chapter 6.3 • Health systems strengthening for cancer control 549 6 C
O N
T R
O L
C H
A P
T E
R 6
.3
create a high probability for further positive developments within the con- text of a national cancer control plan [11–13]. With proper planning and ap- propriate strategies, and availability of trained professionals, developing countries can achieve much higher survival rates, in addition to pre- venting and otherwise controlling cancers. In a comparative study of cancer outcomes across 12 countries in Africa, Asia, and Central America, cancer outcomes correlated with the level of development of health ser- vices [14].
The question is to what extent the models from high-income countries can be replicated in low-income set- tings, and what options are economi- cal and cost-effective in particular situations. For example, which di- agnostic and treatment modalities should be developed? The dilemma for policy-makers and health authori- ties in developing countries is often to determine what comes first. Options may include providing affordable means of treating a portion of the growing number of patients, particu- larly in terms of medicine and health technology such as diagnostic imag- ing and radiotherapy, or adapting the most feasible strategies for cancer prevention and screening services.
The development of radiotherapy capacity is evidentially a cost-effective
and resource-level-appropriate strat- egy in developing countries to initiate capacity building in treatment [15]. As demonstrated in several countries, radiotherapy can serve as an anchor to develop self-sustaining national cancer control programmes, which in turn could help improve and expand other cancer services and infrastruc- ture. Radiotherapy is fundamental to the optimum management of cancer patients, and provision of radiotherapy services is central to national cancer control strategies (Box 6.3.1) [16–20]. Although it requires long-term plan- ning and appropriate assessment of health-care resources, effective radiotherapy for many cancers can be comprehensively provided at moderate cost, without recourse to sophisticated technologies [21]. Most countries have later expanded their initial radiotherapy capacity to add chemotherapy and other essential capacity, including imaging, pathol- ogy, and surgery. Such expanded capacity has often become the focus around which a national cancer centre is established [11].
Health systems strengthened through radiotherapy planning Experience in many developing countries indicates that cancer control cannot achieve its potential
unless the existing health system can sustain and support some key requirements. These include the technical capacity to initiate or manage the national cancer control plan and to deliver certain services. The initiation of a national cancer control plan requires establishing a cancer policy and additional legisla- tive steps, setting or defining some key targets, and/or allocating funds for cancer control activities; this represents a top-down approach. However, quite often, where the leadership of existing cancer clinics or radiotherapy centres is supported by nongovernmental organizations, some members of the community may be particularly active in de- veloping cancer control proposals: a bottom-up approach. For most low- and middle-income countries, a combination of these two ap- proaches is more likely to succeed in initiating cancer control activities [22]. In countries that have been as- sessed, the initiation of cancer con- trol activities and plans has been prompted by the leadership of an existing radiotherapy centre (often the only one in the country) because such centres are the places that im- mediately appreciate the severity of the cancer burden. The tragedy in developing countries is that 80%
Box 6.3.1. Radiation medicine: a vital component of cancer diagnosis and treatment.
The first cure of cancer by radio- therapy was reported in 1899, a few years after Röntgen’s description of X-rays in 1895 and the discovery of radium in 1898 [16]. The tech- nology has evolved radically since the 1950s, and today knowledge of radiation medicine and the avail- ability of relevant technology are indispensable for cancer diagnosis, cure, and care, where radiation and radioactivity play fundamental roles. Depending on the stage and type of cancer, typically some 50–65% of all cancer patients require radio- therapy during the course of their disease, either as a single modal-
ity or in combination with surgery, chemotherapy, hormone therapy, and/or immunotherapy [17].
Worldwide, radiotherapy is a major aspect of investment in health-care systems to treat cancer. Low- and middle-income countries have far to advance if their patients are to have reasonable access to such treatment. Although these countries account for 85% of the world’s pop- ulation, there are only about 4400 megavoltage machines in low- and middle-income countries, less than 35% of the world’s radiotherapy facilities, leaving most cancer pa- tients in these countries without any
access to potentially life-saving ra- diotherapy treatment [18].
The International Atomic Energy Agency (IAEA) emphasizes that radiotherapy treatment is most ef- fective when it is linked to a com- prehensive national cancer control programme. For this reason, in 2004 IAEA established its Programme of Action for Cancer Therapy to help expand radiotherapy capacity in developing countries and build partnerships with WHO and others to address the huge disparities that exist in cancer control and services [19,20].
550
of cancer cases are diagnosed at a very advanced stage of disease.
The burden of late-stage diagno- sis of cancer in developing countries provides strong motivation to move such a country towards cancer con- trol planning rather than attempting to expand clinical facilities to treat the large number of patients. Countries that have committed to this option, and initiated some cancer control ac- tivities, have been more successful in seeking new funding to acquire additional treatment facilities and re- sources. Accordingly, basic strength in the health system, exemplified by a radiotherapy centre, can lead to the initiation of a more robust cancer control plan, which in turn can result in further health systems strengthen- ing through improvements in other essential treatment modalities, and most often the establishment of a na- tional cancer centre [11]. The direc- tors of such national cancer centres in more than 40 low- and middle- income countries have prompted the launch of cancer control plans [23]. Arguably, a minimum level of national capacity and infrastructure must be present within the health system of any country before any
cancer control activities can be initi- ated or launched successfully [13].
WHO-IAEA cooperation on cancer control To build on the premises already described, and after the establish- ment of the Programme of Action for Cancer Therapy (PACT), WHO and IAEA signed an agreement in March 2009 to launch a cooperation on cancer control, aimed at strength- ening and accelerating cancer con- trol in developing countries [24,25]. The formation of a WHO-IAEA cooperation began in May 2005 when the World Health Assembly adopted a far-reaching resolution in response to the marked increase in cancer incidence worldwide. Since then, PACT, WHO, IARC, and other key international cancer organiza- tions have undertaken increasingly productive collaboration. Particular emphasis is placed on providing assistance within a broad, multidis- ciplinary cancer capacity building programme that complements and enhances the clinical and public health impact of treatment.
The WHO-IAEA cooperation provides the framework for the two organizations to integrate their work,
building on their areas of expertise to create a more coordinated and ro- bust approach to combating cancer in low- and middle-income countries. The focus of WHO support within the cooperation is on public awareness, prevention, early detection, and overall care for patients. In particular, WHO emphasizes population-based interventions and the strengthening of public health approaches to pre- vention and control of cancer, with special attention to primary health care. National cancer control pro- grammes fit into the broader WHO framework to strengthen health sys- tems with a major focus on primary health care, and are part of the im- plementation of the Action Plan of the Global Strategy for the Preventions and Control of Noncommunicable Diseases, which was endorsed by the World Health Assembly in May 2008 and has been updated to cover the 2013–2020 period [26,27]. Since the launch of PACT, IAEA requires that all of its assistance related to radiation medicine be channelled through ministries of health to en- sure an integrated and balanced ap- proach. IAEA recommends that low- and middle-income countries should prepare their national cancer control plans and then specifically examine the need for expansion of diagnostic and treatment services based on lo- cal resources, types of cancer, and other relevant conditions.
Cancer control through demonstration projects Efforts in the WHO-IAEA coopera- tion on cancer control have been focusing on eight PACT Model Demonstration Sites (PMDS), in Al- bania, Ghana, Mongolia, Nicaragua, Sri Lanka, the United Republic of Tanzania, Viet Nam, and Yemen [28], most of which have been successful in using their existing radiotherapy programmes to embark on develop- ing cancer control strategies. WHO and IAEA have assisted these coun- tries in needs assessment, evalua- tion of possible strategies, setting of priorities, and selection of resource- level-appropriate interventions.
Fig. 6.3.3. Ghana’s first cancer hospital was established in 1997 with IAEA support. The operation of this cancer centre is helping Ghana to expand its cancer infrastructure and capacity in three regions of the country within the scope of its national cancer control plan.
Chapter 6.3 • Health systems strengthening for cancer control 551 6 C
O N
T R
O L
C H
A P
T E
R 6
.3
Through the cooperation on cancer control, WHO and IAEA have also responded to requests for cancer control assessment and programme development assistance in more than 45 countries since 2005 [29,30]. Each country has particu- lar features in terms of the cancer burden, cancer risk factors, culture, health system, available financial and human resources, and infra- structure. Careful assessment of these elements helps to establish realistic and achievable priorities for action in a specific context [31]. WHO, IAEA, IARC, and partners assist national authorities in their cancer control endeavours, shar- ing the experiences of developing countries that have already achieved successes in cancer control. Once a country’s most pressing needs are identified and prioritized, part- ners and donors are approached for adequate response strategies, which may be on the technical and/ or financial side. Follow-up missions for monitoring progress are also con- ducted regularly.
Early evaluation of PMDS PMDS projects are at different stag- es of implementation, having started at different baselines. However, they all share very high motivation at the professional and political level. Each of the PMDS projects has worked diligently on developing a national cancer control plan and establish- ing a national steering committee to oversee the implementation of the plan. With support from partners, PACT has mobilized funds to support countries with expert advice, human resource development, acquisition of radiation medicine equipment, and cancer control capacity building [19].
As reported by a recent evalu- ation of PMDS [12], it is clear that in each PMDS project, availability of radiation medicine capacity, al- though still limited, has played a pivotal role as an initial driver of the country’s cancer control plan, and in the implementation of elements of each country’s national cancer con- trol programme. There are already signs and actions in PMDS projects demonstrating that health authori- ties have appreciated that health systems strengthening for cancer care and control cannot be achieved without balanced investments in pre- vention, early detection, diagnosis, treatment, and palliative care.
Hence, the success of national cancer control plans is dependent to a large extent on broader reform measures within the health system of each country to address the inequal- ities in health care and services. Equally important is the capacity of ministries of health, and their longer- term commitment to implementation and evaluation of the policies and strategies developed. In particular, if there is no commitment towards efforts aimed at prevention, coordi- nation of services, and affordability of, and access to, health services,
then there is unlikely to be an im- provement in health systems.
To ensure further progress and sustainability of cancer control programmes already initiated, all developing countries should con- sider establishing at least one gov- ernment-supported cancer centre of excellence based on existing can- cer centres to provide high-quality services and act as a driving force for national cancer control [11,23]. Furthermore, in view of prevailing limitations in workforce capability, for the long-term strengthening and success of cancer control, each country should have a 10–15-year training programme [13]. Cancer registries, health data collection, and health services research in all countries require support [32]. In this context, the IARC Global Initiative for Cancer Registry Development in Low- and Middle-Income Countries and the International Cancer Control Partnership, jointly coordinated by the UICC and the United States National Cancer Institute and involv- ing WHO, IAEA-PACT, IARC, and several key cancer organizations and individual experts, are immedi- ately relevant.
Fig. 6.3.4. Mongolia’s National Cancer Centre, established with WHO-IAEA sup- port. As part of implementation of the national cancer control plan, with funding provided by the United Nations Women’s Guild, the National Cancer Centre con- structed this playground on the premises to provide young cancer patients with a diversion from their treatment.
Fig. 6.3.5. The number of cancer patients threatens to overwhelm this children’s cancer ward in Hanoi, Viet Nam, where children sleep on floors with their family members nearby.
552
References
1. WHO (2011). United Nations High-Level Meeting on Noncommunicable Disease Prevention and Control. Available at http://www.who.int/nmh/events/un_ncd_ summit2011/en/.
2. International Network for Cancer Treatment and Research (2013). Cancer in Developing Countries. Available at http://www.inctr.org/about-inctr/cancer- in-developing-countries/.
3. Ferlay J, Shin HR, Bray F et al. (2010). GLOBOCAN 2008 v2.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
4. IAEA (2011). Inequity in Cancer Care: A Global Perspective. IAEA Human Health Reports, No. 3. Available at http://www-pub. iaea.org/books/IAEABooks/8180/Inequity- in-Cancer-Care-A-Global-Perspective.
5. Union for International Cancer Control: http://www.uicc.org.
6. WHO Media Centre (2012). 65th World Health Assembly closes with new global health measures. News release. Available at http://www.who.int/mediacentre/news/ releases/2012/wha65_closes_20120526/ en/index.html.
7. Sullivan R, Purushottham A (2010). Towards an international cancer control plan: policy solutions for the global cancer epidemic. INCTR Magazine, 9:1–8. Available at http://www.inctr.org/network-magazine/ past-editions/.
8. World Bank (2013). World Bank in Albania breast cancer campaign. Available at http:// web.worldbank.org/WBSITE/EXTERNAL/ NEWS/0,,contentMDK:23177166~menuP K:141310~pagePK:34370~piPK:34424~ theSitePK:4607,00.html.
9. Abegunde DO, Mathers CD, Adam T et al. (2007). The burden and costs of chronic diseases in low-income and middle-income countries. Lancet, 370:1929–1938. http:// dx.doi.org/10.1016/S0140- 6736(07)616 96-1 PMID:18063029
10. Levin V, Meghzifene A, Izewska J, Tatsuzaki H (2001). Improving cancer care: increased need for radiotherapy in developing countries. IAEA Bull, 43:25–32.
11. Sloan F, Gelband H, eds; Committee on Cancer Control in Low- and Middle- Income Countries (2007). Cancer Control Opportunities in Low- and Middle-Income Countries, Washington, DC: National Academies Press.
12. IAEA, PACT (2012). Evaluation of PMDS, 2011–2012. IAEA internal report.
13. Anderson B, Ballieu M, Bradley C et al. (2010). Access to cancer treatment in low- and middle-income countries – an essential part of global cancer control. A CanTreat Position Paper. Available at http://ssrn.com/abstract=2055441.
14. Sankaranarayanan R, Swaminathan R, Brenner H et al. (2010). Cancer survival in Africa, Asia, and Central America: a population-based study. Lancet Oncol, 11:165–173. http://dx.doi.org/10.1016/S14 70-2045(09)70335-3 PMID:200005175
15. IAEA (2010). Planning National Radiotherapy Services: A Practical Tool. IAEA Human Health Series, No. 14. Available at http://www-pub.iaea.org/books/ I A E A Books /8 419 / Planning - National - Radiotherapy-Services-A-Practical-Tool.
16. Rickwood P (2001). Saving a mother’s life. Radiotherapy offers new hope for women of child rearing age suffering cervical cancer in developing countries. Available at http://www.iaea.org/About/Policy/GC/ GC45/SciProg/sfradiotherapy.html.
17. IAEA (2008). Setting up a Radiotherapy Programme: Clinical, Medical Physics, Radiation Protection and Safety Aspects. Vienna: International Atomic Energy Agency.
18. Burkart W, Chhem RK, Samiei M (2010). Atoms for health: the IAEA’s contribution to the fight against cancer. In: Health G20: A Briefing on Health Issues for G20 Leaders. Available at http://www-naweb.iaea.org/ na/G20-IAEA-article-atoms-for-health.pdf.
19. IAEA-PACT (2013). International Atomic Energy Agency Programme of Action for Cancer Therapy. Available at http://cancer. iaea.org/.
20. Samiei M (2008). Building partnerships to stop the cancer epidemic in the developing world. UN Special, No. 676, September 2008. Available at http://www.unspecial. org/UNS676/t31.html.
21. Stewart BW, Kleihues P, eds (2003). World Cancer Report. Lyon: IARC.
22. Barton MB, Frommer M, Shafiq J (2006). Role of radiotherapy in cancer control in low-income and middle-income countries. Lancet Oncol, 7:584–595. http://dx.doi. org /10.1016/S1470 -20 45(0 6)70759 - 8 PMID:16814210
23. Samiei M (2013). Challenges of making radiotherapy accessible in developing countries. In: Cancer Control 2013: Cancer Care in Emerging Health Systems. Global Health Dynamics UK and INCTR Belgium, pp. 85–96.
24. WHO Media Centre (2009). WHO, IAEA join forces to fight cancer in developing coun- tries. News release. Available at http://www. who.int/mediacentre/news/releases/2009/ who_iaea_cancer_programme_20090526/ en/.
25. WHO, IAEA (2009). Arrangements be- tween the Directors General of the World Health Organization and the International Atomic Energy Agency for the WHO/IAEA Joint Programme on Cancer Control. Available at http://www.who.int/nmh/ events/2013/2009arrangements.pdf.
26. WHO (2009). 2008–2013 Action Plan for the Global Strategy for the Prevention and Control of Noncommunicable Diseases. Geneva: WHO. Available at http://www.who. int/nmh/publications/9789241597418/en/.
27. WHO (2013). Global Action Plan for the Prevention and Control of Noncommunicable Diseases 2013–2020. Geneva: WHO. Available at http://www.who. int/nmh/en/.
28. IAEA-PACT (2013). PACT Model Demonstration Sites (PMDS). Available at http://cancer.iaea.org/pmds.asp.
29. IAEA-PACT (2013). IAEA imPACT (inte- grated missions of PACT). Available at http://cancer.iaea.org/impact.asp.
30. IAEA-PACT (2012). In Malaysia, IAEA con- ducts 45th imPACT Cancer Assessment Review Mission. Available at http://www. iaea.org/newscenter/news/2012/impact assessment.html.
31. WHO (2011). National Cancer Control Programmes: Core Capacity Self- Assessment Tool. Available at http:// www.who.int/cancer/publications/nccp _ tool2011/en/index.html.
32. Hanna TP, Kangole ACT (2010). Cancer control in developing countries: us- ing health data and health services research to measure and improve ac- cess, quality and efficiency. BMC Int Health Hum Rights, 10:24. http://dx.doi. org/10.1186/1472-698X-10-24
Cancer control experiences 553 6 C
O N
T R
O L
Cancer control in France: towards patient- centred precision medicine Agnès Buzyn
The two successive cancer control plans endorsed by the president of France have helped shape a new era of coordinated and integrated cancer care, centred on the patients’ needs and grounded on evidence-based measures. They have helped to build the components of the current organizational cancer care frame- work in France, which is designed to deliver high-quality services and facilitate the evolution of stratified medicine. This organizational frame- work bridges both patient-centred and tumour-centred approaches, and encompasses the personal- ized care programme, specialized structures or organizations (for rare cancers, oncogeriatrics, oncopaedi- atrics, etc.), and molecular profiling of tumours. The ultimate goal, yet to be attained, is to bring high value to patients and survivors while fighting the roots of health inequities through fair access to innovation.
The personalized care pro- gramme offers patients their road map towards completion of their treatment, and helps them to navi- gate through referral from one point in the process to the next with ease.
The personalized care programme has helped address the fragmenta- tion of the cancer care pathway. As the number of cancer survivors in- creases in most high-income coun- tries, the next challenge is to further develop a survivorship plan that is fully integrated into the continuum of the person’s care, from the onset of the disease.
To strengthen the cancer care framework, specialized networks for rare cancers, and specific organi- zational frameworks for oncogeriat- rics and oncopaediatrics, have been developed based on scientific and clinical excellence, with integrated facilities and dedicated staff. They usually function as a network with a wide geographical coverage. They have close links with biobank re- sources and molecular genetics in- frastructures. The implementation of a quality assurance programme and a collaborative approach warrants that all network centres meet re- quired standards and deliver state- of-the-art cancer care.
Advances in the molecular profil- ing of tumour tissues have opened up an era of personalized cancer
treatment where therapies are matched to the individual tumour. Pivotal to this personalized ap- proach are drugs that have been designed to target a particular mo- lecular pathway affected in a cer- tain cancer type. Targeted therapies have been successfully introduced into clinical practice for breast and stomach cancers (HER2 overex- pression), lung cancer (EGFR muta- tions), and colorectal cancer (KRAS mutations). Molecular characteriza- tion of tumours has become a de- cisive factor in the choice of thera- peutic strategies for cancer patients. To meet this challenge, a network of molecular genetics centres with na- tionwide reach was developed. This organizational framework has now been operational for 4 years and is successful in delivering state-of- the-art targeted cancer treatments. Access time to novel therapies has been reduced. The latest genomic knowledge and technologies (next- generation sequencing) should ac- celerate progress towards precision medicine.
554
Summary • The development and evalua-
tion of effective health services for cancer control is dependent on high-quality prevention and related research, which in turn is reliant on the efficient op- eration of functional research infrastructures.
• Cohorts, biobanking facilities, and cancer registries play a key role in research and the devel- opment and evaluation of cancer control measures.
• From a global perspective, there is an increasing rate of initiation of large population-based co- horts and advanced biobank- ing facilities. International or- ganizations, including IARC, as well as national and other au- thorities operate multiple, often large, biobanks. The necessary resources on which cancer re- search is based are termed the study bases and are formed by longitudinal follow-up of cohorts and biobanks using cancer reg- istry linkages.
• International standardization and networking of infrastructures is required to enable compre- hensive investigations and to ensure the standardization and
6.4 6 CONTROL
Research infrastructure – biobanks, cohorts, registries, and data linkage Joakim Dillner Gustavo Stefanoff (reviewer)
Jim Vaught (reviewer)
comparability of research data and results.
Research infrastructure: essential resources for cancer research Research infrastructure refers to fa- cilities, databases, and collections together with related services used by the scientific community to con- duct research [1]. When operated on an international scale, research infrastructure offers unique research services to users from different countries. In addition, a tangential function of research infrastructure is to engage young investigators seeking to establish careers in sci- ence and so contribute to the growth of scientific communities. This goal can be achieved when research infrastructure serves as a hub for providing services in essential el- ements of research such as (i) ac- cess to high-quality and standard- ized data; (ii) access to high-quality, well-documented, and standardized biospecimens; (iii) access to high- performance molecular analysis platforms and/or molecular data generated on demand; (iv) develop- ment, implementation, and monitor- ing of best practices; and (v) educa- tion, expertise, and advice.
The concept of the knowledge tri- angle involves research, education,
and innovation [2]. A high-quality research infrastructure should be at the centre of the knowledge triangle. A successful research infrastructure does not merely serve to provide materials. Research infrastructure should also contribute by having a role in the production of knowledge through research, the diffusion of knowledge through education, and the application of knowledge through innovation.
To enable comparison of health outcomes and research results from different parts of the world [3], col- laboration between research infra- structure hubs on an international scale is imperative, specifically in
Fig. 6.4.1. Blood samples are taken from a newborn baby in Sweden for the national phenylketonuria (PKU) registry biobank. In Sweden, such blood samples for the screening of metabolic diseases, including PKU, have been stored since 1974 and comprise about 2.7 million samples.
Chapter 6.4 • Research infrastructure – biobanks, cohorts, registries, and data linkage 555 6 C
O N
T R
O L
C H
A P
T E
R 6
.4
particular areas such as standardi- zation, quality control, and coordi- nated development. An international infrastructure for cancer research will be required for progress in in- ternational cancer control.
There is a growing realization that research is an essential compo- nent for delivery of effective health- care and preventive services in any setting, and is not only relevant to high-income countries. Studies that have affected the development of this concept include the Dartmouth Atlas, a systematic inventory of the different health services delivered in various parts of the USA, their cost, and their effectiveness (www. dartmouthatlas.org). These studies have estimated that if the most cost- effective health services were used throughout, the quality of health care would increase, while provision of health care would be up to 30% less expensive [4].
In health services research, the medical practice under study is in place irrespective of whether there is ongoing research. The purpose of such research is simply to study whether the health services deliv- ered actually work, based on imme- diate experience. Although there is current political interest in research on the comparative effectiveness of health services, research on the ef- fectiveness of health services has a long tradition in many countries. One of the best known and estab- lished traditions of health services research is the activities of cancer registries to evaluate the effective- ness of screening and other cancer
control policies. When new policies are implemented, they are frequently evaluated using randomized study designs, in so-called randomized health services studies [5].
Cohorts: studies of health outcomes and disease etiology Cohort studies aim to provide a reliable database for investigation of risk factors for the diseases that affect a population. Basic informa- tion about the diseases affecting the population and the risk factors that cause these diseases is the foundation for planning of preven- tive services. A large number of healthy volunteers in the population provide data about living conditions and lifestyle (such as diet, exercise, smoking, and other habits), donate a series of biospecimens, and are fol- lowed up longitudinally, specifically in relation to disease development. While it may be inferred that most cancer is attributable to the impact of environmental factors rather than, for example, heritable risk, many causal factors remain to be specifi- cally identified. Cohort studies with comprehensive collection of data
and molecular analysis of biospeci- mens may clarify – and, in some in- stances, identify – these factors and the inherent possibility of prevention. This approach was proposed by Doll and Peto more than 30 years ago [6] and has received renewed attention with the emergence of effective and comprehensive resources for mo- lecular analysis of both genetic and environmental risk factors [7].
The Public Population Project in Genomics and Society (www.p3g. org) has a global cohort observa- tory that lists 15 international cohort networks and 79 individual cohorts. Many of these cohorts are extremely well funded. The best known exam- ple is the National Children’s Study in the USA, a US$ 200 million/year cohort study that monitors pregnant women and their children to iden- tify factors contributing to healthy childbearing and child development (www.nationalchildrensstudy.gov).
The largest among the compre- hensive cohorts is the European Prospective Investigation into Cancer and Nutrition (EPIC), designed to in- vestigate the relationships between diet, nutritional status, lifestyle, and other environmental factors and the incidence of cancer and other chronic
Fig. 6.4.2. Liquid nitrogen containers in the biobank of the Coriell Institute for Medical Research in Camden, New Jersey, USA.
Fig. 6.4.3. Analysis of biological samples, in this case by quantitative micrographic determination, may be undertaken in collaboration with remote investigators where tissues are stored, rather than basing collaboration on the transportation of samples to remote locations.
556
diseases. EPIC recruited more than 520 000 people in 10 European coun- tries: Denmark, France, Germany, Greece, Italy, the Netherlands, Nor- way, Spain, Sweden, and the United Kingdom [8]. The EPIC cohort has multiple associated working groups, concerned not only with most major cancer types but also with cardiovas- cular diseases, ageing, and diabetes.
Why are biobanks needed? Collections of human biological spec- imens are essential, both for cancer research and for improved clinical di- agnosis. The revolutionary develop- ment of high-performance molecular analysis platforms in recent years has meant that the rate-limiting step in translating the advances in basic research to cancer control is no longer the molecular analysis, but
access to well-characterized speci- mens with associated follow-up data. Because of biobanking, scientists no longer need to initiate sample collec- tion either individually or within their research groups – relevant samples are already available.
The main advantage of biobank- ing is that a long follow-up on health outcomes for particular specimens is immediately available without a fur- ther requirement for the passage of time. A translational research project need not be restricted to testing new specimens and allowing the passage of a decade or more to determine relevant health outcomes. Rather, the project design may be based on locating specimens that have been stored for decades under circum- stances where the health outcomes after specimen donation have been determined by registry linkages [9].
Prevention research may involve the search for novel biomarkers that can be used for early diagnosis and/or as new screening tests. This goal requires specimens taken before diagnosis, and a situation where diagnoses made at a later time are known. Explanations for differenc- es in cancer occurrence worldwide are central to an understanding of cancer etiology. Such research also requires specimens taken before diagnosis, but with data about the later diagnosis of cancer in particular individuals.
The study base and the role of cancer registry linkages The study base is a term that refers to the specimens and data on which scientific studies are based, i.e. the stored specimens and accessory
Fig. 6.4.4. Map of the Nordic region depicting the enrolment regions for the biobank cohorts that participated in a joint cancer registry linkage that identified more than 2 million sample donors and more than 100 000 prospectively occurring cancer cases.
Chapter 6.4 • Research infrastructure – biobanks, cohorts, registries, and data linkage 557 6 C
O N
T R
O L
C H
A P
T E
R 6
.4
data plus the results of longitudinal follow-up for health outcomes.
A major limitation of biobank- based research is that many biobanks are not regularly linked to the relevant cancer registries and thus the study bases for scientific studies are unknown. Even if the biobanks are regularly followed up, there may be a lack of comprehen- sive and updated overviews for in- vestigators about which study bases are available. The most advanced presentation of such study bases is currently the Danish National Biobank Register (www.biobanks. dk), which has an online, automat- ed service to determine the study bases by on-demand linkages of the major biobanks in Denmark with the cancer registry and other health data registries. Large-scale and standardized cancer registry linkages for many biobanks in the Nordic countries have also been performed by the Finnish Cancer Registry under the auspices of European Union biobanking proj- ects (Fig. 6.4.4) [10], most recently the FP7 project Eurocourse (Europe
against Cancer: Optimisation of the Use of Registries for Scientific Excellence in Research; www.eu rocourse.org). Together with the European Union biobanking plat- form BBMRI (Biobanking and Biomolecular Resources Research Infrastructure; www.bbmri.eu), Euro- course has also developed a for- mat for standardized description of available study bases.
The requirement for follow-up to exploit the potential of biobank spec- imens has important implications. In particular, this has implications for anonymity in respect of specimens. Although scientists who use biobank specimens have access only to cod- ed specimens, the identity of the do- nors must be known to the agency responsible for the follow-up [11].
In many respects, biobanks are similar to cancer registries. Biobanks constitute research infrastructure responsible for collection and stor- age of information about patients. Stringent and similar standards are required of biobanks and registries for strict handling of personal data to protect the integrity of individuals.
As linkage to cancer registries is essential for the follow-up of biobank specimens to define the study base, the standardized overview and de- scription of biobanks and biobank- based study bases can be han- dled by cancer registries. Because cancer registries are the disease registers with the longest follow-up and experience of data handling as well as international networking and standardization of data, it is conceiv- able that they could drive the excel- lence in realizing the potential of biobanks for health. A role model for this concept is the Janus Biobank, a population-based biobank with samples from 350 000 donors that is operated entirely by the Cancer Registry of Norway [12].
Biological resource centres Historically, every research study or diagnostic laboratory developed its own separate system for storing biospecimens. However, this frag- mentation resulted in expensive duplication and severely limited the cumulative use of specimens, which had been handled and stored by
Fig. 6.4.5. Geographical origin of the samples stored at the IARC Biobank (excluding EPIC samples).
558
differing procedures. International studies based on samples from sev- eral sources were not necessarily valid. With the growing operational complexity of biobanks, investigators are increasingly reliant on special- ized service facilities that handle the acquisition, quality control, stor- age, processing, and distribution of biospecimens [13]. These biobank- ing facilities are termed biological resource centres. A biological re- source centre will provide services for all steps in the scientific study, from contacting study participants to archiving research results.
Most biological resource centres collaborate in international biobank- ing networks committed to common international standards for collec- tion, labelling, annotation, process- ing, storage, retrieval, and analysis of the biospecimens, while ensuring biological safety and protection of personal data. Under the auspices of the United States National Cancer Institute, the Cancer Human Biobank (caHUB) is a national biorepository of human tissue, blood, and other bi- ological materials (http://cahub.can- cer.gov). In Europe, major networks of biobanks are the above-men- tioned Biobanking and Biomolecular Resources Research Infrastructure
and the International Society for Biological and Environmental Repositories (www.isber.org). There is increasing emphasis on the view from funding agencies and other stakeholders that publicly funded research materials should be open- ly accessible.
The IARC Biobank is an inter- national biological resource centre serving more than 50 studies from 30 countries worldwide. As is typical
for such centres, study designs in- clude cohorts, case series, preva- lence studies, and case–control studies (Fig. 6.4.5). A total of about 5.0 million samples from 1.5 mil- lion subjects are stored. Most of the samples are body fluids, including plasma, serum, and urine, as well as extracted DNA samples (Fig. 6.4.6). The IARC Biobank has an access policy to encourage proposals for new collaborative studies on exist- ing biospecimen collections (http:// ibb.iarc.fr/).
Cohort and biobanking networks The Low- and Middle-Income Coun- tries Biobank and Cohort Network (BCNet) is an IARC initiative in- tended to support the development of cancer control in low- and middle- income countries. Currently, invest- ments in research infrastructure calculated to assist the development of high-quality science, which will underpin effective cancer control measures, are receiving increased attention. Population cohorts can as- sist countries in providing scientific education, stimulating translational research, and identifying, launching, and evaluating interventions for im- proving health.
Fig. 6.4.6. An indication of the amount and nature of biological materials that may be subject to collection and storage in a biobank. In this instance, the data refer to samples stored at the IARC Biobank.
Fig. 6.4.7. Sharing of sample analysis results rather than samples may accelerate research: the concept of the biological expert centre.
Chapter 6.4 • Research infrastructure – biobanks, cohorts, registries, and data linkage 559 6 C
O N
T R
O L
C H
A P
T E
R 6
.4
The tools and methodologies used to develop a biobank infrastructure and population cohorts are basically similar. Thus, by comparison with earlier endeavours, lower cost and improved effectiveness in the con- text of establishing cohorts may be achieved by the operation of interna- tional networks. Such networks pro- vide initiating countries with guidance for establishing sustainable infra- structure, often involving procedures that have been tested and optimized in a variety of different settings. Because population cohorts and biobanking facilities are not prevalent in low- and middle-income countries, the development of health sciences and evidence-based cancer control may be affected. BCNet provides an opportunity for these countries to co- operate in a coordinated and effective manner to address local shortfalls in biobanking infrastructure.
A successful biobank is one that fosters successful science. Many current cohort studies are designed to investigate several diseases, with associated disease-specific working groups, as noted above in relation to EPIC. Compared with fragmented studies on a single risk determinant or a single predictive factor, the co- ordinated approach is necessary to evaluate whether different risk
determinants studied represent pri- mary risk factors or are the result of confounding. In the future, many dis- ease working groups will formalize their coordination by publishing open calls for proposals that describe the available study bases, with samples and data relating to different diseas- es, and invite scientists to apply for access.
Within the Biobanking and Bio- molecular Resources Research Infrastructure in the European Union, the concept of the biological expert centre has been developed. This centre offers an alternative to time-consuming and costly sam- ple shipments by providing sample analysis results rather than send- ing the samples to the investigator (Fig. 6.4.7). This initiative involves the availability of medical and scientific expertise in biobank-based research on issues such as proper sample selection, study design, and data interpretation [14]. At best, such an approach is optimally supported by in-house state-of-the-art analy- sis platforms that are so standard- ized and quality-assured that the results of analyses can be shared between centres and used for appli- cations that require stringent quality and documentation of results. It is envisaged that a more rapid, more
quality-assured, and less costly de- livery of the final data that the cus- tomer needs could further advance the central usefulness of biobanks for cancer research.
Conclusion The continuing quest to understand the causes of cancer and to develop global cancer control requires com- parisons between disease patterns in different parts of the world [15]. These research goals may be bet- ter achieved through different insti- tutions and countries endorsing the concept that samples and data for research should be openly acces- sible and that international – rather than national (or even subnational) – standards, rules, and procedures are to be preferred [16,17]. Promoting the adoption of common standards is desirable in high-resource countries as well as in low-resource countries. International collaboration to estab- lish essential research infrastruc- ture, including cohorts, biobanking facilities, and registry linkage sys- tems, is key to promote international cancer research, and is particularly relevant for cancer research in low- and middle-income countries.
560
References
1. Tunis SR, Benner J, McClellan M (2010). Comparative effectiveness research: policy context, methods development and research infrastructure. Stat Med, 29: 1963–1976. http://dx.doi.org/10.1002/sim. 3818 PMID:20564311
2. Maassen P, Stensaker B (2011). The knowledge triangle, European higher edu- cation policy logics and policy implementa- tion. High Educ, 61:757–769. http://dx.doi. org/10.1007/s10734-010-9360-4
3. Burgers JS, Fervers B, Haugh M et al. (2004). International assessment of the quality of clinical practice guide- lines in oncology using the Appraisal of Guidelines and Research and Evaluation Instrument. J Clin Oncol, 22:2000–2007. http://dx.doi.org/10.1200/JCO.2004.06.157 PMID:15143093
4. Wennberg JE, Fisher ES, Stukel TA et al. (2004). Use of hospitals, physician visits, and hospice care during last six months of life among cohorts loyal to highly re- spected hospitals in the United States. BMJ, 328:607. http://dx.doi.org/10.1136/ bmj.328.7440.607 PMID:15016692
5. Hakama M, Malila N, Dillner J (2012). Randomised health services studies. Int J Cancer, 131:2898–2902. http://dx.doi. org/10.1002/ijc.27561 PMID:22461063
6. Doll R, Peto R (1981). The causes of cancer: quantitative estimates of avoid- able risks of cancer in the United States today. J Natl Cancer Inst, 66:1191–1308. PMID:7017215
7. Wild CP, Scalbert A, Herceg Z (2013). Measuring the exposome: a powerful ba- sis for evaluating environmental exposures and cancer risk. Environ Mol Mutagen, 54:480– 499. http://dx.doi.org/10.1002/em. 21777 PMID:23681765
8. Gonzalez CA, Riboli E (2010). Diet and cancer prevention: contributions from the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Eur J Cancer, 46:2555–2562. http://dx.doi. org/10.1016/j.ejca.2010.07.025 PMID: 20843485
9. Riegman PH, de Jong BW, Llombart- Bosch A (2010). The Organization of European Cancer Institutes Pathobiology Working Group and its support of European biobanking infrastructures for transla- tional cancer research. Cancer Epidemiol Biomarkers Prev, 19:923–926. http:// dx.doi.org/10.1158/1055 -9965.EPI-10 - 0062 PMID:20332270
10. Pukkala E (2010). Nordic biological speci- men bank cohorts as basis for studies of cancer causes and control: quality control tools for study cohorts with more than two million sample donors and 130,000 pro- spective cancers. In: Dillner J, ed. Methods in Biobanking. New York: Springer, pp. 61–112.
11. Langseth H, Luostarinen T, Bray F, Dillner J (2010). Ensuring quality in studies linking cancer registries and biobanks. Acta Oncol, 49:368–377. http://dx.doi. org/10.3109/02841860903447069 PMID: 20059313
12. Dillner J, ed. (2009). The Janus Serum Bank – From Sample Collection to Cancer Research. Oslo: Cancer Registry of Norway. Available at http://www.kreftregis teret.no/Global/Publikasjoner%20og%20 rappor ter/CIN _ 20 08 _ Special _ Issue _ Janus_web.pdf.
13. Massett HA, Atkinson NL, Weber D et al. (2011). Assessing the need for a standard- ized cancer HUman Biobank (caHUB): findings from a national survey with cancer researchers. J Natl Cancer Inst Monogr, 2011:8–15. http://dx.doi.org/10.1093/jnci monographs/lgr007 PMID:21672890
14. Watson RW, Kay EW, Smith D (2010). Integrating biobanks: addressing the prac- tical and ethical issues to deliver a valuable tool for cancer research. Nat Rev Cancer, 10:646–651. http://dx.doi.org/10.1038/nrc 2913 PMID:20703251
15. Hudson TJ, Anderson W, Artez A et al.; International Cancer Genome Consortium (2010). International network of cancer genome projects. Nature, 464:993–998. http://dx.doi.org/10.1038/nature08987 PMID:20393554
16. Moore HM, Compton CC, Alper J, Vaught JB (2011). International approaches to advancing biospecimen science. Cancer Epidemiol Biomarkers Prev, 20:729–732. http://dx.doi.org/10.1158/1055-9965.EPI- 11-0021 PMID:21430299
17. Vaught JB, Henderson MK, Compton CC (2012). Biospecimens and bioreposito- ries: from afterthought to science. Cancer Epidemiol Biomarkers Prev, 21:253–255. http://dx.doi.org/10.1158/1055-9965.EPI- 11-1179 PMID:22313938
Cancer control experiences 561 6 C
O N
T R
O L
Cancer control in India: cancer care through a four-tier system G.K. Rath
In 2012, there were an estimated 1.01 million new cancer cases in India (age-standardized incidence, 94.0 per 100 000) and 980 000 can- cer-related deaths (age-standard- ized mortality, 64.5 per 100 000), and the 5-year prevalence was 1.79 million (proportion, 202.9 per 100 000) [1]. The cancer pattern is varied in different parts of the coun- try because of diverse lifestyles. The data from 27 population-based cancer registries in India show that the highest incidence of cancer is in Mizoram state, which is in the north-eastern part of India, com- pared with the lowest incidence from a rural registry in Barshi, in the western part of the country. There is an increasing trend of incidence rates for all malignancies except for cervical cancer, which has a down- ward trend.
The government has estab- lished the National Centre for Disease Informatics and Research to develop a national research da- tabase on cancer, diabetes, cardio- vascular diseases, and stroke [2]. India is one of the first countries to initiate a comprehensive national cancer control programme. There are 27 Regional Cancer Centres, including 339 radiotherapy facili-
ties equipped with 481 teletherapy machines. Most Regional Cancer Centres also have medical, surgi- cal, palliative care, imaging, and laboratory facilities. They are ac- tively involved in teaching and re- search as well.
The government has started a National Programme for Prevention and Control of Cancer, Diabe- tes, Cardiovascular Diseases, and Stroke. This has been done with the aim of combating the common risk factors for these diseases. The strategies under this programme include prevention through behav- iour change, early diagnosis, treat- ment, capacity building of human resources, surveillance, monitoring, and evaluation.
The cancer care network is en- visaged to be a four-tier system. Initially, the programme is being implemented in 100 districts across 21 states. Subsequently, this will be expanded to all 640 districts in the country. District hospitals are being strengthened for prevention, early detection, and management of com- mon cancers, especially oral, breast, and cervical cancers. Tertiary care centres will be the referral for the district hospitals and provide com- prehensive cancer care services.
Training will be provided at tertiary care centres/State Cancer Institutes for health-care professionals. State Cancer Institutes will be referral cen- tres for the tertiary care centres and district hospitals, and provide spe- cialized cancer care services. Three apex centres (National Cancer Institutes) will conduct research on the various malignancies that are common and pertinent to our part of the world, i.e. tobacco-related cancers, and cancers of the cervix, gall bladder, and liver. Besides work- ing as referral centres, the National Cancer Institutes will also provide training to generate high-quality hu- man resources. They will help the government in formulating national cancer control policies. State and district noncommunicable disease cells will be established in the se- lected states/districts for monitoring of the programme’s implementation.
References 1. Ferlay J et al. (2013). GLOBOCAN 2012
v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http:// globocan.iarc.fr.
2. Indian Council of Medical Research (2013). National Centre for Disease Informatics and Research – National Cancer Registry Programme. Available at http://www.ncdir india.org/.
562
Summary • The 2011 United Nations
General Assembly High-Level Meeting on the Prevention and Control of Noncommunicable Diseases presented the cancer community with an unprec- edented advocacy opportunity to position cancer as a global health and development issue and to use the momentum gen- erated at the meeting to press for increased efforts to reduce the global cancer burden, par- ticularly in low- and middle-in- come countries.
• In high-income countries, the role of advocacy in ensuring the provision of optimal clini- cal care and other benefits is established; advocacy in wider aspects of cancer control is be- ing increasingly specified as a key consideration.
• Comprehensive national cancer control plans, based on knowl- edge about the cancer burden provided by population-based cancer registries, are funda- mental to cancer prevention and control and should be the focus of cancer advocacy efforts.
• Advocacy is about achieving a change that delivers a desired
6.5 6 CONTROL
Advocacy for cancer control Cary Adams Rebecca Morton Doherty
Bob Chapman (reviewer) Scott Wittet (reviewer)
outcome. Advocacy requires a strategic plan, supported by a strong evidence base, emo- tional appeal, and a convincing financial case, and is most ef- fectively implemented in collab- oration with other like-minded organizations.
• Particular roles for advocacy for cancer control are being iden- tified both as contributing to progress in high-income coun- tries and as being developed in the context of local experi- ence in low- and middle-income countries.
Using advocacy to accelerate the fight against cancer Advocacy is about achieving a change that delivers a desired out- come. Health advocacy aims to raise political awareness and influence public policy decisions in support of an organization’s mission. In the context of cancer control, effective advocacy is required to create an engaged political environment con- ducive to improving the way cancer control knowledge influences policy and practice [1].
In high-income countries par- ticularly, the key role of advocacy in ensuring optimal clinical care is well recognized. Beyond contributing
to improving the circumstances of individual patients, such advocacy may serve, for example, to facilitate and encourage participation in clini- cal trials [2]. Indeed, a role in cancer research is recognized: cancer pa- tient advocates may contribute by offering perspectives different from those derived through government, academic, medical, and scientific approaches [3].
Complementing such perspec- tives, a role of cancer advocacy in cancer control in its broadest con- text, including all aspects of preven- tion, may be recognized. In this con- text, advocacy has a singular role. Thus, for example, in the prevention of human papillomavirus-related ma- lignancies, a field readily identified with major achievements, more con- certed advocacy is encouraged as complementing research and policy initiatives, if progress is to be opti- mal [4].
In September 2011, the United Nations General Assembly held a historic High-Level Meeting on the Prevention and Control of Noncommunicable Diseases. Only the second such health summit in the history of the United Nations, this meeting represented global rec- ognition of the growing burden of cancer and other noncommunicable diseases, including diabetes and car- diovascular and respiratory diseases,
Chapter 6.5 • Advocacy for cancer control 563 6 C
O N
T R
O L
C H
A P
T E
R 6
.5
and positioned cancer control as a global health and development im- perative. This provided the cancer community with an unprecedented advocacy opportunity to build on the momentum generated at the meeting and by the United Nations Political Declaration on noncommunicable diseases that was adopted [5], to press for increased efforts to reduce the cancer burden, particularly in low- and middle- income countries, where the cancer burden is set to rise by 81% by 2030 [6].
Advocacy for cancer control: what is wanted? Calling for national cancer control plans The ultimate goal of advocacy efforts in cancer control is threefold: (i) to ensure that governments develop comprehensive national cancer control plans that span the entire continuum of care; (ii) to ensure that such national plans are informed by population-based cancer registries that provide the necessary knowl- edge about cancer burden, as well as a means to evaluate the impact of implemented activities; and (iii) to ensure that governments finance
and implement interventions known to be cost-effective and productive.
Even in countries with an exist- ing cancer policy, plan, or strategy, these plans are not always sup- ported by the necessary funds, per- sonnel power, or infrastructure. In a recent survey, only 43% of low- income countries reported having operational national cancer control plans (see Chapter 6.2). Similarly, there are stark disparities in cancer registration between high-income and low- and middle-income coun- tries: the percentage of the popu- lation covered by cancer registries in the reference publication Cancer Incidence in Five Continents, Vol. X (2013) is 95% in North America and 42% in Europe but only 8% in Latin America, 6% in Asia, and 2% in Africa [7].
Granted circumstances in sub- Saharan Africa, approaches to mini- mize the cancer burden in the recent past have had little success, among other things because of a poor understanding of the potential for cancer prevention [8]. The situation may be addressed by initiatives in advocacy, research, workforce, care, and funding. Cancer advocates in Africa, for example, may look to
North America and Europe for guid- ance. However, lessons learned from high-income countries are not necessarily applicable, and relevant resources based on Africa are be- coming increasingly available [9].
Population-based cancer reg- istries and national cancer control plans are a vital investment in un- derstanding and responding to the cancer burden in all countries. A well-conceived plan that is based on cancer incidence, prevalence, and survival rate data provided by cancer registries and that outlines evidence-based strategies for pre- vention, early detection, diagnosis, treatment, and palliation can signifi- cantly lower the number of cancer cases and improve the quality of life of cancer patients. In the absence of population-based cancer registries, cancer planners can nevertheless use available data to outline strate- gies for implementing proven cost- effective interventions for reducing the cancer burden. Given the cur- rent economic climate, it is crucial to maximize the population impact of money spent. Investing in the collec- tion of basic cancer information and the development of comprehensive national cancer control plans needs
Supporting national cancer control planning: a toolkit for Civil SoCiety organiSationS (CSoS)
geneva, 2012
CollaBoration
aCtion
Content
evalUation
BUrDen
Fig. 6.5.2. Cover of an advocacy toolkit. This publication from the Union for Inter- national Cancer Control provides support to civil society organizations.
Fig. 6.5.1. The United Nations General Assembly High-Level Meeting on the Prevention and Control of Noncommunicable Diseases, held in September 2011, was only the second such health summit in the history of the United Nations.
564
to be clearly specified to ensure that the greatest impact is achieved for the funds allocated.
There are cost-effective evi- dence-based interventions that can significantly reduce the cancer bur- den. These include screening for cervical cancer and breast cancer, and vaccination against hepatitis B virus and human papillomavirus to protect against infection-related liver cancer and cervical cancer, respec- tively. Cervical cancer screening and hepatitis B virus immunization were underscored as priority interventions in the 2011 United Nations Political Declaration on noncommunicable diseases [5] and have also been highlighted by WHO as not only high- ly cost-effective but also feasible and appropriate to implement within the constraints of health systems in low- and middle-income countries [10].
What does success look like? Advocacy at a global level was key to adoption of the WHO Framework Convention on Tobacco Control, the first treaty negotiated under the aus- pices of WHO. The Convention pro- vides a robust framework to confront the efforts of the tobacco industry, which employs lawyers and market- ing and communications experts to counter health advocates around the globe. The recent major achievement involving Australia’s adoption of legis- lation for plain packaging of tobacco products highlights the impact that a persistent, united, and vocal civil society may have.
Similarly, the NCD Alliance (www. ncdalliance.org), a global network of more than 2000 civil society organi- zations from 170 countries address- ing the four main noncommunicable diseases (cardiovascular disease, diabetes, cancer, and chronic re- spiratory disease), played a pivotal role in the lead-up to the United Nations General Assembly High- Level Meeting on the Prevention and Control of Noncommunicable Diseases, pressing governments to recognize that noncommunicable diseases are a global development priority requiring an urgent response. Since then, the NCD Alliance has
sustained global action on noncom- municable diseases and spearhead- ed remarkable progress, culminating in the adoption of an omnibus resolu- tion on noncommunicable diseases at the 66th World Health Assembly, which will benefit the millions of peo- ple worldwide who are at risk of, or liv- ing with, noncommunicable diseases.
Advocacy for cancer control: how do we do it? Effective advocacy for cancer con- trol requires a carefully constructed strategic plan identifying the prob- lem, advocacy goal, key messages and target audiences, activity plan, and evaluation process [1], and is optimally supported by three core components: a comprehensive evi- dence base, emotional appeal, and a strong financial case. Neglecting both the views of patients and their loved ones and the compelling case for investing in cancer preven- tion and control now will inevitably result in a rise in the cancer burden and require a significantly greater in- vestment later. In addition to these three core components, effective ad- vocacy for cancer control needs to occur at the local, national, regional, and global levels. These efforts can be enhanced through strong part- nerships between local and global advocates, as well as across groups concerned with related diseases, specifically including other noncom- municable diseases, and involv- ing multiple sectors, including the private sector. Thus, in identifying breast cancer control strategies in Asia, Latin America, and the Middle East, a specific role for advocacy was identified [11]. The development of ef- fective strategies may be grounded in the experience of local practition- ers, policy-makers, and advocacy leaders.
Evidence-based advocacy Evidence is critical for influencing public policy decisions; it provides the necessary information for ad- vocates to raise awareness of the nature and extent of a problem and make realistic projections of possible
policy outcomes. Screening for cervi- cal cancer is a good example of the links between research, policy, and practice. Over the previous decades, advocacy has helped to drive the re- search agenda, culminating in large- scale randomized clinical trials to test the “screen and treat” approach using visual inspection with acetic acid. The evidence base generated through these trials, particularly in low-income settings, has resulted in widespread recognition of cervi- cal cancer screening as a proven
Fig. 6.5.3. This model of a cigarette, an inflated balloon, was displayed in Hyder- abad, India, during World No Tobacco Day 2012. World No Tobacco Day is marked on 31 May each year, to create awareness about the harmful effects of tobacco and smoking.
Chapter 6.5 • Advocacy for cancer control 565 6 C
O N
T R
O L
C H
A P
T E
R 6
.5
cost-effective intervention for reduc- ing the global cancer burden, and has enabled the rapid integration of an in- dicator for cervical cancer screening into the Global Monitoring Framework for the Prevention and Control of Noncommunicable Diseases [12] fi- nalized by United Nations Member States in November 2012. Cervical Cancer Action, a global coalition of organizations including the American Cancer Society, PATH (an interna- tional non-profit organization), the Pan American Health Organization, and the Union for International Cancer Control, has been a driving force behind these successful advo- cacy efforts [13].
The power of the patient voice Cancer patients, survivors, and their families are not only key to docu- menting the impact of policies on individuals; these same individuals can also be highly effective and pow- erful advocates for the adoption of cancer prevention and control poli- cies. Patient participation in cancer advocacy is often hindered by the lack of long-term survivors among those diagnosed with particular ma- lignancies, but as medical advances result in increasing survival rates such situations may change, as high- lighted by the rising cohesive body of lung cancer advocacy groups. A good example of this is the Global Lung Cancer Coalition, established in 2001, which now comprises 28 nongovernmental patient organiza- tions from more than 20 countries, and works to improve disease out- comes for lung cancer patients glob- ally [14].
Case study: access to opioids for cancer pain relief Bringing the patient voice to policy discussions, and emphasizing the need for a rights-based approach to address the global inequities in access to pain relief, has been instrumental in effecting policy changes at the highest level. Since 2009, the Union for International Cancer Control has worked with its members to add the cancer voice
to the human rights and palliative care communities’ efforts towards increased access to opioid analge- sics. During this period, the United Nations’s drug policy-making organ, the Commission on Narcotic Drugs, has adopted two resolutions [15,16], emphasizing that more needs to be done to enhance availability of con- trolled medications while prevent- ing their diversion and abuse. The 2011 Commission on Narcotic Drugs resolution specifically requested the United Nations Office on Drugs and Crime, in consultation with the International Narcotics Control Board and WHO, to update its model laws to ensure that they reflect an appropriate balance between ensur- ing adequate access to internation- ally controlled drugs and preventing diversion and abuse.
In 2012, commendable advoca- cy efforts led by the palliative care community resulted in the inclusion of an indicator on access to pal- liative care, assessed by morphine- equivalent consumption of strong opioid analgesics per cancer death, in the Global Monitoring Framework for the Prevention and Control of Noncommunicable Diseases [12].
Building the financial case The World Economic Forum has identified noncommunicable dis- eases, including cancer, as the sec- ond greatest risk to global economic
growth [17]. One half of those who die from noncommunicable diseases are in their most productive years, meaning that the social costs and economic consequences in terms of lost productivity are consider- able. The cost of cancer alone is estimated to reach US$ 458 billion in 2030 [10], yet WHO estimates that a basic package of cost-effective strategies to address the common cancer risk factors (tobacco use, alcohol consumption, unhealthy diet, and physical inactivity) would cost only US$ 2 billion a year [18]. Even so, less than 3% (US$ 503 mil- lion out of US$ 22 billion) of overall development assistance for health was allocated to noncommunicable diseases in 2007 [19], and only 5% of global spending on cancer is in developing countries. This is despite the fact that nearly 80% of the pre- ventable deaths from these diseases occur in developing countries, with this percentage set to rise [20]. The financial case for investing in cancer is strong, and is one that can and must be used by cancer advocates at all levels to support adequate and sustained resourcing of national cancer control plans.
Multisectoral partnerships The United Nations Political Decla- ration on noncommunicable diseas- es clearly articulated the need for multisectoral partnerships, engaging
Fig. 6.5.4. A nurse provides pain relief to a cancer patient in Uganda.
566
both health and non-health actors, including civil society and the private sector, to promote and support the provision of services for noncommu- nicable disease prevention and con- trol. In addition to bolstering global and national advocacy efforts, such partnerships are essential for the im- plementation of cancer interventions at the country level. Given today’s fi- nancial climate, engagement of parts of the private sector, with appropriate safeguards to manage potential con- flicts of interest, is more critical than ever. There is a clear willingness in the private sector to engage at this level; according to a recent survey conducted by Business for Social
Responsibility, 40% of companies expect to increase their commit- ment to global health partnerships focused on noncommunicable dis- eases in the next 5 years [21]. Global health partnerships could play an important role in improving primary health-care systems, which are the front lines – particularly in low- and middle-income countries – for en- gaging communities with prevention, diagnosis, and treatment across a range of diseases, including cancer.
Advocacy for cancer control: the road ahead In May 2013, Member States at the 66th World Health Assembly adopted
an omnibus resolution on noncom- municable diseases that combined major decisions and recommen- dations in a single comprehensive resolution, including (i) adopting a Global Monitoring Framework for the Prevention and Control of Noncommunicable Diseases; (ii) endorsing a Global Action Plan for Noncommunicable Diseases 2013– 2020; and (iii) agreeing to establish a Global Coordination Mechanism for Noncommunicable Diseases [22]. This emerging framework for non- communicable diseases will carve out a new global advocacy space for the cancer community, and an opportunity to ensure that noncom- municable diseases, including can- cer, continue to occupy a place on the global health and development agenda. Ensuring that cancer is part of the 2013 Millennium Development Goal review and the emerging de- bate on universal health coverage, the sustainable development goals, and other development issues will also be vital to ensuring that cancer control remains central to future thinking. Cancer advocates face both new opportunities and chal- lenges as “latecomers” to the devel- opment discourse. Now more than ever, innovative partnerships that go beyond traditional health groups and embrace partners in the develop- ment sphere, including reproductive, maternal, and child health organiza- tions as well as the AIDS community, are critical if the currently predicted cancer burden for future generations is to be reduced.
Fig. 6.5.5. Cancer Council Australia provides a range of resources related to evidence- based cancer control policy and advocacy on its website.
Chapter 6.5 • Advocacy for cancer control 567 6 C
O N
T R
O L
C H
A P
T E
R 6
.5
References
1. Godfrey E et al. (2012). Cancer Advocacy Training Toolkit for Africa. Africa Oxford Cancer Foundation (AfrOx), African Organisation for Research and Training in Cancer (AORTIC), European Society for Medical Oncology (ESMO), and Union for International Cancer Control (UICC). Available at http://www.uicc.org/advocacy/ advocacy-resources/additional-resources.
2. Katz ML, Archer LE, Peppercorn JM et al. (2012). Patient advocates’ role in clinical trials: perspectives from Cancer and Leukemia Group B investigators and advocates. Cancer, 118:4801– 4805. ht t p: //dx.do i.o r g /10.10 0 2 /c nc r. 274 8 5 PMID:22392584
3. Collyar D (2005). How have patient advo- cates in the United States benefited can- cer research? Nat Rev Cancer, 5:73–78. http://dx.doi.org/10.1038/nrc1530 PMID: 15630417
4. Franco EL, de Sanjosé S, Broker TR et al. (2012). Human papillomavirus and cancer prevention: gaps in knowledge and pros- pects for research, policy, and advocacy. Vaccine, 30 Suppl 5:F175–F182. http:// dx.doi.org/10.1016/j.vaccine.2012.06.092 PMID:23199961
5. United Nations (2011). Political Declaration of the High-Level Meeting of the General Assembly on the Prevention and Control of Non-communicable Diseases. New York: United Nations. Available at www. who.int/nmh/events/un_ncd_summit2011/ political_declaration_en.pdf.
6. WHO (2011). Global Status Report on Noncommunicable Diseases 2010. Geneva: WHO. Available at http://www.who.int/nmh/ publications/ncd_report2010/en/.
7. Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
8. Morhason-Bello IO, Odedina F, Rebbeck TR et al. (2013). Challenges and oppor- tunities in cancer control in Africa: a per- spective from the African Organisation for Research and Training in Cancer. Lancet Oncol, 14:e142–e151. http://dx.doi. org /10.1016 /S1470 -20 4 5(12)70 4 82- 5 PMID:23561745
9. Odedina FT, Rodrigues B, Raja P (2013). Setting the stage for cancer advocacy in Africa: how? Infect Agent Cancer, 8 Suppl 1:S6. http://dx.doi.org/10.1186/1750-9378- 8-S1-S6 PMID:23902653
10. Bloom DE, Cafiero ET, Jané-Llopis E et al. (2011). The Global Economic Burden of Noncommunicable Diseases. Geneva: World Economic Forum. Available at www. weforum.org/EconomicsOfNCD.
11. Bridges JF, Anderson BO, Buzaid AC et al. (2011). Identifying important breast cancer control strategies in Asia, Latin America and the Middle East/North Africa. BMC Health Serv Res, 11:227. http://dx. doi.org/10.1186/1472-6963-11-227 PMID: 21933435
12. WHO (2012). Report of the Formal Meeting of Member States to conclude the work on the comprehensive global monitoring framework, including indicators, and a set of voluntary global targets for the preven- tion and control of noncommunicable dis- eases (A/NCD/2). Available at http://apps. who.int/gb/ncds/pdf/A _NCD_2-en.pdf.
13. Cervical Cancer Action (2012). A Global Coalition to Stop Cervical Cancer. Available at http://www.cervicalcancer action.org/about/about.php.
14. Global Lung Cancer Coalition (2012) Available at http://www.lungcancer coalition.org/en.
15. The Commission on Narcotic Drugs (2010). Promoting adequate availability of interna- tionally controlled licit drugs for medical and scientific purposes while preventing their diversion and abuse, Resolution 53/4. Available at http://www.unodc.org/ documents/commissions/CND-Res-2000- until-present/CND53_4e.pdf.
16. The Commission on Narcotic Drugs (2011). Promoting adequate availability of internationally controlled narcotic drugs and psychotropic substances for medical and scientific purposes while preventing their diversion and abuse, Resolution 54/6. Available at http://www.unodc.org/ documents/commissions/CND-Res-2011 to2019/CND54_6e1.pdf.
17. Global Risk Network of the World Economic Forum (2010). Global Risks 2010: A Global Risk Network Report. Geneva: World Economic Forum. Available at http://www3.weforum.org/docs/WEF_ GlobalRisks_Report_2010.pdf.
18. WHO (2011). Scaling Up Action Against Noncommunicable Diseases: How Much Will It Cost? Geneva: WHO. Available at http://www.who.int/nmh/publications/ cost_of_inaction/en/.
19. Nugent RA, Feigl AB (2010). Where Have All the Donors Gone? Scarce Donor Funding for Non-Communicable Diseases. Center for Global Development Working Paper 228. Available at http://www.cgdev. org/publication/where-have-all-donors- gone-scarce-donor-funding-non-communi cable-diseases-working-paper.
20. Knaul FM, Frenk J, Shulman L; for the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries (2011). Closing the Cancer Divide: A Blueprint to Expand Access in Low and Middle Income Countries. Boston, MA: Harvard Global Equity Initiative.
21. Little M, Schappert J (2012). Working to- ward Transformational Health Partnerships in Low- and Middle-Income Countries. Business for Social Responsibility. Available at https://www.bsr.org/en/our- insights/report-view/working-toward-trans formational-health-partnerships.
22. Sixty-sixth World Health Assembly (2013). Resolution WHA66.10. Follow-up to the Political Declaration of the High- level Meeting of the General Assembly on the Prevention and Control of Non- communicable Diseases. Available at http://apps.who.int/gb/e/e_wha66.html.
Websites The NCD Alliance. Putting non-communicable diseases on the global agenda: www.ncdalliance.org
Union for International Cancer Control: www.uicc.org
568
Cancer control in Jordan: goals for low- and middle-income countries Omar Nimri
In Jordan, a country with about 6.5 million people, cancer is the second most frequent cause of death, after heart disease. A total of 6820 new cancer cases were registered by the national Jordan Cancer Registry in 2010. Of these, 72% were among Jordanians and 28% were among non-Jordanians.
The rank order of the five most common cancers affecting Jordanians is: breast cancer (19.6% of all cases), colorectal cancer (11.5%), lymphoma (7.9%), lung cancer (7.8%), and prostate cancer (4.5%). The crude annual incidence rate of all cancers among Jordanians is 79.4 per 100 000 (male, 74.0; female, 85.1). The most frequent paediatric cancers (0–14 years) are leukaemias (38%), brain and central nervous system tumours (27%), lym- phomas (20%), renal tumours (8%), and soft tissue cancers (7%) [1].
Growing and ageing populations are projected to experience dramatic increases in the number of cancer cases and cancer deaths, particular- ly in low- and middle-income coun- tries like Jordan. It is imperative that planning begins now to deal not only with those cancers already occur- ring but also with the larger numbers expected in the future.
A national cancer control plan (NCCP) for Jordan is in the process of being finalized. A simple NCCP existed previously, but the idea of a full, up-to-date NCCP really started in October 2008 at a meeting at- tended by many parties, including the Jordanian Ministry of Health, the WHO office in Amman, the King
Hussein Cancer Center, some non- governmental organizations, and the King Hussein Institute for Cancer Research and Biotechnology). A draft NCCP was prepared and pre- sented to the minister of health, but unfortunately the draft was not en- dorsed and the NCCP was shelved.
In January 2012, the International Atomic Energy Agency, upon offi- cial request of the Ministry of Health, conducted a mission in close col- laboration with the ministry. Based on the findings and recommenda- tions of that mission, a new NCCP is now in preparation. The major partners are the Ministry of Health, the King Hussein Cancer Center, the Jordanian Royal Medical Services, the private medical sec- tor, WHO, nongovernmental or- ganizations such as the Jordan Medical Association and the Jordan Oncology Society, academia and research institutes, smoking cessa- tion programmes, and others. It is hoped that the new, updated NCCP will cover all aspects of cancer, in- cluding prevention, early detection, screening, diagnosis and treatment, palliative care, and rehabilitation.
Most people pick up their bad habits at a younger age, leading to unhealthy diet, smoking, poor hygiene, and many other risk fac- tors. The majority of cancers are preventable, by reducing exposure to factors that are known to cause the disease. Hence, education of the youth should be a priority to decrease these risks in the context of cancer control.
In 2001, a national tobacco control programme was started. In 2004, Jordan was the second coun- try in the local region and the 29th in the world to endorse the WHO Framework Convention on Tobacco Control. The prevalence of smoking among adults (18 years and older) was 29% (male, 51%; female, 7%) in 2007. The prevalence among youth (13–15 years) was 11.5% (male, 17.4%; female, 8.3%) in 2009 [2]. In 2008, a law was passed banning smoking in public places; enforce- ment of this smoke-free legislation is still needed. Also, a repulsive picture of lungs and smoking is required on all cigarette packaging, along with warning statements.
Given that breast cancer is the most common cancer in Jordan, a national programme for breast cancer awareness, early detection, and screening was initiated in 2006, led by the King Hussein Cancer Foundation and Center.
The whole spectrum of cancer treatment is available in Jordan, but the inequality of access is the major obstacle faced.
Palliative care services remain grossly inadequate and represent a further area of priority.
References 1. Jordanian Ministry of Health. Annual
Incidence of cancer in Jordan. Available at http://www.moh.gov.jo/EN/Pages/Publi cations.aspx.
2. WHO (2009). Global Youth Tobacco Survey: Country Fact Sheet – Jordan. Available at http://www.emro.who.int/tobac co/gtss-youth-survey/gyts-factsheets- reports.html.
Chapter 6.6 • Law in cancer control 569 6 C
O N
T R
O L
C H
A P
T E
R 6
.6
Summary • The effective application of law
is essential to cancer control and requires engagement across such diverse areas as trade law, intellectual property law, invest- ment law, human rights law, drug control law, constitutional law, consumer protection law, neg- ligence law, medical law, and criminal law.
• Current challenges to Australia’s world-first plain tobacco pack- aging legislation have become seminal cases on the inter- section between global public health and trade and investment laws and norms, with profound implications for cancer control.
• Overly restrictive laws continue to operate as barriers to the availability of opioid analgesics for the relief of cancer-associ- ated pain. Significant norma- tive progress has recently been made at the international level.
• Efforts to address the unavoid- able tensions between the pub- lic interests in privacy protection and cancer research should ac- knowledge the role of the right to health, understood as a col- lective right of populations, the
6.6 6 CONTROL
Law in cancer control Jonathan Liberman Marianne Hammer (reviewer)
Roger Magnusson (reviewer) Anne Lise Ryel (reviewer)
fulfilment of which depends on high-quality research.
Law is everywhere in cancer control. Some examples of areas are: • the regulation of the behaviour of
the tobacco, alcohol, food, asbes- tos, and sunbed industries;
• the regulation of where people can use products that cause harm to third parties, such as laws prohibit- ing smoking in workplaces, public places, residential facilities, and cars carrying children;
• court challenges to cancer preven- tion measures such as tobacco product display bans (Norway), graphic health warnings (Uruguay and the USA), and plain packaging (Australia);
• litigation against industries whose products and practices cause cancer;
• the application of occupational health and safety laws to exposure to carcinogens in the workplace;
• the relationship between patent law and pharmaceutical research and the affordability of medicines;
• the effect of laws regulating the trade, prescription, and supply of opioid analgesics on their avail- ability for the treatment of cancer- associated pain;
• the regulation of health profes- sionals, including the protection
of patients from inappropriate or dangerous “treatments”;
• the regulation of decision-making in treatment and care, including the end-of-life period;
• the conduct of clinical trials and the relationship between epide- miological research and health information privacy;
• the access of family members to genetic information; and
• intellectual property rights over genetic material and scientific discovery.
In these and a host of other areas, the effective application of law is es- sential to cancer control.
For each of these areas, law has both domestic and interna- tional elements. Domestically, law is constituted by a combination of constitutional arrangements, legis- lation, regulations, court decisions, regulatory practices and policies, and professional and community un- derstandings of what the law is and how it operates in practice. At the international level, law includes trea- ties – notably the WHO Framework Convention on Tobacco Control (FCTC); the United Nations Single Convention on Narcotic Drugs; the International Covenant on Economic, Social, and Cultural Rights, which enshrines the right to the highest attainable standard of health; other international human rights treaties;
570
the World Trade Organization (WTO) agreements; and a raft of regional, plurilateral, and bilateral trade and investment treaties – as well as a wide range of “softer” normative instruments, such as declarations, guidelines, and strategies.
Health law and governance – at both domestic and international levels – is an increasingly crowded and fragmented space. The United Nations Political Declaration on the Prevention and Control of Non- communicable Diseases [1] under- lines that prevention and control of these diseases require multisectoral approaches, including across such sectors as health, education, energy, agriculture, sports, transport, com- munication, urban planning, environ- ment, labour, employment, industry and trade, finance, and social and economic development.
Internationally, WHO is the United Nations’s specialized agen- cy for health, but it is only one of a large number of international agencies whose work directly or indirectly affects health, includ- ing cancer, such as the United Nations Children’s Fund, the Food and Agriculture Organization of the United Nations, the United Nations Environment Programme, the United Nations Development Programme, the United Nations Population Fund, the World Bank, the United Nations
Human Rights Council, the WTO, the United Nations Office on Drugs and Crime, the International Atomic Energy Agency, and the International Labour Organization. These agen- cies have very different mandates, priorities, instruments, and values, meaning that policy and operational coherence in the interests of health is often elusive.
Against this broader backdrop, this chapter focuses on three topi- cal and important areas, relating to prevention, treatment, and research, respectively.
Trade and investment challenges to tobacco control The foreword to the WHO FCTC [2] describes the treaty as having been “developed in response to the glob- alization of the tobacco epidemic”, the spread of which “is facilitated through a variety of complex factors with cross-border effects, including trade liberalization and direct foreign investment”. The WHO FCTC was conceived, in part, as a global re- sponse to these dynamics.
Increasingly, the relationship between the WHO FCTC, as an in- ternational legal instrument, and the international legal instruments that govern trade, intellectual property, and investment is being played out in the form of legal challenges [3].
As this Report goes to press, chal- lenges to Australia’s world-first “plain packaging” laws are the most promi- nent example.
In November 2011, the Australian Parliament enacted the Tobacco Plain Packaging Act 2011 [4]. As of 1 December 2012, all tobacco prod- ucts sold in Australia were required to comply with the legislation. The legislation bans the use of logos, brand imagery, symbols, other im- ages, colours, and promotional text on tobacco products and tobacco product packaging, and requires packaging to be a standard drab dark-brown colour in matte finish. Products are differentiated by the brand and product name, displayed in a standard colour, position, font size, and style. Graphic health warn- ings are required on 75% of the front surface and 90% of the back surface of tobacco packaging. Indeed, the “plain” in “plain packaging” is some- what of a misnomer.
Plain packaging is supported by two sets of guidelines adopted by the governing body of the WHO FCTC – its Conference of the Parties – in November 2008: one on packaging and labelling (Article 11) [5] and the other on tobacco advertising, pro- motion, and sponsorship (Article 13) [6]. The guidelines for the implemen- tation of Article 11 state that plain packaging “may increase the notice- ability and effectiveness of health warnings and messages, prevent the package from detracting atten- tion from them, and address industry design techniques that may suggest that some products are less harmful than others”. The Article 13 guide- lines acknowledge that packaging “is an important element of advertising and promotion”. The Australian leg- islation includes as one of its objects “to give effect to certain obligations that Australia has as a party to the Convention on Tobacco Control”.
Plain packaging has been chal- lenged both in the WTO and under a bilateral investment treaty between Australia and Hong Kong Special Administrative Region. In the WTO, in which dispute settlement is state– state rather than investor–state,
Fig. 6.6.1. Health warning images and plain packaging designs used in Australia. All brand names are specified in a standard colour, position, font size, and style.
Chapter 6.6 • Law in cancer control 571 6 C
O N
T R
O L
C H
A P
T E
R 6
.6
Ukraine, Honduras, the Dominican Republic, Cuba, and Indonesia have claimed that plain packaging breaches Australia’s obligations under the General Agreement on Tariffs and Trade, the Agreement on Technical Barriers to Trade, and the Agreement on Trade-Related Aspects of Intellectual Property Rights. It has been reported that the two multinational tobacco com- panies, Philip Morris and British American Tobacco, have been pro- viding support to countries to chal- lenge plain packaging [7].
Unlike in the WTO, the bilat- eral investment treaty between Australia and Hong Kong Special Administrative Region – like many other investment treaties – enables a foreign investor to sue a govern- ment directly. Philip Morris Asia – which acquired its interest in Philip Morris Australia on 23 February 2011, i.e. after plain packaging had been announced by the Australian Government, in April 2010 – has brought proceedings against the Australian Government claiming “ex- propriation” of its investment and de- nial of “fair and equitable treatment”.
The Australian Government is vigorously defending both the WTO and bilateral investment treaty pro- ceedings. These challenges to plain packaging have become seminal cases on the intersection between global public health and trade and investment laws and norms, touch- ing on broader issues relating to the power of governments to regulate in the public interest, the relationship between private intellectual prop- erty rights and public health, and the nature and role of evidence in policy-making.
Achieving balance in the regulation of opioids Opioid analgesics are regulated inter- nationally under the United Nations Single Convention on Narcotic Drugs, 1961, as amended by its 1972 Protocol [8]. The Single Convention seeks to strike a balance between the two aims of ensuring the availability of opioids for medical use for the relief of pain
and suffering – which it describes as “indispensable” – and preventing their diversion and abuse. In practice, this balance has not been achieved, with poor availability in much of the world and great global disparities. In 2011, Australia, Canada, Japan, New Zealand, the USA, and some European countries accounted for more than 93% of total global mor- phine consumption [9].
The major barriers to the avail- ability of opioid analgesics are well known. They include insufficient training of health-care profession- als in the prescription and admin- istration of opioids; misperceptions among health-care professionals, policy-makers, and patients and their families about the safety of opioids; exaggerated fears about the development of dependence; overly restrictive laws and regula- tions that exceed the requirements of the Single Convention; and unduly harsh sanctions for the unintention- al mishandling of opioids by health workers [10].
The Single Convention estab- lishes a regulatory system that includes requirements for: govern- ment authorization for participa- tion in the trade and distribution of narcotics; provision by states to the International Narcotics Control Board of estimates of need, and statistical returns covering produc- tion, manufacture, consumption, im- port, and export; export and import licences for international transac- tions; licensing and record-keeping for trade or distribution; and medical prescriptions for dispensation.
Many states have adopted overly restrictive laws and regulations that exceed the requirements of the Single Convention and that operate as barriers to availability. Examples include limitations on who can pre- scribe, special prescription proce- dures, restrictions on doses that may be prescribed or on the number of days’ supply that may be provided in a single prescription, arbitrary restrictions on the number of phar- macies that are allowed to dispense opioid medications, and overly bur- densome administrative and bureau- cratic procedures.
The lack of balance in legal and regulatory approaches to opioid analgesics is increasingly being ad- dressed at the international level. Several important documents have been developed and adopted over the past few years: • two resolutions of the United
Nations’s main drug policy-mak- ing organ, the Commission on Narcotic Drugs, in 2010 [11] and 2011 [12], with co-sponsorship by countries from different regions and with different levels of income, which have emphasized that more needs to be done to enhance the availability of controlled medica- tions while preventing their diver- sion and abuse;
• new WHO policy guidelines (2011) [13], which provide guidance on policies and legislation with re- spect to availability, accessibility, affordability, and control of con- trolled medicines;
• a special supplement to the 2010 International Narcotics Control
Fig. 6.6.2. Disparity in morphine consumption in high-income countries (High) versus low- and middle-income countries (LMIC). (A) Morphine consumption (as a percentage of the total number of kilograms consumed) in 2010. (B) Total population of countries reporting opioid consumption to the International Narcotics Control Board in 2010.
A B
572
Board Annual Report [14], which includes recommendations on availability of controlled medicines and their appropriate use, national control systems, and prevention of diversion and abuse;
• a 2011 United Nations Office on Drugs and Crime discussion paper [15], which includes recommen- dations relating to data collec- tion, laws and policies, and public awareness.
The 2011 Commission on Narcotic Drugs resolution requested the United Nations Office on Drugs and Crime, in consultation with the International Narcotics Control Board and WHO, to update its model laws to ensure that they reflect an appropriate balance between en- suring adequate access to interna- tionally controlled drugs and pre- venting diversion and abuse. At the 2013 session of the Commission on Narcotic Drugs, the United Nations Office on Drugs and Crime released proposed amendments to its model drug laws [16], including: • the introduction of substantive
objects into the model law (rather than simply compliance with the
relevant conventions): “ensuring the availability of narcotic drugs and psychotropic substances for medical and scientific use, and preventing the non-medical and non-scientific use of narcotic drugs and psychotropic substances”.
• the use of neutral terminology to describe drugs that have medical or scientific uses but can cause harm when misused. For exam- ple, the proposals describe the Schedule in which morphine is listed as including “drugs and substances having a medical and/ or scientific use which should be subject to control in view of the harms that their non-medical and/ or non-scientific use can cause”. This contrasts with the previous wording, which referred to “strict- ly controlled substances and plants having a medical use” and “substances with a high potential risk to public health but having a medical use” (emphasis added). Under the previous formulation, concern about the risk of mis- use predominated over the need to ensure appropriate medical treatment.
• an explicit recognition that, as rec- ommended by WHO, the capacity to prescribe controlled medicines should not be restricted to a small number of medical specialties (such as oncologists); controlled medicines should be available at all appropriate levels of care, and the amount of medicine pre- scribed, the appropriate formula- tion, and the duration of treatment should be for the health practition- er to decide, based on individual patient needs and sound scientific medical guidance.
Privacy and cancer research High-quality research – clinical, behavioural, and epidemiological – is key to effective cancer control. Epidemiological research provides a wealth of information that is es- sential for planning, resource alloca- tion, programme development and delivery, legal and regulatory inter- ventions, therapy development, and care improvement.
In practice, such research can- not be effectively conducted with- out access to sometimes sensitive
Fig. 6.6.3. Marked differences between countries in relation to medicinal opioid consumption (2010), assessed by morphine-equivalent consumption (in milligrams) of strong opioid analgesics per cancer death.
Chapter 6.6 • Law in cancer control 573 6 C
O N
T R
O L
C H
A P
T E
R 6
.6
personal health information, the dis- closure of which can have a range of consequences for individuals. Public health research and surveillance cannot be effectively performed on the basis of fully de-identified data. De-identified data are not sufficient for the individual–data linkages re- quired for epidemiological research designed to investigate more intri- cate interactions between exposure, lifestyle, preventive interventions, treatment, and outcomes [17]. Even after research has been conduct- ed, data may need to remain re- identifiable to allow for the correc- tion of errors or updating of new data, and for the review of scientific work, thereby guarding against sci- entific fraud [18].
It is not feasible to condition the use of personal health information in public health research on individual consent in all cases. When informa- tion is collected, it is seldom possi- ble to identify all its future valuable uses. When dealing with large data sets, it may not be feasible to obtain consent from each individual whose data is proposed to be used [19]. Further, relying on individual consent can lead to compromised or invalid results because of significant differ- ences between individuals who do and do not grant consent, meaning that research might be conducted on the basis of unrepresentative data [20].
The effect of these research re- alities is an unavoidable tension be- tween the public interests in privacy protection and cancer research. Many cancer (and other public health) researchers believe that regulation is increasingly privileging privacy protection to the detriment of the conduct of research. One of the striking features of the literature on the relationship between information privacy and public health research – both academic articles and legal and policy documents – is that only one species of right, the right to privacy, seems to feature. There is invariably an “interest” against which the right should be balanced, often expressed as public health. While the point is often made that the “public” consists
of individuals, who are the ultimate beneficiaries of public health re- search [21], it is curious that the right to health seems never to be raised.
This is the case notwithstanding the recognition that fulfilment of the right to health depends on the con- duct of research. The International Covenant on Economic, Social, and Cultural Rights is the primary inter- national legal instrument on eco- nomic, social, and cultural rights. The United Nations Committee on Economic, Social, and Cultural Rights’s General Comment on Article 12 (the right of everyone to the enjoy- ment of the highest attainable stan- dard of physical and mental health) states that one of the “core obliga- tions” of the article is “to adopt and implement a national public health strategy and plan of action, on the basis of epidemiological evidence, addressing the health concerns of the whole population” (emphasis added) [22]. The control of diseases required by Article 12.2(c) necessitates “us- ing and improving epidemiological surveillance and data collection on a disaggregated basis”.
Bringing the right to health to bear on matters relating to health informa- tion privacy need not entail a disre- gard, or a downgrading, of the right to privacy or the values it promotes. Rather, it would situate debates in
their proper legal context, and should ultimately influence the way in which public health research is regulated. Too often, public health research ap- pears as an afterthought exception in regimes aimed at strengthening privacy protection, often driven by concerns about such matters as pub- lication of information about celebri- ties, online security, and identity theft. Acknowledging the role that the right to health has to play here touches on broader debates about the nature of the right to health and the need to conceive of it as not just an individual right but a collective right of popula- tions [23].
Conclusion Cancer control requires expertise and capacity in law. But there is no single discipline of “cancer control law”. Cancer control law requires en- gagement at both domestic and in- ternational levels and across such di- verse areas as trade law, intellectual property law, investment law, human rights law, drug control law, constitu- tional law, consumer protection law, negligence law, medical law, and criminal law. The need for capac- ity in cancer control law continues to grow. The global cancer control community has begun to recognize and respond to this challenge.
Fig. 6.6.4. An opium poppy field in Tasmania, Australia, where opium is grown for me- dicinal morphine. The opium poppies are usually genetically modified so that extraction of opiate can be undertaken only in a laboratory.
574
References
1. United Nations (2011). Political Declaration of the High-Level Meeting of the General Assembly on the Prevention and Control of Non-communicable Diseases. New York: United Nations. Available at www. who.int/nmh/events/un_ncd_summit2011/ political_declaration_en.pdf.
2. WHO (2003). WHO Framework Convention on Tobacco Control. Opened for signature 16 June 2003, 2305 UNTS 166 (entered into force 27 February 2005). Geneva: WHO. Available at http://www.who.int/fctc/ text_download/en/index.html.
3. WHO (2012). Confronting the Tobacco Epidemic in a New Era of Trade and Investment Liberalization. Geneva: WHO Tobacco Free Initiative. Available at http:// www.who.int/tobacco/publications/industry/ trade/confronting_tob_epidemic/en/.
4. Australian Government (2011). Tobacco Plain Packaging Act 2011. Available at http://www.comlaw.gov.au/Details/C2011 A00148.
5. The Conference of the Parties to the WHO FCTC (2008). Guidelines for Implementation of Article 11 of the WHO Framework Convention on Tobacco Control on Packaging and Labelling of Tobacco Products, Decision FCTC/ COP3(10). Available at http://www.who. int/fctc/guidelines/adopted/article_11/en/.
6. The Conference of the Parties to the WHO FCTC (2008). Guidelines for Implementation of Article 13 of the WHO Framework Convention on Tobacco Control on Tobacco Advertising, Promotion and Sponsorship, Decision FCTC/COP3(12). Available at http://www.who.int/fctc/guidelines/adopted/ article_13/en/.
7. Thompson C (2012). Big Tobacco backs Australian law opposers. Financial Times, 29 April 2012.
8. United Nations (1961). Single Convention on Narcotic Drugs, 1961, as amended by the 1972 Protocol amending the Single Convention on Narcotic Drugs, 1961. Opened for signature 25 March 1972, 520 UNTS 204 (entered into force 8 August 1975). New York: United Nations. Available at https://www.unodc.org/unodc/en/treaties/ single-convention.html.
9. International Narcotics Control Board (2012). Estimated World Requirements for 2013 – Statistics for 2011. New York: United Nations. Available at http://www.incb.org/ incb/en/narcotic-drugs/Technical_Reports/ narcotic_drugs_reports.html.
10. Hogerzeil HV, Liberman J, Wirtz VJ et al.; Lancet NCD Action Group (2013). Promotion of access to essential medicines for non-communicable diseases: practical implications of the UN political declara- tion. Lancet, 381:680– 689. http://dx.doi. org /10.1016/S0140 - 673 6(12) 62128 -X PMID:23410612
11. The Commission on Narcotic Drugs (2010). Promoting adequate availability of interna- tionally controlled licit drugs for medical and scientific purposes while prevent- ing their diversion and abuse, Resolution 53/4. Available at http://www.unodc.org/ documents/commissions/CND-Res-2000- until-present/CND53_4e.pdf.
12. The Commission on Narcotic Drugs (2011). Promoting adequate availability of inter- nationally controlled narcotic drugs and psychotropic substances for medical and scientific purposes while preventing their diversion and abuse, Resolution 54/6. Available at http://www.unodc.org/docu ments/commissions/CND-Res-2011to2019/ CND54_6e1.pdf.
13. WHO (2011). Ensuring Balance in National Policies on Controlled Substances: Guidance for Availability and Accessibility of Controlled Medicines. Geneva: WHO. Available at http://www.who.int/medicines/ areas/quality _safety/guide_nocp_sanend/ en/.
14. International Narcotics Control Board (2011). Report of the International Narcotics Control Board on the Availability of Internationally Controlled Drugs: Ensuring Adequate Access for Medical and Scientific Purposes. Available at http://www.incb.org/incb/en/ p ub li c at i o n s /a n nu a l - r e p o r t s /a n nu a l - report.html.
15. United Nations Office on Drugs and Crime (2011). Ensuring Availability of Controlled Medications for the Relief of Pain and Preventing Diversion and Abuse: Striking the Right Balance to Achieve the Optimal Public Health Outcome. Vienna: United Nations Office on Drugs and Crime. Available at www.unodc.org/docs/treat- ment/Pain/Ensuring_availability _of_con trolled_medications_FINAL_15_March_ CND_version.pdf.
16. United Nations Office on Drugs and Crime (2011). Revision of parts of the model law related to availability and accessibility to controlled drugs for medical purposes. Vienna: United Nations Office on Drugs and Crime.
17. Hakulinen T, Arbyn M, Brewster DH et al. (2011). Harmonization may be counter- productive – at least for parts of Europe where public health research operates ef- fectively. Eur J Public Health, 21:686–687. http://dx.doi.org/10.1093/eurpub/ckr149 PMID:22080476
18. Stenbeck M, Gissler M, Haraldsdóttir S et al. (2011). The planned changes of the European Data Protection Directive may pose a threat to important health re- search, Open Letter to European Decision Makers. Available at http://ki.se/content/1/ c6/13/68/01/Appendix3_Open%20letter% 20to%20decision%20makers%20data%20 protection%20directive.pdf.
19. Xafis V, Thomson C, Braunack-Mayer AJ et al. (2011). Legal impediments to data linkage. J Law Med, 19:300–315. PMID: 22320005
20. Coleman MP, Evans BG, Barrett G (2003). Confidentiality and the public in- terest in medical research – will we ever get it right? Clin Med, 3:219–228. http:// dx.doi.org/10.7861/clinmedicine.3-3-219 PMID:12848254
21. Institute of Medicine (2009). Beyond the HIPAA Privacy Rule: Enhancing Privacy, Improving Health through Research. Washington, DC: National Academies Press. Available at http://www.iom.edu/ Reports/2009/beyond-the-HIPAA-Privacy- Rul e - Enh anc in g - Pr i vac y - I m p r ov in g - Health-Through-Research.aspx.
22. United Nations Committee on Economic, Social, and Cultural Rights (2000). The Right to the Highest Attainable Standard of Health (Article 12 of the International Covenant on Economic, Social, and Cultural Rights), General Comment No. 14. Available at http://www.ohchr.org/EN/ HRBodies/CESCR/Pages/CESCRIndex. aspx.
23. Meier BM, Mori LM (2005). The highest attainable standard: advancing a collec- tive human right to public health. Columbia Human Rights Law Rev, 37:101.
Website
McCabe Centre for Law and Cancer: http://www.mccabecentre.org/
Cancer control experiences 575 6 C
O N
T R
O L
Cancer control in Morocco: action in harmony with the socioeconomic and cultural context Rachid Bekkali
Cancer control in Morocco is a na- tional priority. Created in November 2005, the Lalla Salma Foundation for Cancer Prevention and Treatment has conducted, in collaboration with the Moroccan Ministry of Health, 15 studies to analyse the situation re- garding all aspects related to cancer. The findings were the basis for the development of the National Cancer Prevention and Control Plan 2010– 2019 (NCPCP), ratified by the gov- ernment in March 2010. The vision of the NCPCP is to control cancers nationwide through a multisectoral approach, proposing concrete and sustainable actions and making the best use of available resources, while being in harmony with the socioeconomic and cultural con- text. The values of the NCPCP are equity, solidarity, quality, and excel- lence. The objective of the NCPCP is to reduce cancer morbidity and mortality, and to improve the quality of life of patients and their relatives.
The NCPCP action strategy con- sists of 78 operational measures to be undertaken in four strategic
fields: prevention, early detection, di- agnosis and treatment, and palliative care. These strategic components are supported by the activities of communication, social mobilization, regulation, training, and research. All these components are being integrated around a national and international mobilization and thus constitute the conceptual framework of the NCPCP. Since the creation of the Lalla Salma Foundation, several projects have been accomplished in partnership with the Ministry of Health, national and internation- al nongovernmental organizations, governmental and private agencies, and benefactors. As of the end of 2012, of the 78 measures of the NCPCP, 72 measures had been initiated, among which 51 were well advanced.
In the area of prevention, public awareness campaigns are organized every year, a programme of smoke- free schools and businesses has been implemented, a human papil- lomavirus vaccination programme is planned to be generalized in 2014,
and the strengthening of regulations is ongoing. In the field of health care, several oncology centres were built and equipped, as well as “Houses of Life” to accommodate patients and their families during the treatment period. A programme of access to medication for low-income patients has been established. Human skills have been developed. Therefore, between 2006 and the end of 2012, for example, Morocco improved from two oncology centres to nine, from two accelerators to 22, from one “House of Life” to nine, from 50 on- cologists to more than 150, and from 11 500 patients treated to 23 000. In addition, a detection programme for breast and cervical cancers has been implemented, a palliative care network project was initiated, and an ambitious plan to structure can- cer research in Morocco has been started.
The situation analysis reports, the NCPCP, and more details about cancer control activities in Morocco can be found at http://www.contrele cancer.ma/en/documents/.
576
Summary • The 14 million estimated new
cases of cancer worldwide an- nually inflict a crushing burden of economic costs and human suffering. A significant part of this burden could be avoided by expanding coverage of preven- tion, early detection, and treat- ment for specific cancers.
• The estimated total annual eco- nomic cost of cancer was ap- proximately US$ 1.16 trillion in 2010 – the equivalent of more than 2% of total global gross domestic product. Even this im- pressively high figure is a lower bound, as it does not include the substantial longer-term costs to families and caregivers.
• Between one third and one half of cancer deaths could be avoided with prevention, early detection, and treatment – be- tween 2.4 million and 3.7 mil- lion avoidable deaths per year, 80% of which occur in low- and middle-income countries.
• Investing strategically in cancer care and control more than pays for itself. A reasonable estimate shows that the world could have saved between US$ 100 billion and US$ 200 billion in 2010 by
6.7 6 CONTROL
The global economic burden of cancer Felicia Marie Knaul Héctor Arreola-Ornelas Oscar Méndez Marcella Alsan Janice Seinfeld Andrew Marx Rifat Atun
Hideyuki Akaza (reviewer)
investing in prevention, early de- tection, and effective treatment of cancer.
• The ability to prevent, detect, and treat many cancers has improved over time, and many of these advances have led to reductions in costs. Harnessing markets and increasing access can also bring down prices.
Human life and well-being have an intrinsic and immeasurable value. They also have economic value. Viewing health as an investment, rather than a cost, is the philosophy that today inspires human, econom- ic, and environmental global devel- opment agendas.
Yet this investment philosophy remains largely ignored in formulat- ing global and national policies to deal with cancer and other chronic illnesses, leaving wide open the agenda for research into the eco- nomics of cancer. Impressive oppor- tunities exist to develop and expand the knowledge needed to better understand how to reduce the eco- nomic burden of this set of diseases.
This chapter highlights compel- ling economic arguments for invest- ments to increase global access to cancer care and control, drawing heavily on published research and
analysis of the Global Task Force on Expanded Access to Cancer Care and Control [1]. Estimates of the eco- nomic value of the avoidable cancer burden are provided, and these are compared with potential savings based on the current costs of cancer care and control. These estimates are illustrative of how economics can contribute to a deeper understanding of the global cancer burden and to more effective strategies for reduc- ing it.
The economics of investing in cancer care and control The estimated 14 million new cases of cancer worldwide annually lead to enormous economic cost as well as incalculable human suffering [2,3]. The economic consequences of each cancer case include the di- rect and indirect costs of treatment, the income forgone by patients and their families unable to work during periods of treatment and illness, and, most importantly in economic terms, the productivity lost due to premature death, disability, and suf- fering. Further, catastrophic health spending undermines the economic stability of families, often generating dynamic losses and forcing families into impoverishment and economic ruin that may extend far beyond the course of the disease. Broader es- timates of economic consequences
Chapter 6.7 • The global economic burden of cancer 577 6 C
O N
T R
O L
C H
A P
T E
R 6
.7
also, and appropriately, take into ac- count the perceived costs of human suffering. While a full estimate of the economic burden of cancer would take into account each of these cost components, several of them are dif- ficult to calculate with available data.
A significant part of the economic and health burden of cancer could be avoided by expanding cover- age of prevention, early detection, and treatment for specific cancers. Indeed, it is reasonable to consider that the additional financial invest- ments in cancer care and control would be more than counterbal- anced by reductions in the economic toll caused by the disease. This hy- pothesis is analysed below.
The economic cost of cancer A first approximation of the total annual economic cost of cancer is US$ 1.16 trillion, which is more than 2% of total global gross domestic product. This figure is the sum of the costs of prevention and treatment, plus the annual economic value of disability-adjusted life years (DALYs) lost (a measure that combines years of life lost due to premature mortality
and years of healthy life lost due to disability) as a result of cancer.
The estimate is derived as fol- lows: First, the value of lost produc- tivity due to premature death and disability is taken from a study of DALYs for 17 categories of cancer covering all cancer sites, which pro- duced an estimate of US$ 921 billion for 2010 [3,4]. Second, based on an- other recent study, the annual global economic cost of treating new (inci- dent) cancer cases was assumed to be US$ 310 billion for 2010 [2,5,6]. Of this, 53% (US$ 163 billion) is due to medical costs, and 23% to the time of caregivers and the cost of transportation to treatment facili- ties. The remaining 24% is attributed to productivity losses from time in treatment and associated disability; since this portion is also covered in the estimate of DALYs, it was not in- cluded in the summation. Finally, the cost of prevention is assumed to be 7% of the cost of treatment. This is a rough estimate that corresponds to the ratio of prevention to health spending in Canada [5,7]. Applied to the global treatment estimate of US$ 163 billion, the cost of preven- tion of cancer is assumed to be just more than US$ 11.4 billion.
Although impressively high, the figure of US$ 1.16 trillion underes- timates the total annual economic cost of cancer, for several reasons. The most important factor is lack of data on the substantial longer-term costs to families and caregivers, which often extend well beyond the first year of treatment. The figure also fails to account for the value that patients and families place on human suffering, which may be far higher than the productivity losses measured by DALYs.
An alternative method for esti- mating the cost of cancer is to use a value of a statistical life (VSL) ap- proach. This methodology, which includes substantially more of the variety of costs that are incurred by patients and their families, attempts to account for the value individuals themselves place on lost income, out-of-pocket spending on health, and pain and suffering. Based on an- other recent study, the total 2010 VSL estimate of the global cost of cancer is US$ 2.5 trillion. Of this, close to US$ 1.7 trillion is attributed to high- income countries and the remaining US$ 800 billion to low- and middle- income countries (LMICs) [8].
Working from these estimates of the economic costs of cancer, it is possible to calculate the potential return on investments in expanding cancer care and control by determin- ing values for two other key factors: the economic value of lives that could be saved by expanding ac- cess to prevention, early detection, and treatment; and the costs of the expansion of cancer care and con- trol that would be required to achieve these gains.
The “avoidable” cancer burden A significant proportion of the cancer burden could be avoided through prevention, early detection, and treatment. Calculating the propor- tion of years of healthy, productive life that could be saved and their value to both the economy and the individual is the key to estimating the economic benefit of investing in cancer care and control.
Fig. 6.7.1. A nurse delivers chemotherapy to a patient with Kaposi sarcoma at Neno District Hospital in Malawi. Kaposi sarcoma is among the most common cancers in Malawi, but it is highly treatable – chemotherapy can control Kaposi sarcoma cancer growth and help patients return to normal lives.
578
These calculations require an assumption about the proportion of deaths that can be avoided. The literature on avoidable deaths has typically established premature death using an empirical approach to set an upper limit – for example, 64 years. Under this scenario, a death that oc- curred before the age of 65 years (or any other upper limit) is considered avoidable. The selection of cancers that are considered either prevent- able or treatable or both preventable and treatable (with or without earlier detection that might have resulted in either a cure or a significant increase in healthy life expectancy) is based on earlier research [9–15].
Each life expectancy scenario is applied to countries’ income-group- specific GLOBOCAN estimates of mortality and age at death by can- cer type [16,17]. Using a minimum standard of 65 years of life expec- tancy, an estimated 32% of deaths can be avoided. Setting the standard for life expectancy at the level of the best-performing countries in each income region, 36% of deaths could be avoided. Finally, using life expec- tancy of 75 years as the standard, an estimated 49% of cancer deaths
are considered avoidable with pre- vention, early detection, and/or treatment.
These estimates suggest, re- spectively for each scenario, that there are 2.4, 2.7, and 3.7 million avoidable deaths from cancer per year. LMICs account for approxi- mately 80% of this avoidable mortal- ity in each life expectancy scenario.
Costs of treatment and prevention Another factor that affects the cal- culations of returns on investments is the dynamic nature of the costs of treatment and prevention. The estimate of the costs of treatment presented above and used as the baseline here – US$ 310 billion – does not account for the possibil- ity of more effective primary and secondary prevention becoming available.
Several of the cancers that gen- erate significant global investment in treatment are preventable, either by reducing exposure to risk factors, such as tobacco, or by vaccination, as for cervical cancer. Preventing the majority of these cancers means avoiding a considerable proportion
of treatment costs. For several can- cers with a high burden in LMICs – including Kaposi sarcoma and can- cers of the cervix, liver, and (most importantly in terms of burden) lung – a considerable proportion of cases could be prevented or detected in precancerous stages, avoiding the far higher costs of treatment. A 90% reduction in cases (through prevention) for Kaposi sarcoma and cancers of the cervix, liver, and lung implies a reduction of at least 20% in the total estimated costs of treating cancer – or approxi- mately US$ 65 billion per year [18]. Prevention can also extend to other cancers with a lower overall burden, such as head and neck cancers, as well as stomach cancers (by treat- ing Helicobacter pylori) [5].
Hence, with more effective pre- vention, the global cost of treatment for cancer would be significantly less than what is currently being spent: rather than the estimated US$ 310 billion, it would be ap- proximately US$ 246 billion. Adding the estimated cost of prevention (US$ 11.4 billion) produces an over- all figure of about US$ 257 billion [19].
Fig. 6.7.2. Cancer patients participate in an art activity at the Advanced Paediatric Centre of the Post Graduate Institute of Medical Education and Research in Chandigarh, India. The activity was part of events to mark International Childhood Cancer Day.
Fig. 6.7.3. In China, participants in a clini- cal trial for liver cancer prevention wait to have their blood drawn.
Chapter 6.7 • The global economic burden of cancer 579 6 C
O N
T R
O L
C H
A P
T E
R 6
.7
Estimating the return on investments in expanding cancer care and control The estimates in Table 6.7.1 com- pare the economic value of lives saved in DALYs and VSL to the total costs of treatment and prevention. These estimates provide approxi- mations of what the world could have saved in 2010 by investing in cancer care and control. They range from the most optimistic re- turns of US$ 230 billion and almost US$ 1 trillion (US$ 940 billion) in terms of DALYs and VSL [18], re- spectively, to the lower bounds of US$ 10 billion and US$ 531 billion [20], respectively.
A longer-term view Taking a longer-term, dynamic view of costs and benefits of treatment for cancer provides a more complete vision of the costs and benefits of investing in expanded access to cancer care and control. Existing studies are scarce; this section is restricted to a discussion of issues, and further research in this area is recommended.
Estimates of the total value of lost output from cancer, based on
macroeconomic modelling for 2011 to 2030 – an approach that differs from the ones presented above by taking into account accumulated losses over time – suggest substan- tially higher cumulative economic losses of US$ 2.9 trillion for LMICs and US$ 5.4 trillion for high-income countries [8]. These macroeconomic models show that between 2011 and 2030, noncommunicable diseases – including cancer, cardiovascular dis- ease, chronic respiratory disease, diabetes, and mental health – repre- sent a global cumulative output loss of up to US$ 47 trillion [8].
A further, and significant, con- sideration in modelling the costs and benefits of investing in cancer care and control is that the dimen- sions and boundaries of prevention and treatment change over time. Cancers that are considered largely untreatable (liver cancer), or costly to treat (cervical cancer), can now be prevented by applying relatively new medical technologies and vaccines. As science develops, cancers of in- fectious origin that tend to be more common in LMICs will become in- creasingly susceptible to prevention, and the global costs of treatment can
be expected to decline [1]. Hence, estimates of future global costs of cancer treatment may be overstated as science progresses and identifies new options for prevention that are less costly than treating the cancers that primarily affect people in LMICs.
The cost of producing and de- livering drugs and vaccines can fall over time, especially as innova- tive global financing platforms are developed and implemented [21]. This has been amply demonstrated by the experience with antiretroviral drugs for HIV/AIDS and by huge re- ductions in the cost of the hepatitis B virus vaccine, achieved in part through the GAVI Alliance, a pub- lic–private partnership committed to increasing access to immunization in LMICs. A notable example is the recent GAVI-spurred drop of more than 95% in the price of the human papillomavirus vaccine, from the US$ 130 per dose that prevails in high-income countries to less than US$ 5 per dose in the lowest-income countries. Earlier, the Pan American Health Organization Revolving Fund had garnered an almost 90% reduc- tion, to US$ 14 per dose. Although the vaccine is still unaffordable for
Table 6.7.1. Sensitivity analysis of economic returns to investing in cancer care and control
Cost of care and controla Economic cost of cancer (billions of US dollars)
DALYs: valued at US$ 921 billion
VSL less OOP: valued at US$ 2.37 trillion
Avoidable deaths Avoidable deaths
49%b 36%c 49%b 36%c
Scenario 1. Assuming full cost of treatment based on Bloom et al. (2011) [8] + cost of prevention: 310 + 11 = US$ 321 billion
130 10 839 531
Scenario 2. Scenario 1 with reduced costs of treatment based on preventing 90% of Kaposi sarcoma and cancers of the cervix, liver, and lung: (310 – 64) + 11 = US$ 257 billion
194 75 904 596
Scenario 3. Scenario 1 with reduced costs of treatment based on preventing 90% of all potentially preventable cancers: (310 – 100) + 11 = US$ 221 billiond
230 110 940 632
DALYs, disability-adjusted life-years; OOP, out-of-pocket health spending by families; VSL, value of a statistical life. a Each cell equals: [(economic cost of cancer) * (% mortality avoided with treatment or prevention)] – (medical and non-medical costs of treating new cancer cases + costs of prevention). b 49% of cancer mortality is assumed avoidable using a scenario of achieving the levels of the best-performing countries – the social justice approach. c 36% of cancer mortality is assumed avoidable using a scenario of achieving the levels of the best-performing country in each income region. d 90% reduction in incidence and hence treatment costs for Kaposi sarcoma and cancers of the cervix, larynx, liver, lung, nasopharynx, other pharynx, and stomach.
580
Towards understanding the economics of cancer: priority areas for research Felicia Marie Knaul
Research in health economics – especially towards a better under- standing of the global distribution of costs and the most effective oppor- tunities to achieve impacts – could greatly contribute to reducing both the economic and the health burden of cancer. Yet, economic analysis of cancer is a nascent area of re- search that to date has focused on high-income countries.
Based on an initial review of the literature, there has never been a comprehensive body of work on the economics of the global cancer bur- den. Core analysis is, however, be- ing undertaken for an upcoming vol- ume as part of the third round of the Disease Control Priorities project. This is a follow-up volume to the second edition of Disease Control Priorities in Developing Countries (published in 2006), which includes several chapters applying econom- ics to cancer care and control [1–3].
Although overall work on the economics of cancer is sparse, a noteworthy exception is the litera- ture on risk factors (or risky be- haviour), and especially tobacco consumption and its control [4–8]. Indeed, the existing literature about risky behaviour speaks to the rich array of techniques that econom- ics, and especially health econom- ics, can contribute to reducing the global burden of cancer and closing the cancer divide between rich and poor.
To fully understand the econom- ics of cancer, it will be necessary to analyse each component of the cancer care and control continuum: primary prevention, secondary prevention, diagnostics, treatment, survivorship care, pain control, and palliative care [9]. This is particu- larly important because cancer covers a complex set of diseases: some are associated with infection or behaviour; others are largely de- termined by stage at detection; and
yet another, overlapping, subset is susceptible to treatment.
There are a host of opportuni- ties for research on the econom- ics of cancer, based on the field of health economics – an area that has blossomed over the past two to three decades [10]. The research should consider both economics of health (determinants of cancer outcomes) and economics of health care (the market for the provision of cancer care, including global health and health systems), including: • the economics of global cancer
and the relationship between the disease and economic develop- ment, using dynamic analysis of the impact on economic growth and the spillover effects on hu- man and social development and the “costs of inaction” [2];
• the analysis of current and fu- ture demand for cancer care and control globally, to project global cancer care spending growth un- der different scenarios for prices and investment in prevention and treatment;
• determinants of cancer incidence and outcomes, the effect of pov- erty and socioeconomic status, and the social and economic de- terminants of access to and use of health services;
• spending patterns by families over time on cancer care and control and under different incen- tive and insurance scenarios, and over time on long-term care;
• analysis of the supply of cancer care and control services glob- ally, and the role of incentives for expanding access, with quality, at the least possible cost and with greatest equity, considering all aspects of service provision and delivery: pharmaceuticals, pri- mary care-based services, hos- pital-based services, the health- care workforce, technologies, and so on;
• intellectual property and patent markets for drugs and vaccines, and the role of innovative financ- ing platforms such as the GAVI Alliance [11];
• the economics of pain control and palliation, which spans the global regulatory environment and eco- nomics of access;
• the medical workforce and the optimal supply of human re- sources to respond to current and future needs for cancer care and control;
• health system financing, covering innovations that would optimize the incentive structure for cancer care and control in different re- source settings;
• health system reform analysis, including the design, evalua- tion, and implementation of pro- grammes and policies; and
• other research that could both benefit from and contribute to some of the more novel areas of current health systems work – such as the diagonal approach [9,12,13] and delivery science – and avoid single-intervention analysis by considering value chains to analyse complex pro- cesses and diseases and by using the diagonal approach to merge horizontal and vertical interventions to facilitate syner- gistic investments that generate system-wide improvements [14].
Applying techniques of health economics, and integrating them with novel frameworks in health systems research, would generate new knowledge especially applica- ble to improving the capacity to re- spond to the global cancer burden with more appropriate strategies for lower-income settings. Yet to truly improve our understanding of the economic burden of cancer globally and identify the most appropriate ways to face this burden, a multi- disciplinary approach is required that links economists to a myriad of
Chapter 6.7 • The global economic burden of cancer 581 6 C
O N
T R
O L
C H
A P
T E
R 6
.7
many countries, this price reduction marks a huge step forward and was accomplished in only half a decade [22].
Further, the costs of care – and hopes for cure – for several preva- lent cancers like breast, colorectal, and cervical cancers depend on the stage at which they are diagnosed. Thus, investing in earlier detection, much of which depends on devel- oping and implementing innovative delivery models, can greatly reduce the cost per year of life saved. At the same time, population-wide screen- ing for cancer can be very costly, making it a priority to develop inno- vations for earlier detection, many of which will be especially important in LMIC settings.
Many of the dynamic aspects of improving cancer care and control – such as global innovations in financ- ing platforms and country-level inno- vations in early detection – are part of developing a science of delivery around cancer [23].
Research in health economics, largely unexplored to date, could inform policy and provide recom- mendations that would help reduce
the financial and human burden of cancer. Establishing economic in- centives that help change behaviour and increasing the efficacy of ser- vices are two avenues (see “Towards understanding the economics of cancer: priority areas for research”).
Conclusions Health is an investment, rather than a cost. Yet this idea has not suffi- ciently permeated the discussions on cancer care and control. Planning for cancer prevention and manage- ment must be integrated in a for- ward-looking manner into all policy agendas to achieve the most effec- tive investment of health spending. Adopting an investment approach to health reshapes human, economic, social, and sustainable development agendas.
Chronic disease, including can- cer, is a leading global economic risk. The drain of tobacco alone on the global economy exceeds the to- tal annual expenditure on health of all LMICs. This makes investment in both prevention and treatment a priority for health and for economic development.
Between one third and one half of cancer deaths can be avoided with prevention, early detection, and treatment. Thus, between 2.4 mil- lion and 3.7 million deaths are avoidable per year, 80% of which occur in LMICs.
Given the huge and avoidable suffering caused by cancer, meet- ing the unmet need for cancer care and control in LMICs is a moral im- perative. In addition, and from an economic standpoint, expanding prevention, detection, and treat- ment of cancer yields benefits that exceed the costs.
The total annual economic cost of cancer – not including longer-term costs to families and caregivers – was approximately US$ 1.16 trillion in 2010, which amounts to more than 2% of total global gross domestic product. By contrast, investing in cancer care and control yields a positive annual return on prevention and treatment because of the number of deaths that are potentially avoid- able. Global economic savings of at least US$ 100–200 billion could be achieved by avoiding deaths, and
other disciplines in health and even in other sectors. The challenge ahead is to both increase econom- ics research and integrate it into efforts to better understand cancer globally.
References 1. Brown ML et al. (2006). Health service in-
terventions for cancer control in develop- ing countries. In: Jamison DT et al., eds. Disease Control Priorities in Developing Countries, 2nd ed. New York: Oxford University Press, pp. 569–589.
2. Anand S et al. (2012). The Cost of Inaction: Case Studies from Rwanda and Angola. Cambridge, MA: Harvard University Press.
3. Foley KM et al. (2006). Pain control for people with cancer and AIDS. In: Jamison DT et al., eds. Disease Control Priorities in Developing Countries, 2nd ed. New York: Oxford University Press, pp. 981–993.
4. Jha P, Chaloupka FJ (2000). BMJ, 321:358–361. http://dx.doi.org/10.1136/ bmj.321.7257.358 PMID:10926598
5. Jha P, Chaloupka FJ, eds (1999). Curbing the Epidemic: Governments and the Economics of Tobacco Control. Washington, DC: World Bank.
6. World Bank (2003). The Economics of Tobacco Use and Tobacco Control in the Developing World. Washington, DC: World Bank. Available at http://ec.europa.eu/ health/archive/ph_determinants/life_style/ tobacco/documents/world_bank_en.pdf.
7. WHO (2011). Scaling Up Action Against Noncommunicable Diseases: How Much Will It Cost? Geneva: WHO. Available at http://www.who.int/nmh/publications/ cost_of_inaction/en/.
8. Lee K (2008). PLoS Med, 5:e189. http:// dx.doi.org/10.1371/journal.pmed.0050189 PMID:18798688
9. Knaul FM et al. (2012). Health system strengthening and cancer: a diagonal response to the challenge of chronic- ity. In: Knaul FM et al., eds. Closing the Cancer Divide: An Equity Imperative. Based on the work of the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries. Cambridge, MA: Harvard Global Equity Initiative, pp. 95–122.
10. Pauly MV et al. (2012). Handbook of Health Economics, Vol. 2. Oxford: North Holland.
11. Atun R et al. (2012). Lancet, 380:2044– 2049. http://dx.doi.org/10.1016/S0140- 6736(12)61460-3 PMID:23102585
12. Sepúlveda J et al. (2006). Lancet, 368:2017–2027. http://dx.doi.org/10.1016/ S0140-6736(06)69569-X PMID:17141709
13. Frenk J (2006). Lancet, 368:954–961. http://dx.doi.org/10.1016/S0140-6736(06) 69376-8 PMID:16962886
14. Kim JY et al. (2013). Lancet, 382:1060– 1069. http://dx.doi.org/10.1016/S0140- 6736(13)61047-8 PMID:23697823
582
hence lost healthy years of pro- ductive life, through treatment and prevention. Taking into account the human cost of suffering – the value that individuals place on reduced suffering and illness – the savings are at least US$ 500 billion and could reach almost US$ 950 billion.
Yet, economic analysis of the global cancer burden is a nascent field, and most research to date has focused on high-income countries. Research in this area could great- ly contribute to reducing both the economic and health burden. The agenda for this research is ample, and one of the main limiting factors is the lack of adequate data. Indeed,
better data are required to promote a better understanding of the econom- ics of cancer, and any initiative to collect these data could and should be linked to existing initiatives to im- prove global cancer registries [24].
To truly improve our understand- ing of the economic burden of cancer globally, a multidisciplinary approach is required that links economists to a myriad of other disciplines. Future efforts should seek to increase economics research and also to in- tegrate it into efforts to better under- stand cancer globally.
Planning for the future requires harnessing markets in ways that can stimulate innovation, encourage
savings that reduce prices, promote investments that generate system- wide improvements that reduce the impact of cancer and also accrue to other diseases, spread benefits, and reduce costs. These economic benefits could be much greater if the potential cost savings from innova- tive delivery and financing, com- bined with more equitable pricing of drugs and other therapies, could be achieved. Adopting a diagonal approach using care delivery value chains can ensure that the potential synergies and shared benefits of in- terventions are maximized and fully taken into account.
A future where prevention, early detection, and treatment become more accessible to patients and health systems in LMICs is one that builds on the “economics of hope”. Neither the costs of prevention nor the potential benefits of extending cancer care and control should be taken as fixed, given the many op- portunities that exist to increase ac- cess and reduce the global burden of cancer.
This chapter is adapted from Chapter 3 in Knaul FM, Gralow JR, Atun R, Bhadelia A, eds (2012). Closing the Cancer Divide: An Equity Imperative. Based on the work of the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries. Cambridge, MA: Harvard Global Equity Initiative. Distributed by Harvard University Press. © 2012 by President and Fellows of Harvard College acting through the Harvard Global Equity Initiative. Reprinted by permis- sion of the Harvard Global Equity Initiative.
Fig. 6.7.4. A female adolescent receives an intramuscular immunization from a nurse. The availability of vaccines to prevent cervical cancer represents one of the most im- mediate advances in cancer prevention achieved in the past decade.
Chapter 6.7 • The global economic burden of cancer 583 6 C
O N
T R
O L
C H
A P
T E
R 6
.7
References
1. Knaul FM, Arreola-Ornelas H, Atun R et al. (2012). Investing in cancer care and control. In: Knaul FM, Gralow JR, Atun R, Bhadelia A, eds. Closing the Cancer Divide: An Equity Imperative. Based on the work of the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries. Cambridge, MA: Harvard Global Equity Initiative, pp. 71–91.
2. Beaulieu N, Bloom D, Bloom R, Stein R (2009). Breakaway: The Global Burden of Cancer – Challenges and Opportunities. A Report from the Economist Intelligence Unit. London: The Economist Group.
3. John RM, Ross H (2008). Economic value of disability-adjusted life years lost to can- cers. Available at http://media.marketwire. com/attachments/EZIR/627/18192_Final JournalManuscript.pdf.
4. Shafey O, Eriksen M, Ross H, Mackay J (2009). The Tobacco Atlas, 3rd ed. Atlanta, GA: American Cancer Society. Available at http://www.tobaccoatlas.org/.
5. Nikolic IA, Stanciole AE, Zaydman M (2011). Chronic Emergency: Why NCDs Matter. Health, Nutrition and Population (HNP) Discussion Paper. Washington, DC: World Bank.
6. Chand S (2012). Silent Killer, Economic Opportunity: Rethinking Non-Communicable Disease. Centre on Global Health Security, Briefing Paper. Available at http://www. chathamhouse.org/publications/papers/ view/181471.
7. OECD (2010). OECD Stat Extracts data- base. Paris: Organisation for Economic Co-operation and Development. Available at http://stats.oecd.org/Index.aspx.
8. Bloom DE, Cafiero ET, Jané-Llopis E et al. (2011). The Global Economic Burden of Non-communicable Diseases. Geneva: World Economic Forum. Available at www. weforum.org/EconomicsOfNCD.
9. Gispert R, Serra I, Barés MA et al. (2008). The impact of avoidable mortality on life expectancy at birth in Spain: changes between three periods, from 1987 to 2001. J Epidemiol Community Health, 62:783–789. http://dx.doi.org/10.1136/jech. 2007.066027 PMID:18701727
10. Gómez-Arias RD, Bonmatí AN, Pereyra- Zamora P et al. (2009). Design and compar- ative analysis of an inventory of avoidable mortality indicators specific to health condi- tions in Colombia [in Spanish]. Rev Panam Salud Publica, 26:385–397. http://dx.doi. org/10.1590/S1020-49892009001100002 PMID:20107689
11. Humblet PC, Lagasse R, Levêque A (2000). Trends in Belgian premature avoidable deaths over a 20 year period. J Epidemiol Community Health, 54:687–691. http://dx.doi.org/10.1136/jech.54.9.687 PMID:10942448
12. Weisz D, Gusmano MK, Rodwin VG, Neuberg LG (2008). Population health and the health system: a comparative analy- sis of avoidable mortality in three nations and their world cities. Eur J Public Health, 18:166–172. http://dx.doi.org/10.1093/eur pub/ckm084 PMID:17690129
13. de Martel C, Ferlay J, Franceschi S et al. (2012). Global burden of cancers attribut- able to infections in 2008: a review and synthetic analysis. Lancet Oncol, 13:607– 615. http://dx.doi.org/10.1016/S1470-2045 (12)70137-7 PMID:22575588
14. Knaul FM, Adami HO, Adebamowo C et al. (2012). The global cancer divide: an equity imperative. In: Knaul FM, Gralow JR, Atun R, Bhadelia A, eds. Closing the Cancer Divide: An Equity Imperative. Based on the work of the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries. Cambridge, MA: Harvard Global Equity Initiative, pp. 29–70.
15. Gralow GR, Krakauer E, Anderson BO et al. (2012). Core elements for provision of cancer care and control in low and middle income countries. In: Knaul FM, Gralow JR, Atun R, Bhadelia A, eds. Closing the Cancer Divide: An Equity Imperative. Based on the work of the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries. Cambridge, MA: Harvard Global Equity Initiative, pp.123–165.
16. WHO (2011). Global Status Report on Noncommunicable Diseases 2010. Geneva: WHO. Available at http://www.who.int/nmh/ publications/ncd_report2010/en/.
17. World Bank (2010). World Development Indicators 2010. Washington DC: World Bank. Available at http://data.worldbank. org/sites/default/files/wdi-final.pdf.
18. Global Risk Network of the World Economic Forum (2010). Global Risks 2010: A Global Risk Network Report. Geneva: World Economic Forum. Available at www3. weforum.org/docs/ WEF_GlobalRisks_ Report_2010.pdf.
19. DeVol R, Bedroussian A (2007). An Unhealthy America: The Economic Burden of Chronic Disease – Charting a New Course to Save Lives and Increase Productivity and Economic Growth. Santa Monica, CA: Milken Institute.
20. Stuckler D (2008). Population causes and consequences of leading chronic dis- eases: a comparative analysis of prevail- ing explanations. Milbank Q, 86:273–326. ht tp://dx.doi.org/10.1111/j.1468 - 0 0 09. 2008.00522.x PMID:18522614
21. Atun R, Knaul FM, Akachi Y, Frenk J (2012). Innovative financing for health: what is truly innovative? Lancet, 380:2044–2049. http:// dx.doi.org/10.1016/S0140-6736(12)61460-3 PMID:23102585
22. Konduri N, Quick J, Gralow JR et al. (2012). Access to affordable medicines, vaccines, and health technologies. In: Knaul FM, Gralow JR, Atun R, Bhadelia A, eds. Closing the Cancer Divide: An Equity Imperative. Based on the work of the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries. Cambridge, MA: Harvard Global Equity Initiative, pp.197–256.
23. Kim JY, Farmer P, Porter ME (2013). Redefining global health-care delivery. Lancet, 382:1060–1069. http://dx.doi. org /10.1016/S 014 0 - 673 6 (13) 610 47- 8 PMID:23697823
24. Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
584
Cancer control in Peru: “El Plan Esperanza” – the Hope Plan Carlos Vallejos
Peru has a population of about 30 million people, and the region of Lima (the capital city) is home to about one third of the national population. Peru allocates 5.1% of its gross domestic product to ex- penditure on health. The growth of the Peruvian economy in the past two decades has led to an increase in life expectancy, from 66.7 years in 1995 to 74.1 years in 2010. Prostate cancer (25.1% of all cases), stom- ach cancer (15.7%), non-Hodgkin lymphoma (5.8%), and lung cancer (5.4%) are the most frequent ma- lignant tumours in men, and cervi- cal cancer (19.5%), breast cancer (18.9%), stomach cancer (11.5%), and colorectal cancer (5.1%) are the most common in women. Peru’s first institution for specialized cancer treatment was the Instituto Nacional de Enfermedades Neoplásicas (INEN), which was founded in 1939 and is supported by government funds. INEN was designated as an autonomous institution in 1985 and
is required to formulate, regulate, and advise on the development of health policies in the field of cancer.
Problems frequently faced by oncologists in Peru have been: (i) tumours in advanced stages, mainly because there is a long pe- riod between the first signs of cancer and seeking specialist attention; (ii) lack of adherence to treatment, or delay in the treatment, due to costs and because sometimes patients have to travel long distances to re- ceive treatment; and (iii) inadequate follow-up. These three problems are certainly related to the geographical distance between a patient’s home and a treatment facility. Until 2000, INEN was the only public centre for cancer care in Peru. The develop- ment of new policies for the decen- tralization of cancer care led to the inauguration of a new specialized centre covering a major population.
In 2007, Peru implemented the first national cancer control plan in Latin America. Priority was accorded
to the most relevant cancers in Peru – cancers of the breast, cervix, stomach, lung, and prostate – and beneficial results were obtained. In addition, new policies and strat- egies were developed to provide access to medical attention for the poorest people under a universal insurance coverage. In 2011, a newly fledged budgetary strategic programme (budget by results) for cancer control was released, which provided US$ 10.3 million, with the benefits being directed mainly to 10 regions (selected according to pri- orities in terms of population size and cancer incidence). Meeting strategic goals has led to an in- crease in the budget for nationwide activity to US$ 29.8 million for 2012. These state policies established in the past 10 years have established the foundation for the current “Plan Esperanza” (Hope Plan).
Cancer control experiences 585 6 C
O N
T R
O L
Cancer control in Turkey: an encouraging national cancer control plan for the future Ezgi Hacikamiloglu, Guledal Boztas, Murat Gultekin, and Murat Tuncer
Turkey’s national cancer control programme, which was prepared and initiated in cooperation with national and international organi- zations to decrease the number of cancer-related deaths, consists of four main pillars: cancer registry activities, cancer prevention, can- cer screening and early diagnosis, and cancer treatment and palliative care.
Cancer registry activities Since 2002, in conjunction with passive registration, active cancer registration has also been imple- mented. There are 16 active cancer registration centres in 15 cities in Turkey (50% population cover- age). Currently, four of the active registration centres are accredited by IARC. The cancer registration centre in the city of İzmir is desig- nated an international training cen- tre (hub) by IARC. The implementa- tion of active registration centres in all 81 cities, with a 100% coverage rate, is planned to be completed by the end of 2015.
Cancer prevention Cancer prevention in Turkey is man- aged under certain programmes. The nationwide programmes are the Alcohol Control Programme; the Healthy Nutrition, Obesity Con- trol, and Promotion of Physical Activity Programmes; the Reduction of Excessive Salt Consumption Programme; the Strategic Asbes- tos Control Programme; the Ra- don Mapping Programme; and the National Tobacco Control
Programme. The National Tobacco Control Programme, which was start- ed in 2008, is extremely successful with excellent outcomes. Tobacco consumption rates decreased from 33.6% to 27.5%, and Turkey be- came the first country in the world to implement all the tobacco control measures of MPOWER [1–3].
Cancer screening and early diagnosis A Cancer Early Diagnosis, Screen- ing, and Training Centre has been opened in each city in Turkey. Population-based screening and public training programmes about breast, cervical, and colorectal cancers are being organized within these centres totally free of charge. In 2012, the national cancer screen- ing standards were revised as fol- lows: breast cancer screening every 2 years from the age of 40 years; colorectal cancer screening, by using immunological faecal occult blood testing in conjunction with colonoscopy, for all people aged 51 years and 61 years; and cervi- cal cancer screening implementing human papillomavirus (HPV) DNA testing. Nationwide software pack- ages have been produced to follow European Union quality criteria on screening. By means of 130 new mobile screening units, digitiza- tion of mammography, and central mammography reporting units, and by evoking the potential help of family physicians and screening via HPV tests, Turkey aims to raise the population screening rates to 70% by 2015.
Cancer treatment and palliative care Cancer treatment is free of charge for all citizens in Turkey, and all standard treatments are available in the country. Treatment facilities are implemented according to the 2023 strategic investment plans, according to the disease burden, geographical transportation, and population dynamics. Also, domes- tic production of some chemother- apeutics and opioids has already started. One of the basic essentials for palliative care, which was start- ed in 2010, is providing health care at home. This service is also being provided for terminal-stage cancer patients. Palliative care legislation has been prepared and is planned to be launched by the end of 2013. In addition, opioid legislation is be- ing evaluated to increase acces- sibility. Turkey aims to implement a population-based palliative care system based on family physicians and home care teams supported by more than 200 palliative care units within the next 5 years.
References 1. WHO (2012). Global Adult Tobacco Survey.
Turkish Statistical Institute. Available at http://www.who.int/tobacco/surveillance/ gats_turkey/en/index.html.
2. WHO (2013). WHO Report on the Global Tobacco Epidemic, 2013: Enforcing Bans on Tobacco Advertising, Promotion and Sponsorship. Geneva: WHO. Available at http://apps.who.int/iris/bitstream/10665/ 85380/1/9789241505871_eng.pdf.
3. WHO (2008). MPOWER: A Policy Package to Reverse the Tobacco Epidemic. Geneva: WHO. Available at http://www.who.int/to bacco/mpower/mpower_english.pdf.
Summary Recent studies of the hazards of people in the United Kingdom, USA, or Japan who began smok- ing in adolescence or early adult life show loss of about 10 years of life expectancy if they continue, and avoidance of more than 90% of the excess risk if they stop before age 40 (and preferably well before age 40). Smokers who did not start in early adult life have much smaller hazards in middle and old age. Hence, when smoking becomes common among a population of young adults, the full eventual effects of tobacco on mor- tality rates in middle and old age take more than half a century to emerge in that population. For women in many developed countries and men in many other countries, there will be a large increase in tobacco- attributed mortality over the next few decades as a result of increas- es in smoking that have already happened, unless there is wide- spread cessation.
Cancer mortality and overall mortality Taking all countries in the world to- gether, smoking is by far the most important cause of cancer, and al- though tobacco-attributed cancer
Richard Peto is a professor of
medical statistics and epidemiology
and a co-director of the Clinical Trial
Service Unit at the University of
Oxford. He studied natural sciences at
Cambridge University and obtained his
M.Sc. in statistics at the University of
London. Dr Peto’s work has included
studies of the causes of cancer in
general, and of the effects of smoking
in particular, and the establishment of
large-scale randomized trials of the
treatment of cancer and various other
diseases. He has been instrumental in
introducing combined “meta-analyses”
of results from diverse studies.
Dr Peto is one of the world’s most
cited medical researchers and was
knighted in 1999 for his services to
epidemiology and cancer prevention.
He devotes much of his energy to
advising and providing information on
“avoidable death”. His work continues
to have a direct influence on public
policy and adult mortality
in many countries.
mortality is now falling in some populations in developed countries, worldwide it is increasing. Smoking can cause death not only from lung cancer (the main neoplastic hazard) but also from cancer of the mouth, pharynx, larynx, oesophagus, stom- ach, pancreas, liver, kidney, bladder, or cervix [1]. It also causes more deaths from other diseases (e.g. heart disease, stroke, chronic ob- structive lung disease, tuberculosis, and pneumonia) than from cancer.
This Perspective summarizes the full eventual effects on lung cancer mortality, overall cancer mortality, and all-cause mortality of starting to smoke in early adult life and continuing to do so and, conversely, the full eventual effects among smokers who did start when young of cessation at various ages, in comparison with the hazards they would have faced if they had continued smoking.
In general, among cigarette smokers who started in early adult life, those who stop before middle age (here defined as ages 35–69) gain about 10 years of life expec- tancy. This is now known to be true for both men and women [2–5]. It is likely to be true in any popula- tion where the smokers who have not yet reached middle age started smoking cigarettes in early adult life, even if in that population the
586
The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality
Richard Peto in collaboration with Alan D. Lopez, Hongchao Pan, and Michael J. Thun
effects of smoking on mortality in middle and old age (i.e. in previous generations) are not yet substantial, because those who started in early adult life will, if they continue smok- ing, eventually experience substan- tial hazards in middle and old age.
Effects of smoking cigarettes throughout adult life Cigarette smokers who start in ear- ly adult life are at far greater risk in later adult life than otherwise similar smokers who start somewhat later [6]. This implies a remarkably long delay of half a century or more be- tween cause and full effect. As long as due allowance is made for this delay, reliable quantitative predic- tions can be made of the substantial hazards that will eventually be faced by those who have been cigarette smokers since early adult life but are still only in their twenties, thirties, or forties if they continue to smoke, and of the substantial benefits for such smokers of stopping at various ages.
Cigarette smoking is extraor- dinarily destructive (Box P7.1; Figs P7.1, P7.2). It is common in many populations, and where it has been widespread among young adults for many decades, at least half of all persistent cigarette smokers
are eventually killed by it, unless they stop. Bidi smoking (bidis con- sist of a small amount of tobacco wrapped in the leaf of another plant), which is common in parts of South Asia, can cause similar risks [7–8].
Cigarette smoking causes rel- atively few deaths before about 35 years of age but causes many deaths in middle and old age. Although some of those killed by tobacco in middle age might have died soon anyway, many would have lived on for another 10, 20, 30, or more years (Box P7.1).
British men, the first severely affected population to be studied Studies of British men born in the first few decades of the 20th century were particularly informative about the full lifelong hazards of smoking and the benefits of stopping because this was the first large population in which many had begun to smoke cigarettes in early adult life and con- tinued to do so. By 1970, Britain had the worst tobacco-attributed mortal- ity rates in the world [9–11]. The life- long effects of persistent cigarette smoking, and the corresponding benefits of stopping, can therefore
be illustrated by Doll’s study of male British doctors born during the first few decades of the 20th century and followed prospectively through- out the second half of the century (Fig. P7.1) [2].
The doctors who smoked (and those who later stopped smoking) had on average begun at 18 years of age. There were no big differ- ences between smokers, former smokers, and never-smokers in occupation, obesity, or alcohol consumption [12]. As all were doctors, they were easily traced even if they emigrated, and the underlying causes of most deaths were recorded reliably.
Fig. P7.1A compares the ciga- rette smokers with the never-smok- ers, showing a 10-year difference in life expectancy. During middle age (35–69), 19% of the never-smokers and 42% of the cigarette smokers died (i.e. the respective probabili- ties of survival from age 35 to age 70 were 81% and 58%). Much of this absolute difference of 23% in mortality was actually caused by smoking, because it mainly involved differences in the numbers dying from diseases that can be caused by smoking (lung cancer, heart dis- ease, chronic lung disease, etc.).
The risk is big, if they continue smoking
• At least half are eventually killed by smoking, if they continue. (Among persistent cigarette smokers, male or female, the overall relative risk of death is greater than 2 throughout mid- dle age and well into old age. Thus, among smokers of a giv- en age, more than half of those who die in the near future would not have done so at never- smoker death rates.)
• On average, smokers lose at least 10 years of life. (This av- erage combines a zero loss
for those not killed by tobacco with a loss of much more than 10 years for those who are killed by it.)
Those killed in middle age (35–69 years of age) lose many years of life
• Some of those killed in middle age might have died soon any- way, but others might have lived on for another 10, 20, 30, or more years.
• On average, those killed in mid- dle age lose about 20 years of never-smoker life expectancy.
Stopping smoking works
• Those who stop before age 40 (and preferably well before age 40) avoid more than 90% of the excess risk among those who continue to smoke. Those who stop before age 30 avoid more than 97% of the smokers’ ex- cess risk.
• Those who have smoked ciga- rettes since early adult life but stop at 60, 50, 40, or 30 years of age gain, respectively, about 3, 6, 9, or almost the full 10 years of life expectancy, compared with those who continue smoking.
Box P7.1. Main findings for the individual who becomes a cigarette smoker in adolescence or early adult life.
The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality 587
Twenty-first century hazards in women born around 1940 in the United Kingdom and the USA Among British women, few born early in the 20th century smoked from early adult life, but many born around 1940 did so and were there- fore at high risk in later adult life if they continued to smoke. A recent prospective study of 1.3 million such
women found hazards comparable to those in men [3].
Fig. P7.2A shows the relation- ship between daily cigarette con- sumption and all-cause mortal- ity among these women during the 2000s, when they were in their sixties. On average, smokers had 3 times the overall mortality rate of otherwise comparable never- smokers. This 3-fold relative risk is
standardized for age and for many other factors. It therefore means that about two thirds of the deaths of smokers would not have oc- curred when they did if the smok- ers had had the same death rates as otherwise similar nonsmokers. Smoking even just a few cigarettes a day was sufficient to double the overall mortality rate.
Similarly extreme smoker ver- sus nonsmoker mortality ratios dur- ing the 2000s, and a similar 10-year difference in survival, have recently been reported for men and women in the USA and in Japan [4,5,13].
Cessation, lung cancer mortality, and all-cause mortality in the United Kingdom and the USA Smokers who stop before age 40 (preferably well before age 40) avoid more than 90% of their risk of being killed by tobacco, gaining on aver- age more than 9 extra years of life expectancy. Fig. P7.1B shows that in Doll’s study of British doctors, men who stopped at about 40 (35–44) years of age avoided about 90% of the excess risk they would have suf- fered if they had continued smok- ing. Those who stopped at about 60, 50, 40, or 30 years of age gained,
Fig. P7.1. The full eventual effects of smoking from early adult life in men in the United Kingdom (born 1900–1930), showing the lifelong hazards of smoking and the benefits of stopping at age 40 among male British doctors followed up until old age. (A) Survival from age 35 in continuing smokers and never-smokers, showing 10-year loss. (B) Survival from age 40 in smokers, never-smokers, and those who stopped at age 35–44. Follow-up was from 1951 until 2001, with smoking recorded in 1951 and again every few years until 2001. The smokers and the former smokers had both started at mean age 18 years.
Fig. P7.2. The full eventual effects of smoking from early adult life in women (from the Million Women Study during the 2000s of British women born around 1940): multivariable-adjusted relative risks in never-smokers and in continuing smokers, by daily dose, (A) for all-cause mortality and (B) for lung cancer mortality. The smokers had started at mean age 19 years. For each category, the area of the square is inversely proportional to the variance of the category-specific log risk, which also determines the confidence interval.
588
respectively, about 3, 6, 9, or almost the full 10 years [2].
Fig. P7.3A shows, based on the much larger Million Women Study, more precise evidence that stop- ping at age 40 avoids about 90% of the excess mortality among those who continue smoking [3]. Similarly extreme benefits of having stopped at 30 or 40 years of age have been reported from studies of mortality in the USA during the 2000s [4,5].
Lung cancer is one of the main diseases caused by smoking. Even though it accounts for less than half of all smoking-attributed mortality, when the lung cancer death rates among persistent cigarette smok- ers, former smokers, and never- smokers are compared, the relative risks are so extreme that the long- term hazards of smoking and the benefits of stopping can be seen particularly clearly [3,14,15].
Fig. P7.3B shows the lung can- cer findings for smokers by age at cessation in the Million Women Study [3]. Although the former smokers who stopped at about age 30, 40, or 50 still had a highly signif- icant excess lung cancer risk some
decades after stopping, there is a large absolute difference between the risks in those who stopped at these ages and in those who con- tinued smoking.
Those who stopped at 30 avoid- ed 97% of the excess lung cancer risk a few decades later in those who continued (i.e. they had only 3% of the excess risk, with confi- dence interval 2–4%), and those who stopped at 40 avoided 90% of the excess lung cancer risk in those who continued (i.e. they had only 10% of the smokers’ excess risk).
Comparable findings have re- cently been reported for men and women in the USA [4,5].
Evolution of the epidemic in male and female smokers in the USA The report by Thun et al. [5] is partic- ularly illuminating because it spans a 50-year period (1960–2010) during which there was a more than 10-fold increase in the excess lung cancer mortality among female smokers as the epidemic matured. It describes three separate large studies – one in the 1960s, one in the 1980s, and one
in the 2000s – that recorded each woman’s smoking and then moni- tored lung cancer rates over the next few years among women who were older than 55 years. The age-stand- ardized death rate from lung cancer in women who had never smoked was about the same in all three stud- ies because there has been no big change in nonsmoker lung cancer rates in the USA.
The lung cancer risk ratio com- paring current smokers with other- wise similar never-smokers was, however, very different in the three studies (Table P7.1): only 3-fold in the 1960s, 13-fold in the 1980s, and 26-fold in the 2000s (similar to the 24-fold risk ratio in the Million Women Study in the 2000s in the United Kingdom). This is because in the USA women older than 55 who were smokers in the 1960s had not been smoking ever since early adult life, whereas most who were smokers in the 2000s had been do- ing so.
Note that in the USA many women who were young smokers in the 1960s (when their mothers and grandmothers were enjoying
Fig. P7.3. The benefits of stopping at about 30, 40, or 50 years of age in a population where substantial effects of smoking are already apparent (from the Million Women Study during the 2000s of British women born around 1940): multivariable-adjusted relative risks (1.0 for never-smokers) in ex-smokers and in current smokers, (A) for all-cause mortality and (B) for lung cancer mortality. The group of continuing cigarette smokers and the groups who stopped at ages 25–34, 35–44, or 45–54 had all started at mean age about 19 years and all smoked about 15 cigarettes per day. The area of each square is inversely proportional to the variance of the log relative risk (vs never-smokers), which also determines the confidence interval.
589The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality
relatively low lung cancer rates in middle and old age) had been smoking all their young adult lives, so those of them who kept on smok- ing for the next 40 years went on to become the older smokers in the 2000s with high lung cancer rates (much higher than those in previous generations of women).
Underestimation of eventual hazards in studies of other populations Cigarette consumption was low throughout the world in 1900, but among men in many developed countries, such as the United Kingdom and the USA, it increased substantially during the first few decades of the 20th century [1]. In recent decades it has also increased substantially among women in many developed countries and among men in many developing countries, including China [16].
When in a particular popula- tion there is an upsurge of cigarette smoking among young adults, it will be 40 or more years before the main upsurge of tobacco-related deaths in middle age is seen, and then another 20 years before the main upsurge of tobacco-related deaths in old age. Thus, even in populations where there is not yet a high death rate from smoking (be- cause relatively few people who are now in middle or old age have been cigarette smokers throughout adult life), many of the young adults who smoke began to do so in adoles- cence or early adult life, so decades hence they will face substantial risks in middle and old age if they continue, and they have much to gain from prompt cessation.
Many previous studies of smok- ing and disease were in populations where the middle-aged or, particu- larly, older smokers had not been smoking cigarettes throughout adult life. The risks found by comparing smokers and never-smokers (or for- mer smokers) in those studies may therefore greatly underestimate the risks that the younger ciga- rette smokers of today will eventu- ally face if they continue. Hence, in those previous studies the benefits of cessation appear to be substan- tially less than the benefits that the younger cigarette smokers of today would gain from cessation (in com- parison with the risks they would face if they were to continue). Thus, for example, earlier studies that suggested relatively small excess risks among smokers in Japan have been succeeded by a recent study that shows there are now large risks among those in Japan who contin- ue to smoke [13].
In many past studies the pro- portional excess mortality in mid- dle and old age from smoking was greater among male than among female smokers, but this was chiefly because male smokers had smoked cigarettes more intensively when young than had female smok- ers. Recent studies of women in the United Kingdom and the USA show, however, that smoking cigarettes throughout adult life eventually pro- duces about as great a proportional increase in female as in male over- all death rates, so in terms of years of life expectancy lost, the even- tual hazards of persistent cigarette smoking (and the corresponding benefits of cessation) will be about as great for women as for men.
Likewise in China, the hazards that younger cigarette smokers will face in middle and old age may well be substantially greater than the risks now seen among Chinese smokers in middle and old age [17]. Indeed, for any cigarette smoker, male or female, in any part of the world who started smoking sub- stantial numbers of cigarettes when young and has continued, the even- tual hazards may well be similar: about half will be killed by smoking unless they stop, and cessation be- fore age 40 (preferably well before 40) would avoid more than 90% of that risk.
In any population in the world, therefore, the prevalence of ciga- rette (or bidi) smoking among young adults can be used as a proxy to predict reasonably reliably the eventual future impact of smok- ing on mortality in that population several decades hence if those who now smoke continue to do so, and to predict the importance for those who now smoke of prompt cessation.
Contrasting national trends in tobacco-attributed mortality at ages 35–69 In many countries the trends in over- all cancer mortality in the past few decades have been dominated by the delayed effects of long-past in- creases in cigarette smoking among young adults and, more recently, by the more rapid effects of widespread cessation. The United Kingdom, the USA, and Poland offer contrasting examples of this (Figs P7.4–P7.9).
By 1970, male tobacco-attrib- uted mortality rates in the United Kingdom were the worst in the
Table P7.1. Maturing of the lung cancer epidemic in male and female smokers in the USA from the 1960s to the 2000s: ratio of lung cancer death rates, current smoker versus never-smoker, in three large prospective studies
Sex Ratio (95% confidence interval)a
1959–1965 1982–1988 2000–2010
Male 12 (10–16) 24 (21–28) 25 (22–28)
Female 3 (2– 4) 13 (11–14) 26 (23–28)
a Multivariable-adjusted lung cancer mortality ratio, current smoker versus never-smoker (age > 55 years), and 95% confidence interval. Age-standardized never-smoker lung cancer rates were similar in different time periods (showing no significant trend) and in men and women.
590
world, with smoking causing well over half of all cancer mortality and almost half of all mortality at ages 35–69, and female tobacco-attrib- uted mortality was rising (Figs P7.4, P7.5). Over the past few decades in the United Kingdom, however, there has been widespread ces- sation, a substantial decrease in male tobacco-attributed mortality
and, more recently, some decrease in female tobacco-attributed mor- tality. Had female smokers in the United Kingdom continued smok- ing, there would have been a major increase in female tobacco-attribut- ed mortality in the United Kingdom throughout the past few decades instead of the moderate decrease actually seen.
In the USA (Figs P7.6, P7.7), a rapid increase in male tobacco- attributed mortality was still in progress when it was halted (well before the cancer or overall death rates from smoking had become as high as they were in the United Kingdom) by a substantial decrease in cigarette consumption over the past few decades. In the USA, male
Fig. P7.4. United Kingdom, 1950–2010. Total cancer mortality rates at ages 0–34 and 35–69, with the rates at 35–69 subdivided into parts attributed, and not attributed, to smoking. Rates are calculated from WHO mortality data and UN population estimates. Notes: The rate for a 35-year age range is the mean of the 7 annual death rates in the component 5-year age ranges. (Hence, without other causes of death, a rate of R per 100 000 would imply a 35-year risk of 1 – exp[–35R/100 000].) The mortality attributed to smoking is estimated indirectly from the national mortality statistics, using the absolute lung cancer rate as a guide to the fraction of the deaths from other causes, or groups of causes, attributable to smoking.
Fig. P7.5. United Kingdom, 1950–2010. Probabilities of death at ages 0–34 and 35–69, with probabilities of death from smoking at ages 35–69 shaded. Rates are calculated from WHO mortality data and UN population estimates. Notes: Most of those killed by smoking would otherwise have survived beyond age 70, but a minority (shaded area to right of dotted line) would have died by 70 anyway. The mortality attributed to smoking is estimated indirectly from the national mortality statistics.
591The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality
lung cancer mortality in early mid- dle age (35–44; data not shown) has fallen substantially since 1970, and male lung cancer mortality later in middle age is now falling. In the USA, the death rate from smoking among women was still low in 1950, but it rose rapidly and by the 1990s female tobacco-attributed mortal- ity rates in the USA were among the worst in the world, although (as in men in the USA) the rise was eventually halted by a decrease in
cigarette consumption. Again, had female smokers in the USA contin- ued smoking, the steep rise in fe- male death rates from smoking in the USA would still be continuing.
In Poland (Figs P7.8, P7.9), the main increase in cigarette smok- ing among men took place around the middle of the century and was followed by a large increase in male tobacco-attributed mortality during the second half of the century (to levels comparable with those seen
20 years earlier in the United Kingdom). However, a decrease in smoking since 1990 and changes in the nature of the cigarette have decreased this mortality. In Poland women have thus far been less se- verely affected than men, but fe- male tobacco-attributed mortality is rising steadily, and young Polish women who now smoke will face substantial hazards if they continue.
The methods used in these three populations to estimate smoking-
Fig. P7.6. USA, 1950–2010. Total cancer mortality rates at ages 0–34 and 35–69, with the rates at 35–69 subdivided into parts attributed, and not attributed, to smoking. Rates are calculated from WHO (and 2009–2010 United States National Center for Health Statistics) mortality data and UN population estimates. For notes, see Fig. P7.4.
Fig. P7.7. USA, 1950–2010. Probabilities of death at ages 0–34 and 35–69, with probabilities of death from smoking at ages 35–69 shaded. Rates are calculated from WHO (and 2009–2010 United States National Center for Health Statistics) mortality data and UN population estimates. For notes, see Fig. P7.5.
592
attributed mortality, past and pre- sent, are indirect [9,10], but the over- all patterns should be reasonably trustworthy, particularly for cancer. The findings demonstrate the enor- mous potential relevance of smoking cessation to cancer mortality and to overall mortality rates in such coun- tries, and the practicability of sub- stantial changes in tobacco-attrib- uted mortality accumulating over a period of decades. This is true both
in populations where smoking is al- ready a major cause of death and in populations where it is not but where it will become so if current smoking patterns persist.
Worldwide trends Worldwide, about 100 million peo- ple a year reach adult life. Based on present smoking patterns, about 30 million (50% of the young men and 10% of the young women) will
start to smoke, and more than two thirds will continue because in low- and middle-income countries cessa- tion is uncommon [16]. Of those who continue to smoke cigarettes or bi- dis, whether in Asia, America, Africa, or Europe, about half will eventually be killed by their habit (unless they die of something else before middle age). Hence, if more than 20 million of these 30 million new smokers a year continue smoking, and do not
Fig. P7.9. Poland, 1955–2010. Probabilities of death at ages 0–34 and 35–69, with probabilities of death from smoking at ages 35–69 shaded. Rates are calculated from WHO mortality data and UN population estimates. For notes, see Fig. P7.5.
Fig. P7.8. Poland, 1965–2010. Total cancer mortality rates at ages 0–34 and 35–69, with the rates at 35–69 subdivided into parts attributed, and not attributed, to smoking. Rates are calculated from WHO mortality data and UN population estimates. For notes, see Fig. P7.4.
593The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality
stop, and half of those who do so are killed by their habit, eventually more than 10 million people per year will be killed by tobacco [18].
Based on current smoking pat- terns, where 30% start and most who start will not stop, worldwide annual mortality from tobacco is likely to reach 10 million per year (i.e. 100 million per decade) before the middle of this century [10,18,19], and will rise somewhat further in later decades. Tobacco is therefore expected to cause about 150 million deaths in the first quar- ter of this century (many of which have already happened, as smok- ing is already causing about 6 mil- lion deaths a year worldwide [20]) and 250–300 million in the second quarter. Predictions for the third and, particularly, the fourth quarter of the century are inevitably more speculative. However, due partly to population growth and partly to the maturing of the epidemic, if cur- rent smoking patterns persist then the number of tobacco-attributed deaths is likely to exceed 100 mil- lion per decade throughout the sec- ond half of the century (Table P7.2).
Cessation and not starting The number of tobacco-related deaths predicted to occur before 2050 cannot be greatly reduced un- less a substantial proportion of the
adults who have already been smok- ing for some time give up the habit. A decrease over the next decade or two in the proportion of young peo- ple who become smokers will not have its main effects on mortality until the second half of the century. The effects of adult smokers quit- ting on deaths before 2050 and of young people not starting to smoke on deaths after 2050 will probably be approximately as follows.
Cessation If many of the adults who now smoke were to give up over the next decade or two, thus halving global cigarette consumption per adult by the 2020s, this would prevent about one third of tobacco-related deaths in the 2020s and almost halve tobacco- related deaths thereafter. Within a decade of their occurrence, such changes could avoid 10 or 20 million tobacco-related deaths per decade, and could avoid 100 million tobacco- related deaths in the second quarter of the century.
Not starting If, by reduction over the next decade or two in the global uptake rate of smoking by young people, the pro- portion of young adults who become smokers were to be halved by the 2030s, this would avoid hundreds of millions of deaths from tobacco
in the second half of the century. It would, however, avoid almost none of the 150 million deaths from tobac- co in the first quarter of the century, and would probably avoid “only” a few million deaths from tobacco in the second quarter of the century.
Thus, using widely practicable ways of helping large numbers of young people not to start smoking could avoid hundreds of millions of tobacco-related deaths in the second half of the century, but not before. In contrast, widely practica- ble ways of helping large numbers of adult smokers to quit (prefer- ably before middle age, but also in middle age) could well avoid more than 100 million tobacco-related deaths in the first half of this cen- tury. Large numbers of deaths dur- ing the second half of the century would also be avoided if many of those who, despite everything, still start to smoke in future years could be helped to stop before they are killed by tobacco. Such calculations suggest that the effect of quitting could be more rapidly apparent on a population scale than the effects of not starting to smoke. Both, how- ever, are of great importance.
This Perspective is adapted from IARC (2007) [21]. We are grateful to Jillian Boreham and Kirstin Pirie for help with the figures.
Table P7.2. Projected numbers of deaths from tobacco during the 21st century, if current smoking patterns persista
Period (years) Deaths from tobacco (millions)
2000–2024 ~150
2025–2049 ~250–300
2050–2099 > 500
Total, entire 21st century ~1000
Total, entire 20th century ~100
a Worldwide, about 30% of young adults become smokers, and with the current low cessation rates among smokers in low- and middle-income countries, most who start will not stop.
594
References
1. IARC (2004). Tobacco smoke and involun- tary smoking. IARC Monogr Eval Carcinog Risks Hum, 83:1–1438. PMID:15285078
2. Doll R, Peto R, Boreham J, Sutherland I (2004). Mortality in relation to smoking: 50 years’ observations on male British doc- tors. BMJ, 328:1519–1527. http://dx.doi.org /10.1136/bmj.38142.554479.AE PMID:152 13107
3. Pirie K, Peto R, Reeves GK et al.; Million Women Study Collaborators (2013). The 21st century hazards of smoking and benefits of stopping: a prospective study of one million women in the UK. Lancet, 381:133–141. http://dx.doi.org/10.1016/S0 140-6736(12)61720-6 PMID:23107252
4. Jha P, Ramasundarahettige C, Landsman V et al. (2013). 21st-century hazards of smoking and benefits of cessation in the United States. N Engl J Med, 368:341–350. http://dx.doi.org/10.1056/NEJMsa1211128 PMID:23343063
5. Thun MJ, Carter BD, Feskanich D et al. (2013). 50-year trends in smoking-related mortality in the United States. N Engl J Med, 368:351–364. http://dx.doi.org/10.1056/NE JMsa1211127 PMID:23343064
6. Doll R, Peto R (1981). The causes of cancer: quantitative estimates of avoid- able risks of cancer in the United States today. J Natl Cancer Inst, 66:1191–1308. PMID:7017215
7. Gajalakshmi V, Peto R, Kanaka TS, Jha P (2003). Smoking and mortality from tu- berculosis and other diseases in India: retrospective study of 43000 adult male deaths and 35000 controls. Lancet, 362:507–515. http://dx.doi.org/10.1016/S0 140-6736(03)14109-8 PMID:12932381
8. Jha P, Jacob B, Gajalakshmi V et al.; RGI-CGHR Investigators (2008). A nation- ally representative case-control study of smoking and death in India. N Engl J Med, 358:1137–1147. http://dx.doi.org/10.1056/N EJMsa0707719 PMID:18272886
9. Peto R, Lopez AD, Boreham J et al. (1992). Mortality from tobacco in developed coun- tries: indirect estimation from national vital statistics. Lancet, 339:1268–1278. http:// dx.doi.org/10.1016/0140-6736(92)91600-D PMID:1349675
10. Peto R, Lopez AD, Boreham J et al. (1994). Mortality from Smoking in Developed Countries 1950 –2000: Indirect Estimates from National Vital Statistics. Oxford: Oxford University Press.
11. Thun M, Peto R, Boreham J, Lopez AD (2012). Stages of the cigarette epi- demic on entering its second century. Tob Control, 21:96–101. http://dx.doi.org /10.113 6/tobac c oc ontrol -2011- 0 5 029 4 PMID:22345230
12. Doll R, Peto R, Wheatley K et al. (1994). Mortality in relation to smoking: 40 years’ observations on male British doctors. BMJ, 309:901–911. http://dx.doi.org/10.1136/bmj. 309.6959.901 PMID:7755693
13. Sakata R, McGale P, Grant EJ et al. (2012). Impact of smoking on mortality and life expectancy in Japanese smok- ers: a prospective cohort study. BMJ, 345: e7093. http://dx.doi.org/10.1136/bmj.e7093 PMID:23100333
14. Peto R, Darby S, Deo H et al. (2000). Smoking, smoking cessation, and lung cancer in the UK since 1950: combina- tion of national statistics with two case- control studies. BMJ, 321:323–329. http:// d x .d o i.o r g /10 .113 6 / b m j. 3 21.72 5 7. 3 2 3 PMID:10926586
15. Brennan P, Crispo A, Zaridze D et al. (2006). High cumulative risk of lung can- cer death among smokers and non-smok- ers in Central and Eastern Europe. Am J Epidemiol, 64:1233–1241. http://dx.doi.org /10.1093/aje/kwj340
16. Giovino GA, Mirza SA, Samet JM et al.; GATS Collaborative Group (2012). Tobacco use in 3 billion individuals from 16 countries: an analysis of nationally repre- sentative cross-sectional household sur- veys. Lancet, 380:668–679. http://dx.doi. o r g /10.1016 / S 014 0 - 673 6 (12) 610 8 5 -X PMID:22901888
17. Peto R, Chen ZM, Boreham J (1999). Tobacco – the growing epidemic. Nat Med, 5:15 –17. http://dx.doi.org/10.10 38/4691 PMID:9883828
18. Peto R, Lopez AD, Boreham J et al. (1996). Mortality from smoking world- wide. Br Med Bull, 52:12–21. http://dx.doi. org/10.1093/oxfordjournals.bmb.a011519 PMID:8746293
19. Peto R, Lopez AD (2001). Future world- wide health effects of current smoking pat- terns. In: Koop CE, Pearson C, Schwarz MR, eds. Critical Issues in Global Health. New York: Jossey-Bass, pp. 154–161.
20. Lim SS, Vos T, Flaxman AD et al. (2012). A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 re- gions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet, 380:2224–2260. http:// dx.doi.org/10.1016/S0140-6736(12)61766- 8 PMID:23245609
21. IARC (2007). The hazards of smoking and the benefits of stopping: cancer mortality and overall mortality. In: IARC Handbooks of Cancer Prevention, Vol. 11: Tobacco Control: Reversal of Risk After Quitting Smoking. Lyon: IARC, pp. 15–27.
595The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality
596
Jean-Pierre Abastado Singapore Immunology Network Singapore [email protected]
Cary Adams Union for International Cancer Control Geneva, Switzerland [email protected]
Isaac F. Adewole University of Ibadan Ibadan, Nigeria [email protected]
Hideyuki Akaza Research Center for Advanced Science and Technology The University of Tokyo Tokyo, Japan [email protected]
Naomi E. Allen University of Oxford Oxford, United Kingdom [email protected]
Marcella Alsan Stanford University Stanford, CA, USA [email protected]
Nada Al Alwan Iraqi National Cancer Research Center Baghdad University Medical College Baghdad, Iraq [email protected]
Mahul B. Amin Cedars-Sinai Medical Center Los Angeles, CA, USA [email protected]
Benjamin O. Anderson University of Washington School of Medicine Fred Hutchinson Cancer Research Center Seattle Cancer Care Alliance Seattle, WA, USA [email protected]
Ahti Anttila Mass Screening Registry/Finnish Cancer Registry Helsinki, Finland [email protected]
Daniel A. Arber Stanford University Medical Center Stanford, CA, USA [email protected]
Bruce K. Armstrong Sydney School of Public Health The University of Sydney Sydney, Australia [email protected]
Héctor Arreola-Ornelas Fundación Mexicana para la Salud Mexico City, Mexico [email protected]
Silvina Arrossi Centro de Estudios de Estado y Sociedad (CEDES/CONICET) National Program on Cervical Cancer Prevention Ministry of Health/National Cancer Institute Buenos Aires, Argentina [email protected]
Rifat Atun Imperial College London London, United Kingdom and Harvard School of Public Health Harvard University Boston, MA, USA [email protected]
Robert A. Baan International Agency for Research on Cancer Lyon, France [email protected]
Yung-Jue Bang Seoul National University College of Medicine Seoul National University Hospital Seoul, Republic of Korea [email protected]
Emmanuel Barillot Institut Curie Paris, France [email protected]
Jill Barnholtz-Sloan Case Comprehensive Cancer Center Case Western Reserve University School of Medicine Cleveland, OH, USA and Central Brain Tumor Registry of the United States Hinsdale, IL, USA [email protected]
Laura E. Beane Freeman National Cancer Institute Bethesda, MD, USA [email protected]
Rachid Bekkali Lalla Salma Foundation for Cancer Prevention and Treatment Rabat, Morocco [email protected]
Agnès Binagwaho Ministry of Health Kigali, Rwanda [email protected]. edu
Contributors
World Cancer Report 597
Elizabeth H. Blackburn University of California San Francisco, CA, USA [email protected]
Evan Blecher American Cancer Society Atlanta, GA, USA [email protected]
Ron Borland Cancer Council Victoria Carlton, Australia [email protected]
Fred T. Bosman University Institute of Pathology Lausanne, Switzerland [email protected]
Peter Bouwman Netherlands Cancer Institute Division of Molecular Pathology Cancer Genomics Centre Netherlands & Cancer Systems Biology Center Amsterdam, Netherlands [email protected]
Guledal Boztas Cancer Control Department Public Health Institute Turkish Ministry of Health Ankara, Turkey [email protected]
Elisabeth Brambilla Centre Hospitalier Universitaire Albert Michallon Grenoble, France [email protected]
Freddie Bray International Agency for Research on Cancer Lyon, France [email protected]
Paul Brennan International Agency for Research on Cancer Lyon, France [email protected]
Louise A. Brinton National Cancer Institute Bethesda, MD, USA [email protected]
Nathalie Broutet World Health Organization Geneva, Switzerland [email protected]
Michael P. Brown Royal Adelaide Hospital Cancer Centre Adelaide, Australia [email protected]
Heather Bryant Canadian Partnership Against Cancer Toronto, Canada Heather.bryant@partnershipagainst cancer.ca
Nikki Burdett Royal Adelaide Hospital Adelaide, Australia [email protected]
Robert C. Burton School of Public Health and Preventive Medicine Monash University The Alfred Centre Melbourne, Australia [email protected]
Agnès Buzyn Institut National du Cancer Paris, France [email protected]
Kenneth P. Cantor National Cancer Institute Bethesda, MD, USA [email protected]
Federico Canzian German Cancer Research Center Heidelberg, Germany [email protected]
Fátima Carneiro IPATIMUP & Medical Faculty of the University of Porto Centro Hospitalar de São João Porto, Portugal [email protected]
Webster K. Cavenee University of California at San Diego La Jolla, CA, USA [email protected]
Eduardo L. Cazap National Cancer Institute Ministry of Health Buenos Aires, Argentina and Union for International Cancer Control Geneva, Switzerland [email protected]
Frank J. Chaloupka University of Illinois at Chicago Chicago, IL, USA [email protected]
Stephen J. Chanock National Cancer Institute Bethesda, MD, USA [email protected]
Bob Chapman American Cancer Society Cancer Action Network Washington, DC, USA [email protected]
Simon Chapman Sydney School of Public Health The University of Sydney Sydney, Australia [email protected]
Chien-Jen Chen Academic Sinica Taipei, Taiwan, China [email protected]
Il Ju Choi Centre for Gastric Cancer National Cancer Centre Goyang, Republic of Korea [email protected]
Rafael Moreira Claro Center for Epidemiological Studies in Health and Nutrition University of São Paulo São Paulo, Brazil [email protected]
Hans Clevers Hubrecht Institute University Medical Center Utrecht Utrecht, Netherlands [email protected]
598
Vincent Cogliano U.S. Environmental Protection Agency, Integrated Risk Information System Arlington, VA, USA [email protected]
Aaron J. Cohen Health Effects Institute Boston, MA, USA [email protected]
Carlo M. Croce Department of Molecular Virology, Immunology and Medical Genetics The Ohio State University Medical Center Columbus, OH, USA [email protected]
Min Dai Cancer Hospital Chinese Academy of Medical Sciences Beijing, China [email protected]
Sarah C. Darby Clinical Trial Service Unit Nuffield Department of Public Health University of Oxford Oxford, United Kingdom [email protected]
Peter B. Dean International Agency for Research on Cancer Lyon, France [email protected]
Lynette Denny University of Cape Town/Groote Schuur Hospital Cape Town, South Africa [email protected]
Ethel-Michele de Villiers German Cancer Research Center Heidelberg, Germany [email protected]
John E. Dick Campbell Family Institute Ontario Cancer Institute Princess Margaret Cancer Centre University Health Network Toronto, Canada [email protected]
Joakim Dillner Karolinska Institute Stockholm, Sweden [email protected]
Susan M. Domchek Basser Research Center Abramson Cancer Center University of Pennsylvania Philadelphia, PA, USA [email protected]
Tandin Dorji Department of Public Health Ministry of Health Thimphu, Bhutan [email protected]
Roland Dray International Agency for Research on Cancer Lyon, France [email protected]
Majid Ezzati Imperial College London London, United Kingdom [email protected]
Lee Fairclough Canadian Partnership Against Cancer Toronto, Canada Lee.Fairclough@partnership againstcancer.ca
Jacques Ferlay International Agency for Research on Cancer Lyon, France [email protected]
David Forman International Agency for Research on Cancer Lyon, France [email protected]
Silvia Franceschi International Agency for Research on Cancer Lyon, France [email protected]
A. Lindsay Frazier Dana-Farber Cancer Institute Boston, MA, USA [email protected]
Christine M. Friedenreich University of Calgary Alberta Health Services – Cancer Care Calgary, Canada christine.friedenreich@albertahealth services.ca
Søren Friis Danish Cancer Society Research Center Danish Cancer Society Copenhagen, Denmark [email protected]
Tamara S. Galloway College of Life and Environmental Science University of Exeter Exeter, United Kingdom [email protected]
Maurice Gatera Rwanda Biomedical Center Kigali, Rwanda [email protected]
Wentzel C.A. Gelderblom South African Medical Research Council Cape Town, South Africa [email protected]
Margaret A. Goodell Stem Cells and Regenerative Medicine Center Baylor College of Medicine Houston, TX, USA [email protected]
Sharon Lynn Grant International Agency for Research on Cancer Lyon, France [email protected]
Mel Greaves Centre for Evolution and Cancer Division of Molecular Pathology The Institute of Cancer Research Sutton, United Kingdom [email protected]
Adèle C. Green Queensland Institute of Medical Research Brisbane, Australia and University of Manchester Manchester Academic Health Sciences Centre Manchester, United Kingdom [email protected]
World Cancer Report 599
Murat Gultekin Cancer Control Department Public Health Institute Turkish Ministry of Health Ankara, Turkey [email protected]
Christine Guo Lian Brigham and Women’s Hospital Harvard Medical School Boston, MA, USA [email protected]
Prakash C. Gupta Healis, Sekhsaria Institute for Public Health Navi Mumbai, India [email protected]
Ezgi Hacikamiloglu Cancer Control Department Public Health Institute Turkish Ministry of Health Ankara, Turkey [email protected]
Andrew J. Hall Senior Visiting Scientist International Agency for Research on Cancer Lyon, France [email protected]
Stanley R. Hamilton University of Texas MD Anderson Cancer Centre Houston, TX, USA [email protected]
Marianne Hammer Norwegian Cancer Society Oslo, Norway Marianne.Hammer@kreftforeningen. no
Curtis C. Harris National Cancer Institute Bethesda, MD, USA [email protected]
Takanori Hattori Shiga University of Medical Science Tokyo, Japan [email protected]
Zdenko Herceg International Agency for Research on Cancer Lyon, France [email protected]
Hector Hernandez Vargas International Agency for Research on Cancer Lyon, France [email protected]
Rolando Herrero International Agency for Research on Cancer Lyon, France [email protected]
David Hill Melbourne School of Population and Global Health and Melbourne School of Psychological Sciences The University of Melbourne Melbourne, Australia [email protected]
Martin Holcmann Institute of Cancer Research Medical University of Vienna Vienna, Austria [email protected]. at
James F. Holland Icahn School of Medicine at Mount Sinai Mount Sinai Medical Center New York, NY, USA [email protected]
Ralph H. Hruban Johns Hopkins University School of Medicine Baltimore, MD, USA [email protected]
Thomas J. Hudson Ontario Institute for Cancer Research Toronto, Canada [email protected]
Peter A. Humphrey Washington University School of Medicine St. Louis, MO, USA [email protected]
Elaine S. Jaffe National Cancer Institute Bethesda, MD, USA [email protected]
Prabhat Jha St. Michael’s Hospital Dalla Lana School of Public Health University of Toronto Toronto, Canada [email protected]
Jos Jonkers Netherlands Cancer Institute Division of Molecular Pathology Cancer Genomics Centre and Cancer Systems Biology Center Amsterdam, Netherlands [email protected]
Margaret R. Karagas Norris Cotton Cancer Center Dartmouth Medical School Lebanon, NH, USA [email protected]
Michael Karin Laboratory of Gene Regulation and Signal Transduction Department of Pharmacology and Pathology University of California San Diego School of Medicine La Jolla, CA, USA [email protected]
Namory Keita Service de Gynécologie/ Obstétrique Université de Conakry Conakry, Guinea [email protected]
Ausrele Kesminiene International Agency for Research on Cancer Lyon, France [email protected]
Michael C. Kew University of Cape Town Cape Town, South Africa [email protected]
Tim Key University of Oxford Oxford, United Kingdom [email protected]
Thiravud Khuhaprema National Cancer Institute Bangkok, Thailand [email protected]
600
Paul Kleihues Medical Faculty University of Zurich Zurich, Switzerland [email protected]
Günter Klöppel Department of Pathology Technical University of Munich Munich, Germany [email protected]. de
Felicia Marie Knaul Harvard Global Equity Initiative Harvard Medical School Boston, MA, USA and Cáncer de mama: Tómatelo a Pecho Competitividad y Salud, Fundación Mexicana para la Salud Mexico City, Mexico and Global Task Force on Expanded Access to Cancer Care and Control [email protected]
Manolis Kogevinas Centre for Research in Environmental Epidemiology Hospital del Mar Research Institute Barcelona, Spain [email protected]
Nobuo Koinuma Tohoku University School of Medicine Sendai, Japan [email protected]
Barnett S. Kramer National Cancer Institute Rockville, MD, USA [email protected]
Guido Kroemer Institut national de la santé et de la recherche médicale, Institut Gustave Roussy University of Paris Descartes Centre de Recherche des Cordeliers Hopital Européen George Pompidou Paris, France [email protected]
James R. Krycer Garvan Institute Sydney, Australia [email protected]
Sunil R. Lakhani University of Queensland Centre for Clinical Research The Royal Brisbane and Women’s Hospital Brisbane, Australia [email protected]
René Lambert International Agency for Research on Cancer Lyon, France [email protected]
Johanna W. Lampe Fred Hutchinson Cancer Research Centre Seattle, WA, USA [email protected]
Robert R. Langley University of Texas MD Anderson Cancer Center Houston, TX, USA [email protected]
Lawrence H. Lash Wayne State University School of Medicine Detroit, MI, USA [email protected]
Mathieu Laversanne International Agency for Research on Cancer Lyon, France [email protected]
Eric Lavigne Public Health Agency of Canada Ottawa, Canada [email protected]
Eduardo Lazcano Ponce Instituto Nacional de Salud Pública Cuernavaca, Mexico [email protected]
Maria E. Leon International Agency for Research on Cancer Lyon, France [email protected]
Alex C. Liber American Cancer Society Atlanta, GA, USA [email protected]
Jonathan Liberman McCabe Centre for Law and Cancer Cancer Council Victoria and Union for International Cancer Control Carlton, Australia [email protected]
Dongxin Lin Cancer Institute and Hospital Chinese Academy of Medical Sciences Beijing, China [email protected]
Dana Loomis International Agency for Research on Cancer Lyon, France [email protected]
Alan D. Lopez Melbourne School of Population and Global Health The University of Melbourne Melbourne, Australia [email protected]
Joannie Lortet-Tieulent International Agency for Research on Cancer Lyon, France [email protected]
Judith Mackay World Lung Foundation Hong Kong Special Administrative Region, China [email protected]
Roger Magnusson Sydney Law School The University of Sydney Sydney, Australia [email protected]
Reza Malekzadeh Tehran University of Medical Sciences Tehran, Islamic Republic of Iran [email protected]
Andrew Marx Harvard Global Equity Initiative Boston, MA, USA [email protected]
World Cancer Report 601
Beela Sarah Mathew Regional Cancer Centre Trivandrum, India [email protected]
James D. McKay International Agency for Research on Cancer Lyon, France [email protected]
David Melzer University of Exeter Medical School Exeter, United Kingdom [email protected]
Oscar Méndez Fundación Mexicana para la Salud Mexico City, Mexico [email protected]
Kathryn R. Middleton Warren Alpert Medical School of Brown University The Miriam Hospital Providence, RI, USA [email protected]
Martin C. Mihm Jr Brigham and Women’s Hospital Harvard Medical School Boston, MA, USA [email protected]
Anthony B. Miller Dalla Lana School of Public Health University of Toronto Toronto, Canada [email protected]
J. David Miller Carleton University Ottawa, Canada [email protected]
Saverio Minucci Istituto Europeo di Oncologia Milan, Italy [email protected]
Holger Moch University Hospital of Zurich Zurich, Switzerland [email protected]
Elizabeth A. Montgomery Johns Hopkins Medical Institutions Baltimore, MD, USA [email protected]
Rebecca Morton Doherty Union for International Cancer Control Geneva, Switzerland [email protected]
Raul Hernando Murillo Moreno Instituto Nacional de Cancerología Bogotá, Colombia [email protected]
Liam J. Murray Centre for Public Health School of Medicine, Dentistry and Biomedical Sciences Queen’s University Belfast Belfast, United Kingdom [email protected]
Richard Muwonge International Agency for Research on Cancer Lyon, France [email protected]
Robert Newton University of York York, United Kingdom [email protected]
Fidele Ngabo Ministry of Health of Rwanda Kigali, Rwanda [email protected]
Mark J. Nieuwenhuijsen Center for Research in Environmental Epidemiology Barcelona, Spain [email protected]
Jane R. Nilson American Cancer Society Atlanta, GA, USA [email protected]
Omar Nimri Cancer Prevention Department Jordan Cancer Registry Ministry of Health Amman, Jordan [email protected]
Folakemi T. Odedina Pharmaceutical Outcomes and Policy College of Pharmacy Radiation Oncology, College of Medicine Health Disparities, Shands Cancer Center University of Florida Gainesville, FL, USA Prostate Cancer Transatlantic Consortium Seminole, FL, USA [email protected]
G. Johan Offerhaus University Medical Center Utrecht Utrecht, Netherlands [email protected]
Hiroko Ohgaki International Agency for Research on Cancer Lyon, France [email protected]
Magali Olivier International Agency for Research on Cancer Lyon, France [email protected]
Jørgen H. Olsen Danish Cancer Society Research Center Danish Cancer Society Copenhagen, Denmark [email protected]
Hongchao Pan Clinical Trial Service Unit and Epidemiological Studies Unit University of Oxford Oxford, United Kingdom [email protected]
Pier Paolo Pandolfi Beth Israel Deaconess Cancer Center Harvard Medical School Boston, MA, USA [email protected]
Gianpaolo Papaccio Section of Histology and Embryology Tissue Engineering and Regenerative Medicine Division Second University of Naples Naples, Italy [email protected]
602
Chris Paraskeva School of Cellular and Molecular Medicine University of Bristol Bristol, United Kingdom [email protected]
Werner Paulus Institute of Neuropathology Münster, Germany [email protected]
José Rogelio Pérez Padilla Instituto Nacional de Enfermedades Respiratorias Mexico City, Mexico [email protected]
Richard Peto Clinical Trial Service Unit and Epidemiological Studies Unit University of Oxford Oxford, United Kingdom [email protected]
Paul Pharoah Department of Public Health and Primary Care Department of Oncology University of Cambridge Cambridge, United Kingdom [email protected]
Gérald E. Piérard University Hospital of Liège Liège, Belgium and University of Franche-Comté Besançon, France and Laboratory of Skin Bioengineering and Imaging Department of Dermatopathology University Hospital Sart Tilman Liège, Belgium [email protected]
Christopher J. Portier Senior Contributing Scientist Environmental Defense Fund New York, NY, USA [email protected]
Jaime Prat Hospital de la Santa Creu i Sant Pau Autonomous University of Barcelona Barcelona, Spain [email protected]
Rachel Purcell International Agency for Research on Cancer Lyon, France [email protected]
Pekka Puska National Institute for Health and Welfare (THL) Helsinki, Finland [email protected]
You-Lin Qiao Department of Cancer Epidemiology Cancer Institute Chinese Academy of Medical Sciences Peking Union Medical College Beijing, China [email protected]
Thangarajan Rajkumar Cancer Institute Madras, India [email protected]
Kunnambath Ramadas Regional Cancer Centre Trivandrum, India [email protected]
G.K. Rath Department of Radiotherapy All India Institute of Medical Sciences New Delhi, India [email protected]
Cecily S. Ray Healis, Sekhsaria Institute for Public Health Navi Mumbai, India [email protected]
Jürgen Rehm Social and Epidemiological Research Department Population Health Research Group Centre for Addiction and Mental Health Toronto, Canada and Dalla Lana School of Public Health University of Toronto Toronto, Canada and PAHO/WHO Collaborating Centre for Mental Health & Addiction Epidemiological Research Unit Technische Universität Dresden Klinische Psychologie & Psychotherapie Dresden, Germany [email protected]
Luis Felipe Ribeiro Pinto Brazilian National Cancer Institute (INCA) Rio de Janeiro, Brazil [email protected]
Elio Riboli Imperial College London London, United Kingdom [email protected]
Leanne Riley World Health Organization Geneva, Switzerland [email protected]
Ronald T. Riley United States Department of Agriculture – Agricultural Research Service Athens, GA, USA [email protected]
Eve Roman Epidemiology and Cancer Statistics Group, Department of Health Sciences University of York Heslington, United Kingdom [email protected]
Isabelle Romieu International Agency for Research on Cancer Lyon, France [email protected]
Hana Ross American Cancer Society Atlanta, GA, USA [email protected]
Lesley Rushton Imperial College London London, United Kingdom [email protected]
Anne Lise Ryel Norwegian Cancer Society Oslo, Norway [email protected]
Jonathan M. Samet University of Southern California Los Angeles, CA, USA [email protected]
World Cancer Report 603
Massoud Samiei International Atomic Energy Agency-Programme of Action for Cancer Therapy Vienna, Austria [email protected]
Rengaswamy Sankaranarayanan International Agency for Research on Cancer Lyon, France [email protected]
Rodolfo Saracci International Agency for Research on Cancer Lyon, France [email protected]
Guido Sauter University Medical Center Hamburg-Eppendorf Hamburg, Germany [email protected]
Catherine Sauvaget International Agency for Research on Cancer Lyon, France [email protected]
Fabio Savarese Institute of Cancer Research Medical University of Vienna Vienna, Austria fabio.savarese@boehringer- ingelheim.com
Augustin Scalbert International Agency for Research on Cancer Lyon, France [email protected]
Ghislaine Scelo International Agency for Research on Cancer Lyon, France [email protected]
Mark Schiffman National Cancer Institute Rockville, MD, USA [email protected]
Stuart J. Schnitt Beth Israel Deaconess Medical Center Harvard Medical School Boston, MA, USA [email protected]
Joachim Schüz International Agency for Research on Cancer Lyon, France [email protected]
Janice Seinfeld Universidad del Pacifico Lima, Peru seinfeld_ [email protected]
Surendra S. Shastri WHO Collaborating Centre for Cancer Prevention, Screening and Early Detection Tata Memorial Centre Mumbai, India [email protected]
Kevin Shield Centre for Addiction and Mental Health Toronto, Canada [email protected]
Maria Sibilia Institute for Cancer Research Medical University of Vienna Vienna, Austria [email protected]
Jack Siemiatycki University of Montreal Montreal, Canada [email protected]
Ronald Simon University Medical Center Hamburg-Eppendorf Hamburg, Germany [email protected]
Pramil N. Singh Loma Linda University School of Public Health Loma Linda, CA, USA [email protected]
Rashmi Sinha National Cancer Institute Rockville, MD, USA [email protected]
Nadia Slimani International Agency for Research on Cancer Lyon, France [email protected]
Avrum Spira Boston University Medical Centre Boston, MA, USA [email protected]
Gustavo Stefanoff Brazilian National Cancer Institute (INCA) Rio de Janeiro, Brazil [email protected]
Eva Steliarova-Foucher International Agency for Research on Cancer Lyon, France [email protected]
Bernard W. Stewart Cancer Control Program South East Sydney Public Health Unit and School of Women’s and Children’s Health University of New South Wales Sydney, Australia [email protected]. nsw.gov.au
Kurt Straif International Agency for Research on Cancer Lyon, France [email protected]
Simon B. Sutcliffe Terry Fox Research Institute Vancouver, Canada [email protected]
Rosemary Sutton Children’s Cancer Institute Australia Lowy Cancer Research Centre University of New South Wales Sydney, Australia [email protected]
Steven H. Swerdlow University of Pittsburgh School of Medicine Pittsburgh, PA, USA [email protected]
604
Neil D. Theise Beth Israel Medical Center Icahn School of Medicine at Mount Sinai New York, NY, USA [email protected]
David B. Thomas Fred Hutchinson Cancer Research Center Seattle, WA, USA [email protected]
Lester D.R. Thompson Woodland Hills Medical Center Woodland Hills, CA, USA [email protected]
Michael J. Thun American Cancer Society Atlanta, GA, USA [email protected]
Mark R. Thursz Imperial College London London, United Kingdom [email protected]
Massimo Tommasino International Agency for Research on Cancer Lyon, France [email protected]
William D. Travis Memorial Sloan-Kettering Cancer Center New York, NY, USA [email protected]
Edward L. Trimble NCI Center for Global Health National Cancer Institute Rockville, MD, USA [email protected]
Giorgio Trinchieri National Cancer Institute Bethesda, MD, USA [email protected]
Ugyen Tshomo Jigme Dorji Wangchuck National Referral Hospital Thimphu, Bhutan [email protected]
Murat Tuncer Hacettepe University Ankara, Turkey [email protected]
Andreas Ullrich World Health Organization Geneva, Switzerland [email protected]
Toshikazu Ushijima National Cancer Center Research Institute Tokyo, Japan [email protected]
Carlos Vallejos Latin American and Caribbean Society of Medical Oncology (SLACOM) Buenos Aires, Argentina and Oncosalud-AUNA Lima, Perú [email protected]
Piet van den Brandt Maastricht University Maastricht, Netherlands pa.vandenbrandt@maastricht university.nl
James W. Vardiman University of Chicago Medical Center Chicago, IL, USA [email protected]
Jim Vaught National Cancer Institute Bethesda, MD, USA [email protected]
Cesar G. Victora Universidade Federal de Pelotas Rio Grande do Sul, Brazil [email protected]
Paolo Vineis Imperial College London London, United Kingdom [email protected]
Lawrence von Karsa International Agency for Research on Cancer Lyon, France [email protected]
Melanie Wakefield Cancer Council Victoria Carlton, Australia [email protected]
Guiqi Wang Cancer Hospital/Institute Chinese Academy of Medical Sciences Peking Union Medical College Beijing, China [email protected]
Frank Weber Medical Faculty University of Duisburg-Essen Essen, Germany [email protected]
Elisabete Weiderpass Cancer Registry of Norway Oslo, Norway and Arctic University of Norway, University of Tromsø Tromsø, Norway and Karolinska Institute Stockholm, Sweden [email protected]
Zena Werb University of California San Francisco, CA, USA [email protected]
Theresa L. Whiteside University of Pittsburgh Cancer Institute Pittsburgh, PA, USA [email protected]
Christopher P. Wild International Agency for Research on Cancer Lyon, France [email protected]
Walter C. Willett Harvard School of Public Health Boston, MA, USA [email protected]
Dillwyn Williams Strangeways Research Laboratories Cambridge, United Kingdom [email protected]
World Cancer Report 605
Rena R. Wing Brown University Providence, RI, USA [email protected]
Deborah M. Winn National Cancer Institute Bethesda, MD, USA [email protected]
Martin Wiseman World Cancer Research Fund International London, United Kingdom [email protected]
Scott Wittet MalariaCare and Cervical Cancer Prevention Programs, PATH Seattle, WA, USA [email protected]
Magdalena B. Wozniak International Agency for Research on Cancer Lyon, France [email protected]
Hai Yan Duke University Medical Center Durham, NC, USA [email protected]
Teruhiko Yoshida National Cancer Center Research Institute Tokyo, Japan [email protected]
Jiri Zavadil International Agency for Research on Cancer Lyon, France [email protected]
Harald zur Hausen German Cancer Research Center Heidelberg, Germany [email protected]
606
Héctor Arreola-Ornelas reports that his unit at the Mexican Health Foundation benefited from research funding from GlaxoSmithKline, Sanofi S.A., and from Avon Mexico.
Yung-Jue Bang reports that his unit at the Seoul National University College of Medicine benefited from research funding from AstraZeneca, GlaxoSmithKline, Merck, Novartis, Pfizer, Roche, Sanofi-Aventis, Bayer, Bristol-Myers Squibb, Boehringer Ingelheim, Lilly, Otsuka, Hanmi, Green Cross, and from Merrimack; Dr Bang reports receiving personal consultancy fees from AstraZeneca, GlaxoSmithKline, Merck, Novartis, Pfizer, Roche, Sanofi-Aventis, Bayer, Bristol-Myers Squibb, Boehringer Ingelheim, Lilly, Otsuka, Taiho, Macrogenics, Hanmi, and from Green Cross; Dr Bang reports benefiting from support for travel and accom- modation from AstraZeneca, GlaxoSmithKline, Merck, Novartis, Pfizer, Roche, Boehringer Ingelheim, and from Lilly; Dr Bang reports receiving personal speaker’s fees from GlaxoSmithKline, Pfizer, and from Roche.
Elizabeth H. Blackburn reports owning shares in Telomere Health Inc.
Fred T. Bosman reports benefiting from research funding from Pfizer.
Michael P. Brown reports benefiting from research funding from Novartis; Dr Brown reports receiving personal consultancy fees from Amgen, Bayer, Pfizer, Novartis, GlaxoSmithKline, Bristol-Myers Squibb, and from Roche; Dr Brown reports receiving personal speaker’s fees from Bristol-Myers Squibb and from Lilly.
Agnès Buzyn reports receiving personal consultancy fees from Novartis, Bristol-Myers Squibb, and from Amgen; Dr Buzyn reports receiving personal speaker’s fees from Novartis, Bristol-Myers Squibb, and from Amgen.
Eduardo L. Cazap reports that his unit at the Latin American and Caribbean Society of Medical Oncology (SLACOM) benefited from research funding from Poniard Pharmaceuticals and from Daiichi Sankyo Pharma; Dr Cazap reports receiving personal consul- tancy fees from Bayer and from Schering Pharma; Dr Cazap reports receiving personal speaker’s fees from Bayer, Bristol-Myers Squibb, and from Fresenius.
Lynette Denny reports benefiting from research fund- ing from MSD and from GlaxoSmithKline; Dr Denny re- ports receiving personal speaker’s fees from MSD and from GlaxoSmithKline.
Joakim Dillner reports that his unit at the Karolinska Institute benefited from research funding from Sanofi Pasteur, MSD and from Merck.
Susan M. Domcheck reports that her unit at the University of Pennsylvania benefited from research funding from Astra Zeneca and from AbbVie.
Adèle C. Green reports that her unit at the Queensland Institute of Medical Research benefited from research funding from L’Oréal Recherche; Dr Green reports ben- efiting from support for travel and accommodation from L’Oréal Recherche.
James F. Holland reports holding intellectual property rights in two patents owned by his employer, the Icahn School of Medicine at Mount Sinai, on the structure of human mammary tumour virus, and means to detect it.
Ralph H. Hruban reports holding intellectual property rights in a patent owned by Myriad Genetics on PALB2- based diagnostic methods for pancreatic cancer.
Felicia Marie Knaul reports that her unit at the Harvard School of Public Health benefited from research fund- ing from Sanofi S.A., Goldman Sachs Gives, and that her unit at the Mexico Health Foundation benefited from research funding from Sanofi S.A., Avon Mexico, and from GlaxoSmithKline; Dr Knaul reports receiving personal consultancy fees from the Institute for Applied Economics, Global Health and the Study of Business Enterprises, Johns Hopkins University.
Oscar Méndez reports that his unit at the Mexican Health Foundation benefited from research funding from GlaxoSmithKline, Sanofi S.A., and from Avon Mexico.
Fabio Savarese reports being currently employed by Boehringer Ingelheim.
Disclosures of interests
World Cancer Report 607
Mark Schiffman reports that his unit at the United States National Cancer Institute benefited from non- financial research support from Qiagen, Roche, and from GlaxoSmithKline.
Kevin Shield reports that his unit at the Centre for Addiction and Mental Health, Toronto, benefited from research funding from Lundbeck A.S.; Dr Shield reports receiving personal consultancy fees from Lundbeck A.S.; Dr Shield reports benefiting from support for trav- el and accommodation from Lundbeck A.S.
Avrum Spira reports being a founder of and owning shares in Allegro Diagnostics Inc.; Dr Spira reports receiving personal consultancy fees from Allegro Diagnostics Inc.
Mark R. Thursz reports receiving personal consultan- cy fees from Gilead, Bristol-Myers Squibb, and from Janssen Pharmaceuticals; Dr Thursz reports receiv- ing personal speaker’s fees from Gilead, Bristol-Myers Squibb, and from Janssen Pharmaceuticals.
Hai Yan reports benefiting from research funding from Gilead; Dr Yan reports receiving personal con- sultancy fees from Sanofi; Dr Yan reports holding in- tellectual property rights in patents owned by Agios Pharmaceuticals, by Eli Lilly, and by Sanofi-Aventis.
Frank J. Chaloupka, Hans Clevers, Fidele Ngabo, and Elio Riboli did not submit a Declaration of Interest.
608
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
Sources
Boxes 2.3.1 All relative risk functions obtained from Corrao G, Bagnardi V, Zambon A, La Vecchia C (2004). A meta-analysis of alcohol con- sumption and the risk of 15 diseases. Prev Med, 38:613–619. http:// dx.doi.org/10.1016/j.ypmed.2003.11.027 PMID:15066364
P7.1 Compiled from Doll R, Peto R, Boreham J, Sutherland I (2004). Mortality in relation to smoking: 50 years’ observations on male British doctors. BMJ, 328:1519–1527. http://dx.doi.org/10.1136/ bmj.38142.554479.AE PMID:15213107; Pirie K, Peto R, Reeves GK et al.; Million Women Study Collaborators (2013). The 21st cen- tury hazards of smoking and benefits of stopping: a prospective study of one million women in the UK. Lancet, 381:133–141. http:// dx.doi.org/10.1016/S0140-6736(12)61720-6 PMID:23107252; Jha P, Ramasundarahettige C, Landsman V et al. (2013). 21st-century hazards of smoking and benefits of cessation in the United States. N Engl J Med, 368:341–350. http://dx.doi.org/10.1056/NEJMsa1211128 PMID:23343063; Thun MJ, Carter BD, Feskanich D et al. (2013). 50-year trends in smoking-related mortality in the United States. N Engl J Med, 368:351–364. http://dx.doi.org/10.1056/NEJMsa1211127 PMID:23343064; Sakata R, McGale P, Grant EJ et al. (2012). Impact of smoking on mortality and life expectancy in Japanese smokers: a prospective cohort study. BMJ, 345:e7093. http://dx.doi.org/10.1136/ bmj.e7093 PMID:23100333
Charts 5.1.1 GLOBOCAN 2012a
5.1.2 & 5.1.3 Cancer Incidence in Five Continents, Vol. Xb
5.2.1 GLOBOCAN 2012a
5.2.2 & 5.2.3 Cancer Incidence in Five Continents, Vol. Xb
5.3.1 GLOBOCAN 2012a
5.3.2 & 5.3.3 Cancer Incidence in Five Continents, Vol. Xb
5.4.1 GLOBOCAN 2012a
5.4.2 & 5.4.3 Cancer Incidence in Five Continents, Vol. Xb
5.5.1 GLOBOCAN 2012a
5.5.2 & 5.5.3 Cancer Incidence in Five Continents, Vol. Xb
5.6.1 GLOBOCAN 2012a
5.6.2 & 5.6.3 Cancer Incidence in Five Continents, Vol. Xb
5.7.1 GLOBOCAN 2012a
5.7.2 & 5.7.3 Cancer Incidence in Five Continents, Vol. Xb
5.8.1 & 5.8.2 GLOBOCAN 2012a
5.8.3 & 5.8.4 Cancer Incidence in Five Continents, Vol. Xb
5.9.1 GLOBOCAN 2012a
5.9.2 & 5.9.3 Cancer Incidence in Five Continents, Vol. Xb
5.10.1 GLOBOCAN 2012a
5.10.2 & 5.10.3 Cancer Incidence in Five Continents, Vol. Xb
5.11.1 GLOBOCAN 2012a
5.11.2 & 5.11.3 Cancer Incidence in Five Continents, Vol. Xb
5.11.4 GLOBOCAN 2012a
5.11.5 & 5.11.6 Cancer Incidence in Five Continents, Vol. Xb
5.12.1–5.12.3 GLOBOCAN 2012a
5.12.4 –5.12.6 Cancer Incidence in Five Continents, Vol. Xb
5.13.1 GLOBOCAN 2012a
5.13.2 & 5.13.3 Cancer Incidence in Five Continents, Vol. Xb
5.13.4 GLOBOCAN 2012a
5.13.5 & 5.13.6 Cancer Incidence in Five Continents, Vol. Xb
5.14.1 GLOBOCAN 2012a
5.14.2 & 5.14.3 Cancer Incidence in Five Continents, Vol. Xb
5.15.1 GLOBOCAN 2012a
5.15.2 & 5.15.3 Cancer Incidence in Five Continents, Vol. Xb
5.16.1 & 5.16.2 Cancer Incidence in Five Continents, Vol. Xb
World Cancer Report 609
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
Figures 1.1.1 © 2005 Pierre Thiriet, Courtesy of Photoshare.
1.1.2 Courtesy of Rosss via Wikipedia. License: CC BY-SA 2.05.
1.1.3 Courtesy of Carolina Antunes, www.Morguefile.com.
1.1.4 & 1.1.5 GLOBOCAN 2012a
1.1.6 Cancer Incidence in Five Continents, Vol. Xb
1.1.7 GLOBOCAN 2012a
1.1.8 Cancer Incidence in Five Continents, Vol. Xb
1.1.9–1.1.13 GLOBOCAN 2012a
1.1.14 Cancer Incidence in Five Continents, Vol. Xb
1.1.15 –1.1.17 GLOBOCAN 2012a
1.1.18 Cancer Incidence in Five Continents, Vol. Xb
1.1.19–1.1.21 GLOBOCAN 2012a
1.1.22 Cancer Incidence in Five Continents, Vol. Xb
1.1.23–1.1.25 GLOBOCAN 2012a
1.1.26 Cancer Incidence in Five Continents, Vol. Xb
1.1.27–1.1.29 GLOBOCAN 2012a
1.1.30 Cancer Incidence in Five Continents, Vol. Xb
1.1.31–1.1.33 GLOBOCAN 2012a
1.1.34 Cancer Incidence in Five Continents, Vol. Xb
1.1.35 –1.1.37 GLOBOCAN 2012a
1.1.38 Cancer Incidence in Five Continents, Vol. Xb
1.1.39 GLOBOCAN 2012a
1.1.40–1.1.47 Cancer Incidence in Five Continents, Vol. Xb
1.2.1 © 2006 Vinoth Vijayaraghavan, Courtesy of Photoshare.
1.2.2 Data compiled from the Global Health Observatory Data Repository.
1.2.3 & 1.2.4 Data compiled from the United Nations Development Programme.
1.2.5 –1.2.8 GLOBOCAN 2012a
1.2.9 & 1.2.10 Data compiled from GLOBOCAN 2012a and the United Nations Development Programme.
1.2.11 GLOBOCAN 2012a
1.2.12 © 1986 Andrea Fisch, Courtesy of Photoshare.
1.2.13 © 2012 Kyalie Photography, Courtesy of Photoshare.
1.2.14 Freddie Bray.
1.2.15 Data compiled from the Global Health Observatory Data Repository.
1.3.1 GLOBOCAN 2012a
1.3.2 Compiled from Moreno F, Loria D, Abriata G, Terracini B; ROHA network (2013). Childhood cancer: incidence and early deaths in Argentina, 2000–2008. Eur J Cancer, 49:465–473. http:// dx.doi.org/10.1016/j.ejca.2012.08.001 PMID:22980725; Moradi A, Semnani S, Roshandel G et al. (2010). Incidence of childhood cancers in Golestan province of Iran. Iran J Pediatr, 20:335–342. PMID:23056726; Howlader N, Noone AM, Krapcho M et al., eds (2013). SEER Cancer Statistics Review, 1975–2010. Bethesda, MD: National Cancer Institute. Available at http://www.seer.cancer.gov/ csr/1975_2010/browse_csr.php?section=29&page=sect_29_table.02. html; Lacour B, Guyot-Goubin A, Guissou S et al. (2010). Incidence of childhood cancer in France: National Children Cancer Registries, 2000–2004. Eur J Cancer Prev, 19:173–181. http://dx.doi.org/10.1097/ CEJ.0b013e32833876c0 PMID:20361423; Kaatsch P, Spix C (2012). German Childhood Cancer Registry Annual Report 2011 (1980 –2010). Mainz: Institute of Medical Biostatistics, Epidemiology and Informatics at the University Medical Center of the Johannes Gutenberg University. Available at http://www.kinderkrebsregister.de/extern/veroeffentlichun gen/jahresberichte/aktueller-jahresbericht/index.html?L=1; Wiangnon S, Veerakul G, Nuchprayoon I et al. (2011). Childhood cancer incidence and survival 2003–2005, Thailand: study from the Thai Pediatric Oncology Group. Asian Pac J Cancer Prev, 12:2215–2220. PMID:22296359; Fajardo-Gutiérrez A, Juárez-Ocaña S, González- Miranda G et al. (2007). Incidence of cancer in children residing in ten jurisdictions of the Mexican Republic: importance of the Cancer reg- istry (a population-based study). BMC Cancer, 7:68. PMID:17445267; Swaminathan R, Rama R, Shanta V (2008). Childhood cancers in Chennai, India, 1990–2001: incidence and survival. Int J Cancer, 122:2607–2611. http://dx.doi.org/10.1002/ijc.23428 PMID:18324630; Baade PD, Youlden DR, Valery PC et al. (2010). Trends in incidence of childhood cancer in Australia, 1983–2006. Br J Cancer, 102:620–626. http://dx.doi.org/10.1038/sj.bjc.6605503 PMID:20051948; Parkin DM, Ferlay J, Hamdi-Chérif M et al. (2003). Cancer in Africa: Epidemiology and Prevention. Chapter 5: Childhood cancer. Lyon: IARC (IARC Scientific Publications Series, No. 153), pp. 381–396; Bao PP, Zheng Y, Wang CF et al. (2009). Time trends and characteristics of childhood cancer among children age 0–14 in Shanghai. Pediatr Blood Cancer, 53:13–16. http://dx.doi.org/10.1002/pbc.21939 PMID:19260104.
1.3.3 & 1.3.4 From Steliarova-Foucher E, O’Callaghan M, Ferlay J et al. (2012). European Cancer Observatory: Cancer Incidence, Mortality, Prevalence and Survival in Europe, version 1.0. European Network of Cancer Registries, IARC. Available at http://eco.iarc.fr.
1.3.5 Adapted by permission of Oxford University Press. Stiller CA, Kroll ME, Eatock EM (2007). Survival from childhood cancer. In: Stiller CA, ed. Childhood Cancer in Britain: Incidence, Survival, Mortality. Oxford: Oxford University Press, pp. 131–204.
1.3.6 Compiled from GLOBOCAN 2012a and Cancer Incidence in Five Continents, Vol. Xb
1.3.7 WHO Mortality Database. Available at http://www.who.int/health info/mortality _data/en/ and http://www-dep.iarc.fr; and adapted from Pritchard-Jones K, Pieters R, Reaman GH et al. (2013). Sustaining innovation and improvement in the treatment of childhood cancer: les- sons from high-income countries. Lancet Oncol, 14:e95–e103. http:// dx.doi.org/10.1016/S1470-2045(13)70010-X PMID:23434338, with permission from Elsevier.
1.3.8 Courtesy of Hospital 57357 - Children’s Cancer Hospital, Egypt.
1.3.9 Courtesy of the Uganda Child Cancer Foundation.
2.1.1 © 2001 Germain Passamang Tabati, Courtesy of Photoshare.
2.1.2 Reproduced from Thun M, Peto R, Boreham J, Lopez AD (2012). Stages of the cigarette epidemic on entering its second century. Tob Control, 21:96–101. http://dx.doi.org/10.1136/tobacco control-2011-050294 PMID:22345230, with permission from BMJ Publishing Group Ltd.
610
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
2.1.3 Stephenie Hollyman/WHO.
2.1.4 Courtesy of Wikipedia user Morio. License: CC BY-SA 3.0.
2.1.5 © Massimo Mazzotta. All rights reserved.
2.1.6A Adapted from Jha P (2009). Avoidable global cancer deaths and total deaths from smoking. Nat Rev Cancer, 9:655–664. http:// dx.doi.org/10.1038/nrc2703 PMID:19693096, by permission from Macmillan Publishers Ltd. © 2009; and adapted from Guérin S, Hill C (2010). Cancer epidemiology in France in 2010: comparison with the USA [in French]. Bull Cancer, 97:47–54. http://dx.doi.org/10.1684/ bdc.2010.1013 PMID:19995688.
2.1.6B Compiled from ERC (2010). World Cigarette Reports 2010. Suffolk, UK: ERC Group; Economist Intelligence Unit (2011). Worldwide Cost of Living Survey. London: The Economist Group; Economist Intelligence Unit (2011). Marlboro cigarette and local ciga- rette prices, Worldwide Cost of Living Survey. London: The Economist Group; International Monetary Fund (2011). World Economic Outlook Database, April 2011 edition. Available at http://www.imf.org/external/ pubs/ft/weo/2011/01/weodata/index.aspx.
2.1.6C Compiled from Blecher EH (2011). The Economics of Tobacco Control in Low- and Middle-Income Countries [thesis]. Cape Town, South Africa: School of Economics, University of Cape Town.
2.2.1 Data from Giovino GA, Mirza SA, Samet JM et al.; GATS Collaborative Group (2012). Tobacco use in 3 billion individuals from 16 countries: an analysis of nationally representative cross-sectional household survey. Lancet, 380:668–679. http://dx.doi.org/10.1016/ S0140-6736(12)61085-X PMID:22901888
2.2.2 Data from Eriksen M, Mackay J, Ross H (2012). The Tobacco Atlas, 4th ed. Atlanta, GA: American Cancer Society, World Lung Foundation, http://www.tobaccoatlas.org/.
2.2.3 Reprinted from U.S. Department of Health and Human Services (2004). The Health Consequences of Smoking: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
2.2.4 © iStockphoto; U.S. Department of Health and Human Services (2004). The Health Consequences of Smoking: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; and IARC (2012). Personal habits and indoor combustions. IARC Monogr Eval Carcinog Risks Hum, 100E:1–575. PMID:23193840
2.2.5 Courtesy of the Library of Congress Prints and Photographs Division, Washington, DC.
2.2.6 Courtesy of Rue & Sue, accessed at www.wedheads.com.
B2.2.1 Adapted from Timofeeva MN, Hung RJ, Rafnar T et al.; Transdisciplinary Research in Cancer of the Lung (TRICL) Research Team (2012). Influence of common genetic variation on lung can- cer risk: meta-analysis of 14 900 cases and 29 485 controls. Hum Mol Genet, 21:4980– 4995. http://dx.doi.org/10.1093/hmg/dds334 PMID:22899653, by permission of Oxford University Press.
2.3.1 & 2.3.2 Data from Lim SS, Vos T, Flaxman AD et al. (2012). A comparative risk assessment of burden of disease and injury attribut- able to 67 risk factors and risk factor clusters in 21 regions, 1990– 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet, 380:2224–2260. http://dx.doi.org/10.1016/S0140-6736 (12)61766-8 PMID:23245609, © 2012 The Lancet.
2.3.3 © iStockphoto/Webphotographeer.
2.4.1 Reprinted from de Martel C, Ferlay J, Franceschi S et al. (2012). Global burden of cancers attributable to infections in 2008: a review and synthetic analysis. Lancet Oncol, 13:607–615. http://dx.doi. org/10.1016/S1470-2045(12)70137-7 PMID:22575588, © 2012, with permission from Elsevier.
2.4.2 Courtesy of Flickr user A.J. Cann. License: CC BY-SA 2.0.
2.4.3 Reprinted from Woodman CB, Collins SI, Young LS (2007). The natural history of cervical HPV infection: unresolved issues. Nat Rev Cancer, 7:11–22. PMID:17186016, by permission from Macmillan Publishers Ltd. © 2007.
2.4.4 Courtesy of Ed Uthman, MD, American Board of Pathology, accessed at commons.wikimedia.org.
2.4.5 Courtesy of C. Goldsmith, accessed at http://phil.cdc.gov/.
2.4.6 © 2002 Amira Roess, Courtesy of Photoshare.
2.5.1 Reprinted from Collaborative Group on Hormonal Factors in Breast Cancer (2012). Menarche, menopause, and breast cancer risk: individual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies. Lancet Oncol, 13:1141–1151. http://dx.doi.org/10.1016/S1470-2045(12)70425-4 PMID: 23084519, © 2012, with permission from Elsevier.
2.5.2 Reprinted from Yang XR, Figueroa JD, Falk RT et al. (2012). Analysis of terminal duct lobular unit involution in luminal A and basal breast cancers. Breast Cancer Res, 14:R64. http://dx.doi.org/10.1186/ bcr3170 PMID:22513288
2.5.3 Reprinted by permission from the American Association for Cancer Research: James RE, Lukanova A, Dossus L et al. (2011). Postmenopausal serum sex steroids and risk of hormone receptor- positive and -negative breast cancer: a nested case-control study. Cancer Prev Res (Phila), 4:1626–1635. http://dx.doi.org/10.1158/1940- 6207.CAPR-11-0090 PMID:21813404.
2.5.4 Reproduced from Fuhrman BJ, Schairer C, Gail MH et al. (2012). Estrogen metabolism and risk of breast cancer in postmenopausal women. J Natl Cancer Inst, 104:326–339. http://dx.doi.org/10.1093/ jnci/djr531 PMID:22232133
2.5.5 Reproduced from Trabert B, Wentzensen N, Yang HP et al. (2013). Is estrogen plus progestin menopausal hormone therapy safe with respect to endometrial cancer risk? Int J Cancer, 132:417– 426. http://dx.doi.org/10.1002/ijc.27623 PMID:22553145.
2.5.6 Reprinted from Chung SH, Franceschi S, Lambert PF (2010). Estrogen and ERalpha: culprits in cervical cancer? Trends Endocrinol Metab, 21:504–511. http://dx.doi.org/10.1016/j.tem.2010.03.005 PMID: 20456973, © 2010, with permission from Elsevier.
2.5.7 Compiled from Brinton LA, Richesson DA, Gierach GL et al. (2008). Prospective evaluation of risk factors for male breast cancer. J Natl Cancer Inst, 100:1477–1481. http://dx.doi.org/10.1093/jnci/djn329 PMID:18840816.
2.5.8 Reproduced from Roddam AW, Allen NE, Appleby P, Key TJ; Endogenous Hormones and Prostate Cancer Collaborative Group (2008). Endogenous sex hormones and prostate cancer: a collabora- tive analysis of 18 prospective studies. J Natl Cancer Inst, 100:170– 183. http://dx.doi.org/10.1093/jnci/djm323 PMID:18230794.
2.5.9 © 2011 Joydeep Mukherjee, Courtesy of Photoshare.
2.6.1 © iStockphoto/Fertnig.
2.6.2 © 2012 Thomas Mukoya/REUTERS, Courtesy of Photoshare.
2.6.3 Compiled from Berrington de Gonzalez A, Hartge P, Cerhan JR et al. (2010). Body-mass index and mortality among 1.46 million white adults. N Engl J Med, 363:2211–2219. http://dx.doi.org/10.1056/ NEJMoa1000367 PMID:21121834
World Cancer Report 611
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
2.6.4 Compiled from Reeves GK, Pirie K, Beral V et al. (2007). Cancer incidence and mortality in relation to body mass index in the Million Women Study: cohort study. BMJ, 335:1134 PMID:17986716; and Renehan AG, Tyson M, Egger M et al. (2008). Body-mass index and incidence of cancer: a systematic review and meta-analysis of pro- spective observational studies. Lancet, 371:569–578. http://dx.doi. org/10.1016/S0140-6736(08)60269-X PMID:18280327
2.6.5 © 2006 Nengah Kartika, Courtesy of Photoshare.
B2.6.1 Reprinted from Illner AK, Freisling H, Boeing H et al. (2012). Review and evaluation of innovative technologies for measuring diet in nutritional epidemiology. Int J Epidemiol, 41:1187–1203. http://dx.doi. org/10.1093/ije/dys105 PMID:22933652, by permission of Oxford University Press.
B2.6.2 Reprinted from Gilsing AM, Berndt SI, Ruder EH et al. (2012). Meat-related mutagen exposure, xenobiotic metabolizing gene poly- morphisms and the risk of advanced colorectal adenoma and cancer. Carcinogenesis, 33:1332–1339. http://dx.doi.org/10.1093/carcin/bgs158 PMID:22552404, by permission of Oxford University Press.
2.7.1 © 2006 Adam Scotti, Courtesy of Photoshare.
2.7.2 © 2009 sandipan majumdar, Courtesy of Photoshare.
2.7.3 Courtesy of Álvaro Daniel Gonzalez Lamarque, www.Morguefile. com.
2.7.4 © 2012 Farid Ahmed, Courtesy of Photoshare.
2.7.5 IARC photo library.
2.7.6 Courtesy of Oregon Department of Transportation. License: CC BY 2.0.
2.8.1 Reproduced from European Commission, Research Directorate- General, European Communities (2005). Health and Electromagnetic Fields: EU-funded research into the impact of electromagnetic fields and mobile telephones on health.
2.8.2 Reprinted from UN Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) (2010). Annex A: Medical radiation exposures. In: Sources and Effects of Ionizing Radiation, UNSCEAR 2008 Report to the General Assembly with Scientific Annexes, Vol. I. New York: UN. Available at http://www.unscear.org/docs/reports/2008/09-86753_ Report_2008_Annex_A.pdf
2.8.3 Courtesy of Clare McLean/University of Washington Medicine.
2.8.4 Courtesy of Kræftens Bekæmpelse (The Danish Cancer Society) and TrygFonden – a Danish Foundation. © 2013. All rights reserved.
2.8.5 Reprinted from Schüz J, Exposure to extremely low-frequency magnetic fields and the risk of childhood cancer: update of the epide- miological evidence. Prog Biophys Mol Biol, 107:339–342. http://dx.doi. org/10.1016/j.pbiomolbio.2011.09.008 PMID:21946043, © 2011, with permission from Elsevier.
2.8.6 Courtesy of Florentina Kindler.
2.9.1 © 2012 Jerome Salem 05, Courtesy of Photoshare.
2.9.2 Reprinted with permission from Brauer M, Amann M, Burnett RT et al. (2012). Exposure assessment for estimation of the global burden of disease attributable to outdoor air pollution. Environ Sci Technol, 46:652–660. http://dx.doi.org/10.1021/es2025752 PMID:22148428. © 2012 American Chemical Society.
2.9.3 © Ludolf Dahmen/Greenpeace.
2.9.4 Data source: WHO/Public Health Information and Geographic Information Systems. © WHO 2012. All rights reserved.
2.9.5 © 2006 Rakesh Yogal Shrestha, Courtesy of Photoshare.
2.9.6 U.S. Military, Department of Defense.
B2.9.1 Courtesy of Flickr user Katerha. License: CC BY 2.0.
B2.9.2 Courtesy of Charles O’Rear, U.S. Department of Agriculture.
2.10.1 © iStockphoto/Luca Zola. 2.10.2 © 2004 David Alexander, Courtesy of Photoshare.
2.10.3 Courtesy of Rhoda Baer, U.S. National Cancer Institute, ac- cessed at http://visualsonline.cancer.gov.
2.10.4 Courtesy of Wikipedia user Sauligno. License: CC-BY-SA-3.0
2.10.5 Courtesy of Wikipedia user Mr Hyde, via Wikimedia Commons.
2.11.1 Adapted with permission from Tracy E. Anderson, via www. Flickr.com. All rights reserved. © 2009.
2.11.2 © 1991 Bill Horn, Courtesy of Photoshare.
2.11.3 Reproduced from Pitt JI, Hocking AD (2009). Fungi and Food Spoilage, 3rd ed., Fig. 8.13, p. 305; with kind permission from Springer Science+Business Media B.V.
2.11.4 Reproduced from Qian X-Z (1996). Colour Pictorial Handbook of Chinese Herbs, by permission from the People’s Medical Publishing House.
3.1.1 Thomas J. Hudson.
3.1.2 Reprinted from Zhang J, Baran J, Cros A et al. (2011). International Cancer Genome Consortium Data Portal – a one-stop shop for can- cer genomics data. Database (Oxford), 2011:bar026. http://dx.doi. org/10.1093/database/bar026 PMID:21930502, by permission of Ox- ford University Press. © 2011.
3.1.3 & 3.1.4 Thomas J. Hudson.
3.1.5 Reprinted from Stratton MR, Campbell PJ, Futreal PA (2009). The cancer genome. Nature, 458:719–724. http://dx.doi.org/10.1038/na ture07943 PMID:19360079, by permission from Macmillan Publishers Ltd. © 2009.
3.1.6 Joint Genome Institute, U.S. Department of Energy Genomic Science Program, accessed at http://genomicscience.energy.gov/.
B3.1.1 Magali Olivier.
3.2.1 Reprinted from Manolio TA, Collins FS, Cox NJ et al. (2009). Finding the missing heritability of complex diseases. Nature, 461:747– 753. http://dx.doi.org/10.1038/nature08494 PMID:19812666, by per- mission from Macmillan Publishers Ltd. © 2009; and adapted from McCarthy MI, Abecasis GR, Cardon LR et al. (2008). Genome-wide as- sociation studies for complex traits: consensus, uncertainty and chal- lenges. Nat Rev Genet, 9:356–369. http://dx.doi.org/10.1038/nrg2344 PMID:18398418, by permission from Macmillan Publishers Ltd. © 2008.
3.2.2 Reprinted from Garcia-Closas M, Chanock S (2008). Genetic susceptibility loci for breast cancer by estrogen receptor status. Clin Cancer Res, 14:8000–8009. http://dx.doi.org/10.1158/1078-0432.CCR- 08-0975 PMID:19088016, by permission from the American Associ- ation for Cancer Research.
3.2.3 Stephen J. Chanock.
B3.2.1 Adapted from Hoeijmakers JH (2001). Genome mainte- nance mechanisms for preventing cancer. Nature, 411:366–374. PMID:11357144, by permission from Macmillan Publishers Ltd. © 2001.
3.3.1 Magdalena B. Wozniak and Paul Brennan.
3.3.2 Reprinted from Trollope AF, Gutièrrez-Mecinas M, Mifsud KR et al. (2012). Stress, epigenetic control of gene expression and mem- ory formation. Exp Neurol, 233:3–11. http://dx.doi.org/10.1016/j.exp neurol.2011.03.022 PMID:21466804, © 2012, with permission from Elsevier.
3.3.3A Reprinted from Bastien RR, Rodríguez-Lescure Á, Ebbert MT et al. (2012). PAM50 breast cancer subtyping by RT-qPCR and concordance with standard clinical molecular markers. BMC Med Genomics, 5:44. http://dx.doi.org/10.1186/1755-8794-5-44 PMID:23035882, © 2012 Bastien et al.; licensee BioMed Central Ltd.
612
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
3.3.3B Reprinted from van ’t Veer LJ, Dai H, van de Vijver MJ et al. (2002). Gene expression profiling predicts clinical outcome of breast cancer. Nature, 415:530–536. http://dx.doi.org/10.1038/415530a PMID:11823860, by permission from Macmillan Publishers Ltd. © 2002.
3.3.4 Reprinted from Esquela-Kerscher A, Slack FJ (2006). Oncomirs – microRNAs with a role in cancer. Nat Rev Cancer, 6:259–269. http:// dx.doi.org/10.1038/nrc1840 PMID:16557279, by permission from Macmillan Publishers Ltd. © 2006.
3.3.5 Magdalena B. Wozniak and Paul Brennan.
3.3.6 Rhoda Baer, courtesy of U.S. National Institutes of Health (NIH).
B3.3.1 Reprinted from Bouwman P, Jonkers J (2012). The effects of de- regulated DNA damage signalling on cancer chemotherapy response and resistance. Nat Rev Cancer, 12:587–598. http://dx.doi.org/10.1038/ nrc3342 PMID:22918414, by permission from Macmillan Publishers Ltd. © 2012.
3.4.1 & 3.4.2 Reprinted from Sawan C, Vaissière T, Murr R, Herceg Z (2008). Epigenetic drivers and genetic passengers on the road to cancer. Mutat Res, 642:1–13. http://dx.doi.org/10.1016/j.mrfmmm.2008.03.002 PMID:18471836, © 2008, with permission from Elsevier.
3.4.3 & 3.4.4 Toshikazu Ushijima and Zdenko Herceg.
3.5.1 Courtesy of Collectif National des Associations d’Obèses (CNAO, France), www.cnao.fr.
3.5.2–3.5.4 Augustin Scalbert and Isabelle Romieu.
3.6.1 Reprinted from O’Connor TP, Crystal RG (2006). Genetic medi- cines: treatment strategies for hereditary disorders. Nat Rev Genet, 7:261–276. http://dx.doi.org/10.1038/nrg1829 PMID:16543931, by per- mission from Macmillan Publishers Ltd. © 2006.
3.6.2 Courtesy of Freda Miller, Hospital for Sick Children (Toronto, Canada).
3.6.3 & 3.6.4 Hector Hernandez Vargas.
3.6.5 Courtesy of Johannes Jungverdorben, Life & Brain Center, Bonn, www.eurostemcell.org.
3.6.6 Reprinted from Reya T, Morrison SJ, Clarke MF et al. (2001). Stem cells, cancer, and cancer stem cells. Nature, 414:105–111. http:// dx.doi.org/ 10.1038/35102167 PMID:11689955, by permission from Macmillan Publishers Ltd. © 2001.
3.7.1 Robert R. Langley.
3.7.2 Figure courtesy of Zhi-Yuan Chen, Sun Yat-sen University Cancer Center; reprinted from Qin L, Bromberg-White JL, Qian CN (2012). Opportunities and challenges in tumor angiogenesis re- search: back and forth between bench and bed. Adv Cancer Res, 113:191–239. http://dx.doi.org/10.1016/B978-0-12-394280-7.00006-3 PMID:22429856, © 2012, with permission from Elsevier.
3.7.3 Reprinted from Davis DW, McConkey DJ, Abbruzzese JL et al. (2003). Surrogate markers in antiangiogenesis clinical tri- als. Br J Cancer, 89:8–14. http://dx.doi.org/10.1038/sj.bjc.6601035 PMID:12838293, by permission from Macmillan Publishers Ltd on be- half of Cancer Research UK. © 2003.
3.7.4 Reprinted from Langley RR, Fidler IJ (2007). Tumor cell-organ microenvironment interactions in the pathogenesis of cancer metasta- sis. Endocr Rev, 28:297–321. http://dx.doi.org/10.1210/er.2006-0027 PMID:17409287. © 2007, The Endocrine Society.
3.7.5 Robert R. Langley.
3.8.1A & B Image from the RCSB Protein Data Bank, September 2008 Molecule of the Month feature by David Goodsell; http://dx.doi. org/10.2210/rcsb_pdb/mom_2008_9.
3.8.1C Courtesy of Thomas Bauer, Medical University of Vienna.
3.8.2–3.8.4 Courtesy of Thomas Bauer, Medical University of Vienna.
3.9.1 Holger Moch.
3.9.2 Courtesy of Wikipedia user Patho. License: CC-BY-SA-3.0.
3.9.3 Adapted from Pardoll DM (2012). Immunology beats cancer: a blueprint for successful translation. Nat Immunol, 13:1129–1132. http://dx.doi.org/10.1038/ni.2392 PMID:23160205, by permission from Macmillan Publishers Ltd. © 2012.
3.9.4 Reprinted from Mellman I, Coukos G, Dranoff G (2011). Cancer immunotherapy comes of age. Nature, 480:480– 489. http://dx.doi. org/10.1038/nature10673 PMID:22193102, by permission from Macmillan Publishers Ltd. © 2011.
4.1.1 Reproduced with permission from the association Droits des Non-Fumeurs (Paris, France), www.dnf.asso.fr.
4.1.2 Reproduced with permission from The GTSS Atlas: Global Youth Tobacco Survey, 2009 © Myriad Editions www.myriadeditions.com.
4.1.3 Reproduced from IARC (2008). IARC Handbooks of Cancer Prevention, Vol. 12: Tobacco Control: Methods for Evaluating Tobacco Control Policies. Lyon: IARC.
4.1.4 Reproduced with permission from the Australian National Preventive Health Agency.
4.1.5 Reproduced with permission from NYC Health and the World Lung Foundation.
4.1.6 WHO © 2013.
B4.1.1 Simon Chapman.
4.2.1 © 2004 Syed Ziaul Habib Roobon, Courtesy of Photoshare.
4.2.2 & 4.2.3 Rena R. Wing and Kathryn R. Middleton.
4.2.4 Courtesy of jusben, www.Morguefile.com.
4.2.5 Crown Copyright 2013, accessed at www.nhs.uk/Change4Life, a website of the National Health Service (NHS).
4.2.6 U.S. Air Force photo by Mike Kaplan, www.defenseimagery.mil.
B4.2.1 Adapted from Pischon T, Boeing H, Hoffmann K et al. (2008). General and abdominal adiposity and risk of death in Europe. N Engl J Med, 359:2105–2120. http://dx.doi.org/10.1056/NEJMoa0801891 PMID:19005195, with permission from the Massachusetts Medical Society.
4.3.1 Australian National Tobacco Campaign, reproduced with per- mission from the Australian Government Department of Health and Ageing.
4.3.2 Courtesy of the Asociación Española Contra el Cáncer (Spanish Association Against Cancer), www.aecc.es.
4.3.3 Courtesy of Association Lalla Salma de Lutte contre le Cancer (ALSC, Morocco).
4.3.4 Courtesy of Hong Kong Cancer Fund. All rights reserved (2013).
4.3.5 Courtesy of Breast Cancer Foundation, Singapore. Designed by DDB.
4.3.6 A campaign of the Centers for Disease Control and Prevention, Division of Adolescent and School Health www.VERBnow.com.
4.4.1 Courtesy of Flickr user Cory Doctorow. Permission is granted to copy, distribute and/or modify this image under the terms of the Attribution-Share Alike 2.0 Generic license (http://creativecommons. org/licenses/by-sa/2.0/deed.en).
4.4.2 Reproduced from “BRASIL – Advertências Sanitárias nos Produtos de Tabaco – 2009”, accessed at www.inca.gov.br.
4.4.3 A campaign of Children’s Healthcare of Atlanta, accessed at www.strong4life.com.
4.4.4 Courtesy of Biswarup Ganguly via Wikimedia Commons. License: CC-BY-SA-3.0.
World Cancer Report 613
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
4.4.5 Reproduced with permission from the Drug and Alcohol Office, Western Australia.
4.5.1 Courtesy of Álvaro Daniel Gonzalez Lamarque, www.Morguefile. com.
4.5.2 WHO © 2013.
4.5.3 © State of New South Wales through the Parliament, 2012.
4.5.4 © 2007 Frederick Noronha, Courtesy of Photoshare.
4.5.5 © Greenpeace/Karen Robinson, accessed at www.greenpeace. org/toxics.
B4.5.1 Rafael Moreira Claro.
4.6.1 Courtesy of Matthias Trischler via Wikipedia. License: CC-BY- SA-3.0.
4.6.2 IARC photo library.
4.6.3 Reprinted from WHO/UNICEF coverage estimates, 1980–2005, as of August 2006, 192 WHO Member States.
4.6.4 Reproduced from Tabrizi SN, Brotherton JM, Kaldor JM et al. (2012). Fall in human papillomavirus prevalence following a nation- al vaccination program. J Infect Dis, 206:1645–1651. http://dx.doi. org/10.1093/infdis/jis590 PMID:23087430, by permission of Oxford University Press.
B4.6.1 Tandin Dorji and the staff of the Jigme Dorji Wangchuck National Referral Hospital.
B4.6.2 © 2004 Eileen Dietrich, Courtesy of Photoshare.
4.7.1 Adapted from de Koning HJ (2009). The mysterious mass(es). [Inaugural address, Professor of Screening Evaluation.] Rotterdam, The Netherlands: Erasmus MC.
4.7.2 Adapted from Vaccarella S, Lortet-Tieulent J, Plummer M et al. (2013). Worldwide trends in cervical cancer incidence: Impact of screen- ing against changes in disease risk factors. Eur J Cancer, 49:3262– 3273. http://dx.doi.org/10.1016/j.ejca.2013.04.024 PMID:23751569, with permission from Elsevier.
4.7.3 & 4.7.4 IARC Screening Group.
4.7.5 Peter B. Dean/IARC.
4.7.6 Rengaswamy Sankaranarayanan/IARC.
4.7.7 IARC Screening Group.
4.7.8 Adapted from von Karsa L (1995). Mammography screening – comprehensive, population-based quality assurance is required! [in German] Z Allgemeinmed, 71:1863–1867, with permission of Deutscher Ärzte-Verlag.
4.8.1 Krittika Pitaksaringkarn/IARC.
4.8.2 UNFPA/Omar Gharzeddine.
4.8.3 Courtesy of Mark B. Brown, from the Lutheran World Foundation.
4.8.4 Eric Lucas/IARC.
4.8.5 Courtesy of Osaka Center for Cancer and Cardiovascular Diseases Prevention, Japan.
5.1.1 © 2008 Sandipan Majumdar, Courtesy of Photoshare.
5.1.2 & 5.1.3 Elisabeth Brambilla and William D. Travis.
5.1.4 Reprinted from Heist RS, Engelman JA (2012). SnapShot: non- small cell lung cancer. Cancer Cell, 21:448. http://dx.doi.org/10.1016/j. ccr.2012.03.007 PMID:22439939, with permission from Elsevier.
5.1.5 Reprinted from Govindan R, Ding L, Griffith M et al. (2012) Genomic landscape of non-small cell lung cancer in smokers and never-smokers. Cell, 150:1121–1134. http://dx.doi.org/10.1016/j.cell. 2012.08.024 PMID:22980976, with permission from Elsevier.
5.1.6 & 5.1.7 Elisabeth Brambilla and William D. Travis.
B5.1.1 Avrum Spira.
B5.1.2 Figure compiled from Loft S, Voboda P, Kasai H (2005). Prospective study of 8-oxo-7,8-dihydro-20-deoxyguanosine excretion and the risk of lung cancer. Carcinogenesis, 27:1245–1250. http:// dx.doi.org/10.1093/carcin/bgi313 PMID:16364924
5.2.1A & B & C Sunil R. Lakhani.
5.2.1D Stuart J. Schnitt.
5.2.2A Stuart J. Schnitt.
5.2.2B Puay Hoon Tan, Singapore General Hospital, Singapore.
5.2.2C Sunil R. Lakhani.
5.2.2D Puay Hoon Tan, Singapore General Hospital, Singapore.
5.2.2E Sunil R. Lakhani.
5.2.2F Jorge S. Reis-Filho, The Breakthrough Breast Cancer Research Centre, Institute of Cancer Research, London, United Kingdom.
5.2.2G Gianni Bussolati, University of Turin, Turin, Italy.
5.2.2H Johannes L. Peterse.
5.2.3 From the University of Alabama at Birmingham Department of Pathology PEIR Digital Library © (http://peir.net).
5.2.4 Reprinted from Sørlie T, Wang Y, Xiao C et al. (2006). Distinct mo- lecular mechanisms underlying clinically relevant subtypes of breast cancer: gene expression analyses across three different platforms. BMC Genomics, 7:127. http://dx.doi.org/10.1186/1471-2164-7-127 PMID:16729877
5.2.5 Reprinted from Stephens PJ, Tarpey PS, Davies H et al.; Oslo Breast Cancer Consortium (OSBREAC) (2012). The landscape of cancer genes and mutational processes in breast cancer. Nature, 486:400– 404. http://dx.doi.org/10.1038/nature11017 PMID:22722201, by permission from Macmillan Publishers Ltd. © 2012.
5.2.6 Courtesy of Petr Kratochvil, accessed at www.publicdomain pictures.net.
5.2.7 Courtesy of Bill Branson, accessed at https://visualsonline. cancer.gov/ (National Cancer Institute).
B5.2.1 From Neilson HK, Friedenreich CM, Brockton NT, Millikan RC (2009). Physical activity and postmenopausal breast cancer: proposed biologic mechanisms and areas for future research. Cancer Epidemiol Biomarkers Prev, 18:11–27. http://dx.doi.org/10.1158/1055-9965.EPI- 08-0756 PMID:19124476
5.3.1 From the University of Alabama at Birmingham Department of Pathology PEIR Digital Library © (http://peir.net).
5.3.2 © 2008 Khalid Mahmood Raja, Courtesy of Photoshare.
5.3.3 Elizabeth A. Montgomery.
5.3.4 Elizabeth A. Montgomery.
5.3.5 Reprinted from Dulak AM, Stojanov P, Peng S et al. (2013). Exome and whole-genome sequencing of esophageal adenocarcinoma iden- tifies recurrent driver events and mutational complexity. Nat Genet, 45:478– 486. http://dx.doi.org/10.1038/ng.2591 PMID:23525077, by permission from Macmillan Publishers Ltd. © 2013.
5.3.6 Adapted from Lao-Sirieix P, Fitzgerald RC (2012). Screening for oesophageal cancer. Nat Rev Clin Oncol, 9:278–287. http://dx.doi. org/10.1038/nrclinonc.2012.35 PMID:22430857, by permission of Macmillan Publishers Ltd. © 2012.
5.4.1 Fátima Carneiro.
5.4.2 Courtesy of Wikipedia user Daderot. This file is made avail- able under the Creative Commons CC0 1.0 Universal Public Domain Dedication (http://creativecommons.org/publicdomain/zero/1.0/deed. en).
614
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
5.4.3 Fátima Carneiro.
5.4.4 Courtesy of Russell Lee, from the U.S. National Archives and Records Administration.
5.4.5 Reproduced from Deng N, Goh LK, Wang H et al. (2012). A com- prehensive survey of genomic alterations in gastric cancer reveals systematic patterns of molecular exclusivity and co-occurrence among distinct therapeutic targets. Gut, 61:673–684. http://dx.doi.org/10.1136/ gutjnl-2011-301839 PMID:22315472, with permission from BMJ Publishing Group Ltd.
B5.4.1 Il Ju Choi.
5.5.1 Fred T. Bosman.
5.5.2 Courtesy of Dale C. Snover, Department of Pathology, Fairview Southdale Hospital, Edina, MN, USA.
5.5.3 Reproduced from Bosman FT (2013). Molecular pathology of colo- rectal cancer. In: Chung L, Eble J, eds. Molecular Surgical Pathology. Springer, pp. 1–17, with kind permission from Springer+Business Media B.V.
5.5.4 Courtesy of University of Lausanne Medical Centre, with per- mission from Fred T. Bosman.
5.6.1 Courtesy of Banchob Sripa, Faculty of Medicine, Khon Kaen University, Khon Kaen, Thailand.
5.6.2 © 2010 Somenath Mukhopadhyay, Courtesy of Photoshare.
5.6.3 Courtesy of Young Nyun Park, Yonsei University College of Med- icine, Seoul, Republic of Korea.
5.6.4 Courtesy of Hirohisa Yano, Kurume University School of Med- icine, Fukuoka-ken, Japan.
5.6.5A Reproduced with permission International Consensus Group for Hepatocellular Neoplasia (2009). Pathologic diagnosis of early hepatocellular carcinoma: a report of the International Consensus Group for Hepatocellular Neoplasia. Hepatology, 49:658–664. http:// dx.doi.org/10.1002/hep.22709 PMID:19177576
5.6.5B Courtesy of Alexander Kagen, Albert Einstein College of Medicine, New York NY, USA.
5.6.6 Reproduced from Kan Z, Zheng H, Liu X et al. (2013). Whole- genome sequencing identifies recurrent mutations in hepatocellular carcinoma. Genome Res, 23:1422–1433. http://dx.doi.org/10.1101/ gr.154492.113 PMID:23788652. This article is licensed under a Creative Commons License.
B5.6.1 © 2009 Nephron, Permission is granted to copy, distribute and/ or modify this image under the terms of the Attribution-Share Alike 3.0 Unported licence (https://creativecommons.org/licenses/by-sa/3.0/ deed.en).
5.7.1 Courtesy of the National Cancer Institute.
5.7.2 Ralph H. Hruban.
5.7.3 Reprinted from Biankin AV, Waddell N, Kassahn KS et al.; Australian Pancreatic Cancer Genome Initiative (2012). Pancreatic cancer genomes reveal aberrations in axon guidance pathway genes. Nature, 491:399– 405. http://dx.doi.org/10.1038/nature11547 PMID:23103869, by permission from Macmillan Publishers Ltd, 2012.
5.7.4 Ralph H. Hruban.
5.7.5 Courtesy of the Armed Forces Institute of Pathology (AFIP Atlas of Tumor Pathology).
5.7.6 Ralph H. Hruban.
5.8.1 Used with permission from Amirsys Publishing; Thompson LD, Wenig BM (2011). Diagnostic Pathology: Head and Neck, © 2012.
5.8.2 Courtesy of Jason C. Fowler, Meadville Medical Center, Meadville, PA, USA.
5.8.3 iStockphoto: Krakozawr © 2008.
5.8.4 Adapted from Argiris A, Karamouzis MV, Raben D, Ferris RL (2008). Head and neck cancer. Lancet, 371:1695–1709, http://dx.doi. org/10.1016/S0140-6736(08)60728-X PMID:18486742, © 2012, with permission from Elsevier.
5.8.5 Lester D.R. Thompson.
5.8.6 Lester D.R. Thompson.
5.9.1 Courtesy of Tony Alter via Flickr. License: CC BY 2.0.
5.9.2 © 2006 Rezaul Haque, Courtesy of Photoshare.
5.9.3 & 5.9.4 Holger Moch
5.9.5 Reprinted from Varela I, Tarpey P, Raine K et al. (2011). Exome sequencing identifies frequent mutation of the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature, 469:539–542. http://dx.doi. org/10.1038/nature09639 PMID:21248752, by permission from Macmillan Publishers Ltd. © 2011.
5.10.1 © 2001 Jean Sack, Courtesy of Photoshare.
5.10.2 Courtesy of KGH via Wikipedia. License: CC BY-SA 3.0.
5.10.3 Reproduced from Eble JN, Sauter G, Epstein JI, Sesterhenn IA, eds (2004). Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. Lyon: IARC, p. 106.
5.10.4 © 2002 Margaret D’Adamo, Courtesy of Photoshare.
5.10.5 Reprinted with permission from Iyer G, Al-Ahmadie H, Schultz N et al. (2013). Prevalence and co-occurrence of actionable genomic alterations in high-grade bladder cancer. J Clin Oncol, 31:3133–3140. http://dx.doi.org/10.1200/JCO.2012.46.5740 PMID:23897969. © 2013 American Society of Clinical Oncology. All Rights Reserved.
5.11.1 Courtesy of Jeremy Sheldon (“Pay No Mind”) via Flickr. License: CC BY-NC-ND 2.0.
5.11.2 & 5.11.3 Peter A. Humphrey.
5.11.4 Reprinted from Epstein JI, Cubilla AL, Humphrey PA (2011). Tumors of the Prostate Gland, Seminal Vesicles, Penis, and Scrotum (AFIP Atlas of Tumor Pathology Series 4, Fascicle 14). Washington, DC: American Registry of Pathology and Armed Forces Institute of Pathology.
5.11.5 Reprinted with permission from Rubin MA, Maher CA, Chinnaiyan AM (2011). Common gene rearrangements in prostate cancer. J Clin Oncol, 29:3659–3668. http://dx.doi.org/10.1200/JCO.2011.35.1916 PMID:21859993. © 2013 American Society of Clinical Oncology. All rights reserved.
5.11.6 & 5.11.7 Peter A. Humphrey.
5.12.1 © 2012 Akintunde Akinleye/NURHI, Courtesy of Photoshare.
5.12.2 Jaime Prat.
5.12.3 Adapted from Crosbie EJ, Einstein MH, Franceschi S, Kitchener HC (2013). Human papillomavirus and cervical cancer. Lancet, 382:889–899. http://dx.doi.org/10.1016/S0140-6736(13)60022-7 PMID: 23618600, © 2013, with permission from Elsevier.
5.12.4 This image was published in Robboy SJ, Mutter GL, Prat J et al., eds (2009). Robboy’s Pathology of the Female Reproductive Tract, 2nd ed., p. 191, © Elsevier 2009.
5.12.5 & 5.12.6 Jaime Prat.
5.12.7 Jaime Prat.
5.12.8 Reprinted from Catasus L, Gallardo A, Prat J (2009). Molecular genetics of endometrial carcinoma. Diagn Histopathol, 15: 554–563, http://dx.doi.org/10.1016/j.mpdhp.2009.09.002, © 2009, with permis- sion from Elsevier.
5.12.9 Jaime Prat.
5.13.1 & 5.13.2 Courtesy of Harald Stein, Pathodiagnostik Berlin, Germany.
World Cancer Report 615
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
5.13.3 Reproduced from Steidl C, Gascoyne RD (2011). The mo- lecular pathogenesis of primary mediastinal large B-cell lymphoma. Blood, 118:2659–2669. http://dx.doi.org/10.1182/blood-2011-05-3 26538 PMID:21700770, with permission from the American Society of Hematology.
5.13.4 Courtesy of Peter Hesseling, Faculty of Medicine and Health Sciences, Stellenbosch University, South Africa.
5.13.5 Reproduced from Mullighan CG (2012). Molecular genetics of B-precursor acute lymphoblastic leukemia. J Clin Invest, 122:3407– 3415. http://dx.doi.org/10.1172/JCI61203 PMID:23023711, with per- mission from the American Society for Clinical Investigation (ASCI).
B5.13.1 Adapted from Szczepański T, Orfão A, van der Velden VH et al. (2001). Minimal residual disease in leukaemia patients. Lancet Oncol, 2:409– 417. http://dx.doi.org/10.1016/S1470-2045(00)00418-6 PMID:11905735. © 2001 with permission from Elsevier.
5.14.1 Courtesy of CDC/Carl Washington, M.D., Emory Univ. School of Medicine; Mona Saraiya, MD, MPH, accessed at phil.cdc.gov/phil.
5.14.2 From the University of Alabama at Birmingham Department of Pathology PEIR Digital Library © (http://peir.net).
5.14.3 iStockphoto: aydinmutlu © 2011.
5.14.4 Reprinted from Vultur A, Herlyn M (2013). SnapShot: melano- ma. Cancer Cell, 23:706. http://dx.doi.org/10.1016/j.ccr.2013.05.001 PMID:23680152, © 2013, with permission from Elsevier.
5.14.5 iStockphoto: .shock © 2010.
5.15.1 & 5.15.4 Frank Weber.
5.15.2 Courtesy of Gil Tudor, IAEA Division of Public Information.
5.15.3 & 5.15.5 Reproduced from Weber F, Eng C (2005). Gene- expression profiling in differentiated thyroid cancer – a viable strategy for the practice of genomic medicine? Future Oncol, 1:497–510. http:// dx.doi.org/10.2217/14796694.1.4.497 PMID:16556026 with permission from Future Medicine Ltd.
5.15.6 Courtesy of Susanne Reger-Tan and Kurt W. Schmid, both of Universitätsklinikum Essen, Germany.
5.16.1 Courtesy of Angela Leuker, IAEA Division of Public Information.
5.16.2–5.16.4 Paul Kleihues.
5.16.5 Reprinted from Kotliarova S, Fine HA (2012). SnapShot: glioblas- toma multiforme. Cancer Cell, 21:710, e1. http://dx.doi.org/10.1016/j. ccr.2012.04.031 PMID:22624719, © 2012, with permission from Elsevier.
5.16.6 Hiroko Ohgaki.
6.1.1 WHO.
6.1.2 WHO, 2010. All rights reserved.
6.1.3 Courtesy of Institut National de Recherche et de Sécurité (INRS, France; www.inrs.fr).
6.1.4 Courtesy of the Iowa Department of Public Health, accessed at www.idph.state.ia.us.
6.1.5 “Nobody’s Immune to Breast Cancer” campaign by DDB MOÇAMBIQUE for the ALCC (Associação da Luta Contra o Cancro; Association for the Fight Against Cancer, Mozambique).
6.2.1 © 2009 Joanne Katz, Courtesy of Photoshare.
6.2.2 © 2004 Philippe Blanc, Courtesy of Photoshare.
6.2.3 Roland Dray/IARC.
6.2.4 WHO (2006). Cancer Control: Knowledge into Action. WHO Guide for Effective Programmes. Module 1: Planning. Geneva: WHO.
B6.2.1 Adapted with permission from WHO. WHO/IARC/APHRC (2012). Prevention of cervical cancer through screening using visual inspection with acetic acid (VIA) and treatment with cryotherapy. A demonstration project in six African countries: Malawi, Madagascar, Nigeria, Uganda, the United Republic of Tanzania, and Zambia WHO, Nigeria; Fig. 2, p. 11.
B6.2.2 & B6.2.3 Reprinted from Ezzati M (2004). Indoor air quality in developing nations, In: Encyclopaedia of Energy, Vol. 3, pp. 343–350, © 2004, with permission from Elsevier.
6.3.1 © Petr Pavlicek/International Atomic Energy Agency.
6.3.2 © Angela Leuker/International Atomic Energy Agency.
6.3.3 © David Kinley/International Atomic Energy Agency.
6.3.4 Courtesy of the National Cancer Centre of Mongolia.
6.3.5 © Elio Omobono/International Atomic Energy Agency.
6.4.1 © 2012 Wikipedia user Armigo. License: CC-BY-SA-3.0.
6.4.2 NIH/Coriell Institute for Medical Research.
6.4.3 Roland Dray/IARC, with permission from Marie-Pierre Cros.
6.4.4 Reproduced from Pukkala E (2010). Nordic biological speci- men bank cohorts as basis for studies of cancer causes and control – quality control of cohorts with more than 2 million sample donors and 100,000 prospective cancers. In: Dillner J, ed. Methods in Biobanking. New York: Springer, pp. 61–112, with kind permission from Springer Science+Business Media.
6.4.5 & 6.4.6 IARC, Joakim Dillner.
6.4.7 Courtesy of Kurt Zatloukal ([email protected]) and Biobanking and Biomolecular Resources Research Infrastructure (www.BBMRI.eu).
6.5.1 UN Photo/Marco Castro.
6.5.2 © Union for International Cancer Control (UICC).
6.5.3 © 2012 Srikrishna Sulgodu Ramachandra, Courtesy of Photoshare.
6.5.4 Courtesy www.MoonshineMovies.com.
6.5.5 Courtesy of Cancer Council Australia, www.cancer.org.au.
6.6.1 Health warning images and plain packaging designs used with the permission of the Australian Government.
6.6.2 Figure created by Martha Maurer, copyright Pain & Policy Studies Group, UWCCC/home of WHO Collaborating Center, 2013. Opioid consumption data from International Narcotics Control Board; population data from the United Nations World Population Prospects 2010 Revision; country income level from World Bank Classification of Countries by Income level, http://data.worldbank.org/about/ country-classifications/country-and-lending-groups.
6.6.3 Figure created by Martha Maurer, copyright Pain & Policy Studies Group, UWCCC/home of WHO Collaborating Center, 2012. Opioid consumption data from International Narcotics Control Board (values represent the aggregate morphine equivalence consump- tion of fentanyl, hydromorphone, morphine, oxycodone and pethi- dine); cancer deaths from Ferlay J, Shin HR, Bray F et al. (2010). GLOBOCAN 2008 v1.2, Cancer Incidence and Mortality Worldwide: IARC CancerBase No. 10 [Internet]. Lyon: IARC. Available at http:// globocan.iarc.fr (cancer death estimates calculated by applying the 2008 crude death rate per 100 000 to the 2010 population to esti- mate raw number of cancer deaths for 2010); population data from the United Nations World Population Prospects 2010 Revision.
6.6.4 Courtesy of Barb Gannon, via Flickr. All rights reserved. © 2012.
6.7.1 © 2012 Victoria A Smith, Courtesy of Photoshare.
6.7.2 © 2007 Pradeep Tewari, Courtesy of Photoshare.
616
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
6.7.3 © 2003 Justin Fahey, Courtesy of Photoshare.
6.7.4 Courtesy of James Gathany and Judy Schmidt, U.S. Centers for Disease Control and Prevention.
P3.1 Adapted from zur Hausen H, de Villiers EM (2005). Virus target cell conditioning model to explain some epidemiologic characteristics of childhood leukemias and lymphomas. Int J Cancer, 115:1–5. http:// dx.doi.org/10.1002/ijc.20905 PMID:15688417 This material is repro- duced with permission of John Wiley & Sons, Inc.
P4.1 Reproduced with permission from Cambridge University Library.
P4.2 Adapted from Potter NE, Ermini L, Papaemmanuil E et al. (2013). Single cell mutational profiling and clonal phylogeny in cancer. Genome Res, [epub ahead of print], http://dx.doi.org/10.1101/gr.159913.113 PMID:24056532, with permission from Cold Spring Harbor Laboratory Press.
P4.3 Reproduced from Greaves M (2000). Cancer: The Evolutionary Legacy. With permission from Oxford University Press.
P4.4A Mel Greaves.
P4.4B Reproduced from Costanzo M, Baryshnikova A, Bellay J et al. (2010). The genetic landscape of a cell. Science, 327:425– 431. http:// dx.doi.org/10.1126/science.1180823 PMID:20093466; with permis- sion from the American Association for the Advancement of Science. All rights reserved.
P4.5 Mel Greaves.
P7.1 Adapted by permission from BMJ Publishing Group Limited. Doll R, Peto R, Boreham J, Sutherland I (2004). Mortality in relation to smok- ing: 50 years’ observations on male British doctors. BMJ, 328:1519– 1527. http://dx.doi.org/10.1136/bmj.38142.554479.AE PMID:15213107, © 2004.
P7.2 & P7.3 Reprinted from Pirie K, Peto R, Reeves GK et al.; Million Women Study Collaborators (2013). The 21st century hazards of smoking and benefits of stopping: a prospective study of one million women in the UK. Lancet, 381:133–141. http://dx.doi.org/10.1016/ S0140-6736(12)61720-6 PMID:23107252, © 2013, with permission from Elsevier.
P7.4–P7.9 Compiled from Peto R, Lopez AD, Boreham J et al. (1992). Mortality from tobacco in developed countries: indirect estimation from national vital statistics. Lancet, 339:1268–1278. http://dx.doi. org/10.1016/0140-6736(92)91600-D PMID:1349675; Peto R, Lopez AD, Boreham J et al. (1994). Mortality from Smoking in Developed Countries 1950 –2000: Indirect Estimates from National Vital Statistics. Oxford: Oxford University Press; WHO mortality data and UN popula- tion estimates.
page xi Courtesy of Russell E. White, Tenwek Mission Hospital, Kenya.
page 15 GLOBOCAN 2012a page 81 Clockwise from top left: Reproduced from www.pixa bay.com; © 2003 Mukunda Bogati, Courtesy of Photoshare; IARC/ Roland Dray; Courtesy of Flickr user Didier Vidal. License: CC BY-SA 2.0; Courtesy of Flickr user Alexis O’Toole. License: CC BY-SA 2.0; Courtesy of Aram Dulyan, Wikimedia Commons.
page 183 Clockwise from top left: Rhoda Baer, National Institutes of Health Clinical Centre; Lance Liotta Laboratory, National Cancer Institute; Susan Arnold, provided by Raowf Guirguis, National Cancer Institute; Stanley J. Korsmeyer, Metabolism Branch, Section of Cellular Immunology, National Cancer Institute; National Institutes of Health; Basic Research Laboratory, Frederick Cancer Research Facility, National Cancer Institute.
page 267 Clockwise from top left: © 2005 Emily J. Phillips, Courtesy of Photoshare; Courtesy of Breast Cancer Foundation, Singapore. Designed by DDB; Courtesy of Association Lalla Salma de Lutte contre le Cancer (ALSC, Morocco); Rhoda Baer, National Cancer Institute; WHO © 2013; Courtesy of Flickr user Cory Doctorow, ac- cessed at www.flickr.com.
page 347 Clockwise from top left: Dr Maria Tsokos, National Cancer Institute; Courtesy of Ed Uthman, MD, American Board of Pathology, accessed at http://web2airmail.net/uthman/; Otis Brawley, National Cancer Institute; National Cancer Institute; Courtesy of Department of Clinical Pathomorphology and Cytology, Medical University, Lodz, Poland; Courtesy of Wikipedia user MBq, via Wikimedia Commons.
page 527 © 2009 Joanne Katz, Courtesy of Photoshare.
Maps 5.1.1 & 5.1.2 GLOBOCAN 2012a
5.2.1 & 5.2.2 GLOBOCAN 2012a
5.3.1 & 5.3.2 GLOBOCAN 2012a
5.4.1 & 5.4.2 GLOBOCAN 2012a
5.5.1 & 5.5.2 GLOBOCAN 2012a
5.6.1 & 5.6.2 GLOBOCAN 2012a
5.7.1 & 5.7.2 GLOBOCAN 2012a
5.8.1 & 5.8.2 GLOBOCAN 2012a
5.9.1 & 5.9.2 GLOBOCAN 2012a
5.10.1 & 5.10.2 GLOBOCAN 2012a
5.11.1–5.11.4 GLOBOCAN 2012a
5.12.1–5.12.3 GLOBOCAN 2012a
5.13.1–5.13.4 GLOBOCAN 2012a
5.14.1 & 5.14.2 GLOBOCAN 2012a
5.15.1 & 5.15.2 GLOBOCAN 2012a
5.16.1 & 5.16.2 GLOBOCAN 2012a
World Cancer Report 617
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
Tables 1.3.1 Compiled from Baade PD, Youlden DR, Valery PC et al. (2010). Population-based survival estimates for childhood cancer in Australia during the period 1997–2006. Br J Cancer, 103:1663–1670. http:// dx.doi.org/10.1038/sj.bjc.6605985 PMID:21063404; Bao PP, Zheng Y, Wu CX et al. (2012). Population-based survival for childhood cancer patients diagnosed during 2002–2005 in Shanghai, China. Pediatr Blood Cancer, 59:657–661. http://dx.doi.org/10.1002/pbc.24043 PMID:22302759; Swaminathan R, Rama R, Shanta V (2008). Childhood cancers in Chennai, India, 1990–2001: incidence and sur- vival. Int J Cancer, 122:2607–2611. http://dx.doi.org/10.1002/ijc.23428 PMID:18324630; and Wiangnon S, Veerakul G, Nuchprayoon I et al. (2011). Childhood cancer incidence and survival 2003–2005, Thailand: study from the Thai Pediatric Oncology Group. Asian Pac J Cancer Prev, 12:2215–2220. PMID:22296359.
2.2.1 Adapted from U.S. Department of Health and Human Services (2004). The Health Consequences of Smoking: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
2.3.1–2.3.3 Jürgen Rehm and Kevin Shield.
2.4.1 From IARC (2012). Biological agents. IARC Monogr Eval Carcinog Risks Hum, 100B:1– 441. PMID:23189750
2.4.2 & 2.4.3 Reprinted from de Martel C, Ferlay J, Franceschi S et al. (2012). Global burden of cancers attributable to infections in 2008: a review and synthetic analysis. Lancet Oncol, 13:607–615. http:// dx.doi.org/10.1016/S1470-2045(12)70137-7 PMID:22575588, © 2012, with permission from Elsevier.
B2.4.1 Reprinted from Schetter AJ, Heegaard NH, Harris CC (2010). Inflammation and cancer: interweaving microRNA, free radical, cy- tokine and p53 pathways. Carcinogenesis, 31:37– 49. http://dx.doi.org/ 10.1093/carcin/bgp272 PMID:19955394
2.5.1 Adapted with permission from Pearce CL, Templeman C, Rossing MA et al.; Ovarian Cancer Association Consortium (2012). Association between endometriosis and risk of histological subtypes of ovarian cancer: a pooled analysis of case-control studies. Lancet Oncol, 13:385–394. http://dx.doi.org/10.1016/S1470-2045(11)70404-1 PMID:22361336, © 2012, with permission from Elsevier.
2.5.2 Adapted from Trabert B, Wentzensen N, Yang HP et al. (2012). Ovarian cancer and menopausal hormone therapy in the NIH-AARP Diet and Health Study. Br J Cancer, 107:1181–1187. http://dx.doi. org/10.1038/bjc.2012.397 PMID:22929888, by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK, © 2012.
2.7.1 & 2.7.2 From IARC. Agents Classified by the IARC Monographs: http://monographs.iarc.fr/ENG/Classification/ClassificationsAlpha Order.pdf.
B2.7.1 Kurt Straif.
B2.7.2 Lesley Rushton.
2.9.1 Kurt Straif.
2.10.1 & 2.10.2 Adapted from Grosse Y, Baan R, Straif K et al.; WHO IARC Monograph Working Group (2009). A review of human carcino- gens – Part A: pharmaceuticals. Lancet Oncol, 10:13–14. http://dx.doi. org/10.1016/S1470-2045(08)70286-9 PMID:19115512, © 2009, with permission from Elsevier.
2.11.1–2.11.4 Ronald T. Riley.
3.1.1 Thomas J. Hudson.
3.6.1 Zdenko Herceg.
3.8.1 Fabio Savarese, Martin Holcmann, and Maria Sibilia.
3.8.2 Adapted from Brown MP, Burdett N (2013). Targeted therapies, aspects of pharmaceutical and oncological management. Cancer Forum, 37:70–80, http://www.cancerforum.org.au/Issues/2013/March/ Forum/Targeted_therapies.htm, with permission from Cancer Forum.
4.1.1 & 4.1.2 Data compiled from IARC (2008). IARC Handbooks of Cancer Prevention, Vol. 12: Tobacco Control: Methods for Evaluating Tobacco Control Policies. Lyon: IARC.
4.5.1 Adapted from Stockwell T (2006). A Review of Research into the Impacts of Alcohol Warning Labels on Attitudes and Behaviour. Centre for Addictions Research of BC. University of Victoria, British Columbia, Canada, with permission from Tim Stockwell.
4.5.2 Adapted from IARC (2010). Some aromatic amines, organic dyes, and related exposures. IARC Monogr Eval Carcinog Risks Hum, 99:1–658. PMID:21528837. Data compiled from the GESTIS database (Institut für Arbeitsschutz der Deutschen Gesetzlichen Unfallversicherung, IFA).
4.5.3 Adapted from Michaels D, Monforton C (2013). Beryllium’s ‘public relations problem’. In: Late lessons from early warnings: sci- ence, precaution, innovation. EEA report 1/2013. Copenhagen: European Environment Agency, with permission from the European Environment Agency.
4.6.1 Modified with permission from Schiller JT, Castellsagué X, Garland SM (2012). A review of clinical trials of human papillomavirus prophylactic vaccines. Vaccine, 30 Suppl 5:F123–F138. http://dx.doi. org/10.1016/j.vaccine.2012.04.108 PMID:23199956, © 2012, with per- mission from Elsevier.
4.7.1 Lawrence von Karsa.
5.1.1 & 5.1.2 Elisabeth Brambilla and William D. Travis.
5.2.1 Adapted from Elston CW, Ellis IO (1991). Pathological prognostic factors in breast cancer. I. The value of histologi- cal grade in breast cancer: experience from a large study with long-term follow-up. Histopathology, 19:403– 410. http://dx.doi. org/10.1111/j.1365-2559.1991.tb00229.x PMID:1757079, by permis- sion from John Wiley and Sons.
5.5.1 Reproduced from Bosman FT, Carneiro F, Hruban RH, Theise ND, eds (2010). WHO Classification of Tumours of the Digestive System, 4th ed. Lyon: IARC, p. 133.
5.5.2 Reproduced from Bosman FT, Carneiro F, Hruban RH, Theise ND, eds (2010). WHO Classification of Tumours of the Digestive System, 4th ed. Lyon: IARC, p. 143.
5.6.1 Neil D. Theise.
5.6.2 Reproduced from Bosman FT, Carneiro F, Hruban RH, Theise ND, eds (2010). WHO Classification of Tumours of the Digestive System, 4th ed. Lyon: IARC, p. 197.
5.7.1 & 5.7.2 Reproduced from Hruban RH, Pitman MB, Klimstra DS (2007). Tumors of the Pancreas (AFIP Atlas of Tumor Pathology Series 4, Fascicle 6). Washington, DC: American Registry of Pathology and Armed Forces Institute of Pathology, with permission from the American Registry of Pathology.
5.7.3 Table adapted from text: Wu J, Matthaei H, Maitra A et al. (2011). Recurrent GNAS mutations define an unexpected pathway for pan- creatic cyst development. Sci Transl Med, 3:92ra66. http://dx.doi. org/10.1126/scitranslmed.3002543 PMID:21775669; and Jiao Y, Shi C, Edil BH et al. (2011). DAXX/ATRX, MEN1, and mTOR pathway genes are frequently altered in pancreatic neuroendocrine tumors. Science, 331:1199–1203. http://dx.doi.org/10.1126/science.1200609 PMID:21252315.
618
a Ferlay J, Soerjomataram I, Ervik M et al. (2013). GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11 [Internet]. Lyon: IARC. Available at http://globocan.iarc.fr.
b Forman D, Bray F, Brewster DH et al., eds (2013). Cancer Incidence in Five Continents, Vol. X [electronic version]. Lyon: IARC. Available at http://ci5.iarc.fr.
5.9.1 Reprinted from Moch H (2013). An overview of renal cell cancer: pathology and genetics. Semin Cancer Biol, 23:3–9. http://dx.doi. org/10.1016/j.semcancer.2012.06.006 PMID:22722066, © 2013, with permission from Elsevier.
5.10.1 Ronald Simon.
5.12.1–5.12.3 Jaime Prat. 5.12.4 Adapted from Prat J (2012). Ovarian carcinomas: five distinct dis- eases with different origins, genetic alterations, and clinicopathologi- cal features. Virchows Arch, 460:237–249. http://dx.doi.org/10.1007/ s00428-012-1203-5 PMID:22322322, with kind permission from Springer Science+Business Media.
5.14.1 Christine Guo Lian and Martin C. Mihm Jr.
5.16.1 & 5.16.2 Paul Kleihues and Jill Barnholtz-Sloan.
6.2.1–6.2.3 Andreas Ullrich and Leanne Riley.
6.7.1 Reproduced from Knaul FM, Gralow JR, Atun R, Bhadelia A, eds (2012). Closing the Cancer Divide: An Equity Imperative. Based on the work of the Global Task Force on Expanded Access to Cancer Care and Control in Developing Countries. Cambridge, MA: Harvard Global Equity Initiative. Distributed by Harvard University Press. © 2012 by President and Fellows of Harvard College acting through the Harvard Global Equity Initiative. Reprinted by permission of the Harvard Global Equity Initiative.
P4.1–P4.3 Mel Greaves.
P7.1 Compiled from Thun MJ, Carter BD, Feskanich D et al. (2013). 50-year trends in smoking-related mortality in the United States. N Engl J Med, 368:351–364. http://dx.doi.org/10.1056/NEJMsa1211127 PMID:23343064
P7.2 Compiled from Peto R, Lopez AD (2001). Future worldwide health effects of current smoking patterns. In: Koop CE, Pearson C, Schwarz MR, eds. Critical Issues in Global Health. New York: Jossey- Bass, pp. 154–161.
Text page 180 Text quotation reprinted from Richter LM, Victora CG, Hallal PC et al. (2012). Cohort profile: the consortium of health-orientated research in transitioning societies. Int J Epidemiol, 41:621–626. http:// dx.doi.org/10.1093/ije/dyq251 PMID:21224276, by permission of Oxford University Press.
page 262 Text quotation reprinted from Greaves M (2005). In ute- ro origins of childhood leukaemia. Early Hum Dev, 81:123–129. PMID:15707724, with permission from Elsevier.
The Editors are grateful to the Charles Rodolphe Brupbacher Stiftung for facilitating inclusion in World Cancer Report 2014 of material based on some contributions to the 2013 Scientific Symposium.
World Cancer Report 619
Subject index A acetaldehyde 101, 102, 175, 377, 425
acoustic neuroma 148
acute lymphoblastic leukaemia (ALL) 70–72, 261, 262, 491, 492
acute myeloid leukaemia (AML) 230, 492
adaptation 338–342
adiposity 129–131
adolescents 70, 72, 74
adult T-cell leukaemia 112
advanced gastric carcinoma 387
advertising campaigns 277, 290, 292–294
advocacy for cancer control 562–566
afatinib 248
aflatoxins 101, 171, 404, 406, 408
aflibercept 451
age-related diseases 78, 79
air pollution 152–157, 307, 312, 358, 543
AKT1 460
alcohol
consumption 96–103, 308, 374, 377, 392, 406, 407, 422, 425, 429
control 301
-related cancers 97–101
warning labels 308
ALDH 377
ALK 355, 356, 490
allergic conditions 513
aluminium production 135
anal cancer 110, 316
analgesics 165, 439
anaplastic astrocytoma 515
anaplastic large cell lymphoma 490
anaplastic thyroid carcinoma 507
androgen insensitivity syndrome 461
androgen receptor 458, 460
androgens 116, 117, 122, 458, 475
angiogenesis 236–241
anogenital cancers 316
antibiotics 171, 254
antineoplastic agents 161
anti-tobacco campaigns 277, 291–294
α-1-antitrypsin deficiency 407
antiviral drugs 408
APC 381, 395–398, 401
areca nut 89, 424
ARID1A 476, 479, 480
Aristolochia 173
aristolochic acid 165, 173
aromatase inhibitors 166, 369, 370
ArrayExpress 212
arsenic 158, 312, 353, 444, 501
aryl hydrocarbon receptor (AhR) 136
asbestos 134, 135, 137, 141, 152, 158, 311
Aspergillus 171
aspirin 166, 255
astrocytoma 514, 515, 518
atrophic gastritis 383
ATRX 515, 518
avoidable cancer burden 576–581, 586–594
azathioprine 162
B bacteria 170
BAP1 442
Barrett oesophagus 379, 380
basal cell carcinoma 500, 501
basal cell naevus syndrome 501
B-cell acute lymphoblastic leukaemia (B-ALL) 230
B-cell non-Hodgkin lymphoma 109
BCL2 488
620
BCR-ABL1 493
behaviour change 268–278, 281–288, 290–295
benefit–risk ratio 161, 166, 325, 326, 525
benzene 135, 153, 156
benzidine 137
benzo[a]pyrene 140
beryllium 312
betel quid 86, 190, 424
bevacizumab 451, 477
biases in screening 324, 523–525
bidis 86, 269, 422, 587
bile ducts 112, 403, 407
biobanks 556–559
biological resource centres 557, 558, 559
biomarkers 172, 184, 187, 190, 206, 207, 210, 218, 219, 224, 225, 354, 358, 428, 442
biomarker validation 225
biomass fuels 543
bisphenol A (BPA) 154
birth cohorts 177–180
Birt–Hogg–Dubé syndrome 441
bladder adenocarcinoma 450
bladder cancer 113, 134, 135, 137, 158, 444–451
prevention 451
screening 451
bladder small cell carcinoma 450
bladder squamous cell carcinoma 450
bladder urothelial carcinoma 448, 450
bone metastases 241, 242, 243
BRAF 377, 396, 399, 462, 478, 485, 488, 499, 500, 507, 508, 515
brain cancer 511–519
brain metastases 241
BRCA1 369, 371, 478, 479
BRCA2 369, 371, 478, 479
breast cancer 55, 57, 58, 61, 64, 97, 100, 102, 115, 116, 118, 120, 124, 126, 127, 130–132, 163–166, 184, 188, 196, 207, 211, 241, 242, 256, 324, 331– 333, 339, 362–372
prevention 369, 370, 568
screening 323–326, 332, 333, 371, 372, 543, 544
breastfeeding 115
breast implants 164
bronchial airway epithelial cells 354
Burkitt lymphoma 70, 111, 490
busulfan 161
1,3-butadiene 153
C CA19-9 411
calcitonin 503, 506, 507
calcium 127, 458, 460
caloric intake 281, 283, 284, 285
caloric restriction 283
calreticulin 259
cancer control 337, 339, 342, 528–585
in Africa 528
in Canada 537
in China 546
in France 553
in India 561
in Jordan 568
in Morocco 575
in Peru 584
in Turkey 585
Cancer Genome Atlas 211
cancer registries 16, 49, 50, 52, 73, 74, 531, 537, 540, 544, 546, 555–557, 561, 563, 568, 582, 585
cancer registry linkages 556, 557
cancer screening 522–525
cancer stem cells 230, 231, 233
cancer vaccines 254, 258
carcinoembryonic antigen (CEA) 507
carcinogenic drugs 161, 162, 165, 166
carcinogens 162, 163
in alcoholic beverages 101
in indoor air pollution 156, 158
in meat 128, 129
in natural products 169–174
in outdoor air pollution 153
in polluted drinking-water 158
in smokeless tobacco 89, 90
in tobacco smoke 88, 90, 107
Carney complex 508
β-carotene 353
World Cancer Report 621
β-catenin 420, 476, 479
CCND1 448, 487, 489
CCNE1 479
CDH1 381, 389, 390, 409
CDK4 499
CDKN2A 357, 416, 417, 427, 448, 495, 499
cell surface markers 231
cervical adenocarcinoma 473
cervical cancer 57, 59, 61, 64, 110, 120, 164, 165, 316, 317, 320, 323, 331, 333, 470–474
screening 323, 324, 326, 332, 333, 474, 541–545, 564
vaccines 317–320, 474
cervical intraepithelial neoplasia 471–473
cervical squamous cell carcinoma 473
cetuximab 247, 248
chemopreventive agents 165, 166, 167
chemotherapeutic agents 165, 171, 259, 371, 372
chewing tobacco 82, 422
childhood malignancies 69–74, 261, 262
chlorambucil 161
chlorofluorocarbons 156
chloroform 158
cholangiocarcinoma 112, 403, 406–409, 411
cholesterol 458
chromophobe renal cell carcinoma 440, 441
chromosomal instability pathway 395
chromothripsis 185
chronic bladder infection 447
chronic gastritis 383
chronic liver infections 315, 408
chronic lymphocytic leukaemia (CLL) 208, 482, 483
chronic myeloid leukaemia (CML) 184, 185, 493
CIC 516, 518
ciclosporin 162, 171
cigarettes 82–86, 90, 268–278, 350, 354, 422, 426, 429, 436, 586–594
cirrhosis 98, 315, 406–408, 410
cisplatin 371
clear cell renal cell carcinoma 440–442
clonal evolution 339–342
coal combustion 543
coal tar 134, 140
cohort studies 177–180, 555, 558, 559
colectomy 396, 397
colitis 397, 401
collecting duct renal cell carcinoma 440
colon cancer 392–401
colonoscopy 323, 326, 585
colorectal cancer 55, 57, 58, 59, 66, 97, 102, 122, 126, 127, 128, 163, 166, 230, 237, 255, 294, 323, 331, 392–401
screening 323, 324, 326, 332, 334, 400
colposcopy 473, 474
combined estrogen–progestogen contraceptives 164, 475
combined estrogen–progestogen menopausal therapy 163, 475
combustion emissions 152, 154, 155, 156, 157
competing endogenous RNA (ceRNA) 216
computed tomography (CT) 144, 145, 324, 360, 525
cost-effectiveness of interventions 533
costs of cancer 529, 533, 565, 576–582
Cowden syndrome 506, 508
COX-1 255
COX-2 241, 255, 397
CpG island methylator pathway 396
creosotes 140
crizotinib 355
Crohn disease 397
cryotherapy 323, 541
cryptorchidism 461
CTNNB1 390, 409, 420, 476, 479, 517
cyclophosphamide 161
cystic kidney disease 439
cytokines 254, 257
D dairy products 126, 458
data linkage 556, 557
DBC1 448
DDR2 356, 357
developmental origins of health and disease 177
diabetes 281, 282, 407, 416, 439, 475
diesel engine emissions 135, 137, 153
622
diet 124–132, 281–288, 294, 300, 309, 375, 386, 388, 432–434, 458, 460, 504, 513
dietary exposures 124–129, 131, 226
dietary fibre 127
diethylstilbestrol 164
diffuse astrocytoma 515
diffuse large B-cell lymphoma 207, 490
digital rectal examination 543
dioxins 136
disability-adjusted life years (DALYs) 99, 100, 577, 579
DNA
damage 145, 147, 165, 198, 205
methylation 214, 216, 218, 219
repair 186, 198, 205, 214, 216, 217, 219, 399
DOCK2 381
dose–response 97
drug resistance 188, 206, 237, 240, 248, 250, 251, 339, 340, 342, 371
ductal carcinoma in situ (DCIS) 365
dysplastic naevi 495, 499
dysplastic naevus syndrome 495
dysplastic nodules 407, 410
E E2F3 448
early detection 322–328, 330–334, 360, 542, 543, 546, 548, 577, 578, 581
early gastric carcinoma 387
early-life exposures 177–180
E-cadherin 389
economic burden of cancer 576–582
economics of chronic and noncommunicable diseases 302
EGFR 355, 356, 427, 474, 515
electromagnetic fields 143, 147, 148, 512
electronic cigarettes 276
ELMO1 381
embryonic stem cells 228
Encyclopedia of DNA Elements 211
endocrine disrupters 120
endogenous hormones 116–120
endometrial cancer 118, 119, 131, 163–165, 474–477
endometriosis 119, 479
endoscopy 378, 381, 388, 390, 429, 546
environmental pollution 151–159, 307
ependymoma 516
epidemiological transition 55–66
epidermal growth factor receptor (EGFR) 237, 241, 244–248, 395, 396, 409
epidermolysis bullosa syndrome 501
epigenetic plasticity 339, 341, 342
epigenetics 214–219
epigenetic therapy 219
epigenomics 216–218
epithelial–mesenchymal transition (EMT) 240
Epstein–Barr virus (EBV) 105, 111, 112, 426, 429, 490
ERBB receptors 244–247
erlotinib 248, 355, 451
estrogen
metabolism 118, 119
receptor 365, 366
therapy 117, 118, 120, 163
estrogen-only menopausal therapy 163
estrogen–progestogen therapy 118, 120, 163
estrogens 116–120, 122, 474, 475
ethanol 101
European Prospective Investigation into Cancer and Nutrition (EPIC) 287, 288, 555
evaluation of prevention campaigns 269, 271, 293, 294
everolimus 420
evolution 337–342
exome 188
exposome 226
EZH2 490, 493
F faecal immunochemical testing 326
faecal occult blood testing (FOBT) 324, 326, 332, 400, 543, 585
familial adenomatous polyposis syndrome 398, 506, 508
familial colorectal cancer syndromes 397–399
familial non-medullary thyroid carcinoma 508
World Cancer Report 623
fermentation 131, 426
FGFR1 356, 357, 359
FGFR3 448
finasteride 122
fine particles 153–156
5-fluorouracil 396
folate 127
folic acid 127, 434
follicular epithelial cells 503
follicular lymphoma 488
follicular thyroid cancer 504, 507
food 124–132, 171–173, 300, 301, 386, 388, 432–434
food contaminants 171, 172
formaldehyde 156
fruit 126, 433, 434
FUBP1 516, 518
fumonisins 172
fundic gland polyposis 389
fungi 171, 172
Fusarium 172
fusion genes 210
G Gardner syndrome 398, 508
gastrectomy 390
gastric adenocarcinoma 106
gastric adenocarcinoma and proximal polyposis of the stomach 389
gastric cancer 383–390
prevention 388, 390
screening 388, 390
gastric lymphoma 108
gastro-oesophageal reflux disease 378, 379, 380
gefitinib 248, 355
gene–environment interactions 196–198
gene expression 197, 199, 203–212, 214–219
Gene Expression Omnibus 212
gene expression profiling 203–207, 210–212
gene fusions 206
genetic susceptibility 339, 340
to breast cancer 362, 369
to cervical cancer 474
to colorectal cancer 397, 398, 399
to haematological malignancies 483, 485, 487, 489, 490
to head and neck cancers 429
to kidney cancer 440, 441
to lung cancer 353
to melanoma 200
to nervous system tumours 514
to oesophageal cancer 378
to ovarian cancer 478
to pancreatic cancer 416
to prostate cancer 459
to skin cancer 499, 501
to stomach cancer 388
to testicular cancer 461
to thyroid cancer 506, 508
to tobacco-related cancers 91
genital warts 318, 319, 473
genome-wide association studies (GWAS) 91, 193–200
genomics 184–191, 193, 195, 200, 232
genotoxicity 154, 161–163
germ cell tumours 461–463
germline mutations 369, 371
gingivo-buccal cancer 190
glioblastoma 246, 515–518
glioma 148, 512–518
global burden of cancer 16–24, 54–67
Global Burden of Disease study 100
GLOBOCAN 16, 17, 20, 57, 348, 349
glycine 224
glycine decarboxylase 224
glycolysis 222, 223, 224
GNAS 419
gonadal dysgenesis with a Y chromosome 461, 462
Gorlin syndrome 501
granulocyte-macrophage colony-stimulating factor 238, 255
griseofulvin 171
GSTM1 447
624
H habitual behaviours 291, 294
haematopoietic and lymphoid malignancies 482–493
haematopoietic stem cells 228, 230, 493
hairy cell leukaemia 485
hamartomatous polyp syndromes 398
head and neck cancers 96, 97, 110, 246, 422–430
health
advocacy 562–566
economics 576–582
information privacy 572, 573
law 569, 570
messages 291–295
systems strengthening 547–551
Helicobacter pylori 106, 253, 380, 384, 386, 387, 388, 489
hepatitis B virus 108, 109, 314–316, 403, 404, 406–408
vaccine 315, 316, 564
hepatitis C virus 108, 109, 199, 403, 404, 406, 408
hepatocellular carcinoma 108, 314, 315, 403–411
prevention 408, 411
screening 410
HER2 355, 448
herbal products 173
hereditary
diffuse gastric cancer 389
haemochromatosis 407
leiomyomatosis and renal cell carcinoma 441
non-polyposis colon cancer 398, 476, 478
papillary renal cell carcinoma 440
heterocyclic amines (HCAs) 128
highly active antiretroviral therapy (HAART) 112
high-voltage power lines 148
histone modifications 215, 216, 219
Hodgkin lymphoma 111, 491
homology-directed repair (HDR) 205
hormonal
agents 161, 162, 163, 165
factors 115–122, 439, 458, 474, 478
therapy 115–122, 162–164, 475
hormone antagonists 166
hot beverages 374
Human Development Index (HDI) 56–67, 69
human epidermal growth factor receptor 2 (HER2) 365, 366, 389
Human Genome Project 184
human immunodeficiency virus (HIV) 112
human mammary tumour virus (HMTV) 111
Human Microbiome Project 108
human papillomavirus (HPV) 314–320, 426–428, 467, 470, 472–474, 501
infection 110, 111, 114, 120, 253, 466, 467, 470–473
testing 332, 333, 474
vaccination 254, 317, 318, 320, 467, 564
vaccines 318, 319, 474
human T-cell lymphotropic virus type 1 (HTLV-1) 112
hypertension 439, 442
hypoxia 236, 238–240
I IARC Handbooks of Cancer Prevention 269
IARC Monographs 135–139, 140, 146, 148, 161, 163–165, 170, 173, 174, 310–312
IDH1 493, 515, 516, 518
IDH2 490, 493, 515, 516
IGH 488, 490
IKZF1 491, 492
IL1B 387
IL1RN 387
imiquimod 257
immune escape 255–257
immunosuppression 255, 256, 258, 467
immunosuppressive agents 162
immunosurveillance 256
immunotherapy 257–259, 500
implementation of screening 330–334
indoleamine 2,3-dioxygenase 257, 258
infections 105–114, 253, 261–263, 514
infiltrating ductal adenocarcinoma 413, 416–418
inflammation 107, 239, 253–255, 257, 259, 358, 395, 397, 401, 458
inflammatory bowel disease 397, 401
insecticides 157
World Cancer Report 625
insulin 167, 475
interferon 257, 262, 263, 408
interleukins 256, 257, 401
International Atomic Energy Agency 548, 550, 551
International Cancer Genome Consortium 185, 188, 190, 211
intestinal cancer 397
intraductal papillary mucinous neoplasms 418, 419
intratubular germ cell neoplasia 461, 462
involuntary smoking 92, 93, 137, 140, 141, 158, 268, 276, 277, 299
iodine 504
ionizing radiation 144–146, 309, 501, 506, 508, 512
ipilimumab 259, 500
J JAK2 493
K Kaposi sarcoma 58, 70, 112
KIAA1549 515
kidney cancer 130, 436–442
KIT 462
KRAS 247, 355, 389, 395, 396, 400, 410, 416, 419, 462, 474, 476, 478, 479
L lapatinib 248, 451
large cell lung carcinoma 359
laryngeal cancer 97, 422, 425, 427, 429
lasiocarpine 173
law in cancer control 569–573
legislative initiatives 305–312
leukaemia 112, 140, 145, 148, 482, 483, 485, 491–493
Lgr5 397
lichen sclerosus 470
lifestyle intervention 271, 278, 281, 282, 285, 291, 294
linkage disequilibrium 193, 194, 197, 199
liver cancer 57, 59, 97, 100, 108, 122, 164, 172, 314–316, 403–411
prevention 408, 411
screening 410
vaccine 315, 316
liver flukes 112, 406, 407
lobular carcinoma in situ (LCIS) 365
lobular involution 116
lung adenocarcinoma 353, 355, 356
lung cancer 57, 58, 59, 66, 91–94, 134, 135, 137, 142, 145, 146, 152, 155, 156, 158, 350–360, 586–592
prevention 359
screening 360
lycopene 458, 460
lymphoblastic lymphoma 492
lymphoma 105, 108, 482, 488–492
lymphoplasmacytic lymphoma 489
Lynch syndrome 398
M macrophages 239, 240
magnetic resonance imaging 372
major histocompatibility complex (MHC) 255, 257
MALT lymphoma 489
mammary gland development 256
mammography 164, 324, 325, 332, 371, 372
mantle cell lymphoma 488, 489
marginal zone lymphoma 489
mass media campaigns 277, 290–294
mass spectrometry (MS) 223, 224, 225, 226
mastectomy 370, 371
matrix metalloproteinases 239, 428
mature B-cell lymphoma 488
meat 126, 128, 129, 375, 387, 388, 392, 393, 433, 437, 458, 514
mediastinal large B-cell lymphoma 490
medical radiation 144, 145, 506, 511
medullary thyroid carcinoma 507–509
medulloblastoma 517
melanoma 147, 207, 250, 495–501
melphalan 161
MEN1 420
626
meningioma 512, 513, 517
menopausal hormone therapy 116, 118, 119, 122, 162, 163
menstrual factors 115, 116, 119, 120
mesothelioma 140
messenger RNA (mRNA) 204, 207, 216
MET 355
metabolic change 222–226
metabolic syndrome 223, 287, 288, 407
metabolism 222–226
metabolome-wide association studies (MWAS) 226
metabolomics 222–226
metastasis 240–243, 339, 340
metformin 166, 370
methoxsalen 165, 174
methylation 185, 189, 216, 493
MGMT 516
microarrays 194, 195, 204, 206–208, 212
microbiome 131
microcystin-LR 171
microRNA (miRNA) 203, 204, 207–209, 216, 220, 448
microsatellite instability pathway 395
minimal residual disease 491
mining 134, 141, 146, 152, 159
minor allele frequency (MAF) 193, 196, 197
mismatch repair deficiency syndrome 398
mitomycin C 418
MLH1 395, 396, 399, 476
mobile phones 148, 513
monoclonal antibodies 237, 247, 258, 259
monoclonal B-cell lymphocytosis (MBL) 482
monocrotaline 173
monophosphoryl lipid A 257
motor vehicle emissions 152
MPOWER measures 94, 585
MUC1 388
mucinous cystic neoplasms 419
multiple endocrine neoplasia type 2 (MEN2) 508, 509
multiple myeloma 487
mutation hotspots 187
MUTYH-associated polyposis (MAP) 399
MYC 427, 460, 490, 499
mycotoxins 171
MYD88 490
N naevi 495, 498, 499
naevoid basal cell carcinoma syndrome 501
2-naphthylamine 135
nasopharyngeal cancer 100, 111, 426, 428
NAT2 447
national cancer control plans 529–535, 538–545, 547–551, 563
national cancer control programmes 528, 537, 546, 553, 561, 575, 584, 585
naturally occurring chemical carcinogens 169–175
natural products 169–174
natural selection 337–342
nervous system tumours 70, 241, 262, 511–519
neuroblastoma 70, 219, 262, 518
neurofibroma 517
neurofibromatosis 515–517
NF1 479
NF2 516
nicotine 88, 89, 269, 274, 276, 278
NKX3.1 460
N-nitroso compounds (NOCs) 128, 131, 513
non-alcoholic fatty liver disease 404, 407, 411
non-cardia gastric adenocarcinoma 106
non-coding RNA 203, 207, 210, 211, 216, 217
noncommunicable diseases 55, 67, 177–180, 298–303, 548, 550, 562–566, 579
non-Hodgkin lymphoma (NHL) 109, 112, 162, 490
non-melanoma skin cancer 500, 501
non-small cell lung carcinoma 237, 246, 247, 353, 355, 356, 359
non-steroidal anti-inflammatory drugs (NSAIDs) 166, 255
NOTCH1 483, 489, 492
nuclear magnetic resonance (NMR) spectroscopy 223, 224, 225
nucleos(t)ide analogues 408
nutrition 124–129, 294, 300, 301, 432–434, 458
nutrition transition 125
World Cancer Report 627
O obesity 116, 118, 122, 130–132, 222, 223, 281–284,
287, 309, 367, 378, 395, 407, 433–436, 439, 474
occupational carcinogens 134–142, 310
occupational exposures 134–142, 145, 148, 149, 310, 513
ochratoxin A 172
oesophageal adenocarcinoma 378–381
oesophageal cancer 57, 97, 100, 130, 374–381, 546
screening 378, 381
oesophageal squamous cell carcinoma 374–378
oligodendroglioma 516, 518
oncogenes 184, 197, 214, 216, 247, 254, 340, 353, 357, 359, 389, 390, 416, 426, 427, 428, 441, 448, 460, 476, 493, 499, 508, 517
oophorectomy 371
opioid analgesics 565, 571, 572
opium 375
optic glioma 515
oral cancer 97, 100, 190, 422–430
screening 324, 428
oral contraceptives 119, 122, 164, 166, 473, 475, 478
oropharyngeal cancer 100, 110, 111, 317, 320, 422–430
ovarian cancer 119, 120, 163, 164, 166, 371, 477–480
overdiagnosis 324, 325, 326, 371, 523, 524
overtreatment 325, 461, 524
overweight 130, 132, 281, 282, 309, 433, 434, 436
P p16 416, 417, 427, 448, 450, 476, 499
pancreatic cancer 97, 99, 102, 130, 413–420
pancreatic neuroendocrine tumours 420
pancreatitis 416
papillary renal cell carcinoma 440, 441
papillary thyroid carcinoma 506–508
papillary urothelial neoplasms of low malignant potential 447, 448
Pap smear 324, 473, 474, 542
α-particles 144, 146
β-particles 144, 146
particulate matter 151–156, 353
PBRM1 442
Penicillium 171
penile cancer 110, 316
pentachlorodibenzofuran 136
peripheral T-cell lymphoma 490
pesticides 157
pharmaceutical drugs 161–167, 170, 171, 174, 439
pharyngeal cancer 97, 422, 425–427
phenacetin 165, 439
Philadelphia chromosome-positive leukaemia 491
Philadelphia translocation 184, 185
phosphoglycerate dehydrogenase 223
photosensitivity 165
phylogenetic tree 339, 342
physical activity 132, 281, 282–285, 288, 294, 300, 301, 367, 368, 433, 434
phytochemicals 126, 131
pickled vegetables 375, 386
PIK3CA 356, 357, 427, 428, 474, 476, 480
pilocytic astrocytoma 514
PIVKA 410
plain cigarette packaging 275, 564, 570, 571
plants 173
plasma cell myeloma 487
platelet-derived growth factor receptor β (PDGFR-β) 243
Plummer–Vinson syndrome 377
poly(ADP-ribose) polymerase (PARP) inhibitors 371, 418, 477
polybrominated biphenyls (PBBs) 137
polychlorinated biphenyls (PCBs) 135, 136, 158
polycyclic aromatic hydrocarbons (PAHs) 88, 90, 128, 135, 140, 153, 155, 156
polycystic ovary syndrome 475
polymerase chain reaction (PCR) 204, 218
polypectomy 396, 400
pooled birth cohorts 179, 180
poorly differentiated thyroid carcinoma 507
population attributable fraction (PAF) 105, 109, 111, 112, 113
population-based screening programmes 322–328, 330–334
population-wide prevention campaigns 290–295
628
population-wide prevention strategies 298–301, 531, 533
PPP2R1A 476
prenatal infections 261–263
prevention
campaigns 276, 278, 290–295
strategies 268–278, 298–303, 305–312, 314–320, 322–328, 330–333, 528, 530, 532, 533, 541, 548, 563–565
primary glioblastoma 515
primary prevention campaigns 290, 293
primary sclerosing cholangitis 406, 407, 410
progesterone receptor 365, 366
progestogens 163, 164
prognostic markers 187, 190, 219
Programme of Action for Cancer Therapy (PACT) 550, 551
prospective cohorts 178, 179
prostate cancer 55, 57, 58, 59, 61, 64, 65, 97, 98, 122, 126, 127, 166, 167, 243, 258, 339, 453–461
screening 326, 460
prostate-specific antigen (PSA) 459, 460, 523
prostatic adenocarcinoma 459
prostatic intraepithelial neoplasia 459
PSCA 388
psychological stress 79
PTCH 501, 517
PTEN 216, 427, 442, 460, 476, 477, 479, 480, 499, 515
Q quality assurance in screening 325, 327, 330, 331
R γ-radiation 144, 145
radiation 143–149, 501, 511
radiofrequency electromagnetic fields 148
radionuclides 144, 146
radiotherapy 144, 512, 549, 550, 551
radon 144, 146
raloxifene 166, 369
RB1 359
receptor tyrosine kinases 244, 245, 389
rectal cancer 392–401
5α-reductase inhibitors 167, 458
regional burden of cancer 25, 26, 61, 64, 65
regulatory initiatives 305–312, 569–571
renal cancer 436–442
renal pelvis cancer 436, 439
reproductive factors 115–122
research infrastructure 554–559
RET 506, 508, 509
retinoblastoma 70
retrospective cohorts 179
rhabdoid tumours 186
ribavirin 408
riddelliine 173
right to health 573
risk prediction models 116, 196
RNA sequencing (RNA-seq) 204, 206, 212
RNA splicing 203
RNF43 419
ROS1 355
S salt-preserved food 426
sarcosine 225
Schistosoma haematobium 113, 447
schistosomiasis-associated adenocarcinoma 450
schwannoma 517
screening 291, 294, 322–328, 330–334, 522–525, 528, 542, 546, 585
secondary glioblastoma 515
secondary prevention campaigns 290
selective estrogen-receptor modulators 166, 369
selective pressures 338, 340
selenium 458, 460
semustine 161
serous cystadenomas 419
SETD2 442
SF3B1 483
sigmoidoscopy 326
signal transduction 244–251
silica 137
World Cancer Report 629
single-nucleotide polymorphisms (SNPs) 128, 193, 194, 195, 196, 197, 198, 199
sipuleucel-T 258
skin cancer 147, 156, 158, 162, 246, 495–501
prevention 500, 501
SMAD4 395, 416, 419
small cell lung carcinoma 353, 357, 359
small lymphocytic lymphoma (SLL) 482
small-molecule inhibitors 237, 247, 248
smoke-free legislation 276, 277, 306–308, 568, 575
smokeless tobacco 82, 85, 86, 88–91, 269, 273, 274, 422
smoking 82–86, 88–94, 268–278, 306, 307, 350, 354, 374, 377, 387, 414, 422, 426, 436, 444, 467, 473, 568, 586–594
smoking-attributed mortality 586–594
smoking cessation 268, 269, 273, 276, 278, 359, 586–594
snuff 89
soil pollution 159
solar radiation 137, 146, 147, 309, 495, 498, 500
solid fuel combustion 543
solid-pseudopapillary neoplasms 420
somatic mutations 186, 188, 200, 368
somatic stem cells 228, 229
soot 135, 140
sorafenib 247, 451
SOX2 357
spinal ependymoma 516
squamous cell carcinoma 165, 246, 356, 357, 422, 427, 428, 465, 466, 500, 501
squamous intraepithelial lesions 467, 470–474
statins 166
stem cells 223, 228–233, 397, 493
sterigmatocystin 172
stomach cancer 57, 58, 59, 106, 158, 383–390
prevention 388, 390
screening 388, 390
Streptomyces 170
sugar-sweetened beverages 129, 311, 433, 434
sunburn 147, 294, 498, 501
sunitinib 247, 420, 451
sunlight 146, 147, 309, 495, 498–501
sun protection 147, 309, 498–501
sun protection campaigns 147, 294, 309
sunscreen 498, 500
survivorship 73, 74
susceptibility alleles 193, 194, 197, 199, 200
SV40 514
synthetic lethality 371
T tamoxifen 165, 166, 369, 475
tanning devices 147, 309, 495, 501
targeted therapy 184, 190, 206, 211, 220, 223, 237, 241, 242, 243, 244–251, 259, 355, 356, 357, 372, 390, 420, 451, 477, 480, 500, 553
taxation 271–273
telomerase 77, 78
telomeres 77–79
telomere syndromes 78
temozolomide 516
TERT 515
testicular cancer 120, 165, 461, 462
testosterone 458
TET2 490, 493
2,3,7,8-tetrachlorodibenzo-para-dioxin (TCDD) 136
thiotepa 161
thyroid cancer 146, 503–509
thyroidectomy 507
tissue-specific stem cells 228, 229
TMPRSS2 460
tobacco
advertising 83, 86, 273, 275, 299
-attributed mortality 586–594
control 83, 84, 86, 94, 268–278, 299, 300, 306–308, 568, 570, 571
economic cost of – 302
-induced injury 354
packaging 274–276, 564, 570, 571
prices 271–273
-related cancers 90, 91, 586–592
smoking 82–85, 88, 90, 186, 190, 268–278, 306, 307, 350, 354, 414, 422, 426, 436, 444, 586–594
-specific nitrosamines 88, 89, 90
taxation 271–273
use 82–86, 90, 268–278, 374, 377, 422
630
TP53 172, 173, 187, 357, 359, 416, 419, 427, 448, 450, 460, 462, 470, 476, 478, 479, 483, 499, 501, 508, 515
transcription 203, 204
transforming growth factor β (TGF-β) 242
transitions in human development 54–67
translation 204, 207
trastuzumab 390, 451
treosulfan 161
trichloroethylene 439, 441
triple-negative breast cancer 366
TSC1 448
tumour antigens 255–259
tumour microenvironment 236–243, 253–256, 258, 260
tumour necrosis factor inhibitors 167
tumour suppressor genes 184, 186, 187, 214, 216, 340, 390, 416, 419, 427, 441, 442, 448, 450, 460, 476, 492, 499, 508, 515
tumour vasculature 236, 237, 238, 241, 243
U ulcerative colitis 397, 401
ultrasonography 372
ultrasound elastography 509
ultraviolet radiation 146, 147, 186, 190, 309, 495, 498, 500, 501
unopposed estrogen therapy 117–119
uterine cancer 474–477
V vaccination 291, 295, 314–320, 546
vaccines 254, 258, 314–320
vaginal cancer 110, 165, 316, 320, 466, 467
vaginal intraepithelial neoplasia 466
value of a statistical life (VSL) 577, 579
vascular endothelial growth factor (VEGF) 236–238, 256, 258, 477
vegetables 126, 375, 426, 433, 434, 458
vemurafenib 250, 500
VHL 419, 440, 441, 442
visual inspection of the cervix with acetic acid (VIA) 324, 333, 541, 542, 544, 564
vitamin D 127
vitamin E 458
von Hippel–Lindau syndrome 440
vulvar cancer 110, 316, 320, 465, 466, 467, 470
vulvar intraepithelial neoplasia 466, 467, 470
vulvar squamous cell carcinoma 466, 467, 470
W Warburg effect 222
water pollution 158, 159, 307, 312
weight loss 281–288
weight management 281–286
Werner syndrome 508
WHO Framework Convention on Tobacco Control 86, 268, 271, 273, 274, 277, 278, 299, 300, 306, 307, 564, 569, 570
WHO Global Burden of Disease study 100
WHO Global Strategy for the Prevention and Control of Noncommunicable Diseases 298, 299
WHO Global Strategy on Diet, Physical Activity, and Health 300
WHO Global Strategy to Reduce the Harmful Use of Alcohol 303, 308
whole-genome sequencing 193, 250, 251
WHO Noncommunicable Disease Country Capacity Survey 538–545
Wilms tumour 436
X xenobiotic metabolizing enzymes (XMEs) 128
xeroderma pigmentosum 501
X-radiation 144, 145
Y young adults 70
- Cover
- Contents
- Foreword
- Preface
- Introduction
- Perspectives
- Reconciling stress and cancer: insights from telomeres
- Early-life exposures, birth cohorts, and noncommunicable diseases (with special reference to cancer)
- Prenatal infections with subsequent immune tolerance could explain the epidemiology of common childhood cancers
- An evolutionary foundation for cancer control
- Diet, nutrition, and cancer: where next for public health?
- Controversies in cancer screening and their resolution: a view from the United States “battleground”
- The full hazards of smoking and the benefits of stopping: cancer mortality and overall mortality
- 1 Cancer worldwide
- 1.1 The global and regional burden of cancer
- Summary
- Global burden of cancer
- Regional patterns of cancer
- Country-specific patterns of cancer
- 1.2 Transitions in human development and the global cancer burden
- Summary
- Epidemiological transitions, noncommunicable diseases, and cancer
- The Human Development Index: a marker of growth and development
- Global cancer statistics by human development level in 2012
- Evidence of cancer transitions: some temporal examples
- Regional cancer profiles and human development
- Cancer incidence by 2025: demographic and trend-based predictions
- Overview
- 1.3 Childhood cancer
- Summary
- Incidence rates
- Outcomes
- Data sources
- 2 Cancer etiology
- 2.1 The global tobacco epidemic
- 2.2 Tobacco smoking and smokeless tobacco use
- Genetic susceptibility to tobacco-related cancers
- 2.3 Alcohol consumption
- 2.4 Infections
- Inflammation and cancer
- Do infections play a role in breast cancer?
- 2.5 Reproductive and hormonal factors
- 2.6 Diet, obesity, and physical activity
- Challenges of measuring diet in cancer epidemiological studies and new perspectives
- Single-nucleotide polymorphisms relevant to meat consumption and cancer risk
- Diet and the gut microbiome
- 2.7 Occupation
- Evaluating carcinogens: dioxins and dioxin-like substances
- Preventing occupational cancer: successes and failures
- 2.8 Radiation: ionizing, ultraviolet, and electromagnetic
- An IARC announcement that made waves
- 2.9 Pollution of air, water, and soil
- Bisphenol A and cancer
- Pesticides and cancer
- 2.10 Pharmaceutical drugs
- Breast implants and cancer
- 2.11 Naturally occurring chemical carcinogens
- 3 Cancer biology
- 3.1 Genomics
- TP53 mutations and human cancer
- 3.2 Genome-wide association studies
- Gene–environment interactions and breast cancer
- DNA repair polymorphisms and human cancer
- 3.3 Gene expression
- Therapeutic implications of dysregulated DNA damage signalling
- 3.4 Epigenetics
- The non-coding RNA revolution in medical research
- Causes and consequences of microRNA dysregulation in cancer
- 3.5 Metabolic change and metabolomics
- 3.6 Stem cells and cancer stem cells
- Stem cells in cancer: determinants of clinical outcome?
- 3.7 Tumour microenvironment
- 3.8 Signal transduction and targeted therapy
- 3.9 Immunology and immunotherapy
- Role of the innate and acquired immune systems in mammary development and breast cancer
- Premortem autophagy and endoplasmic reticulum stress as immunogenic signals in cancer therapy
- 4 Cancer prevention
- 4.1 Changing behaviours – tobacco control
- Australia’s plain packaging of tobacco products Tobacco and China
- 4.2 Changing behaviours – physical activity and weight control
- Energy restriction, age, and cancer risk
- EPIC as a model study vehicle
- 4.3 Designing and evaluating population-wide campaigns
- 4.4 Prevention strategies common to noncommunicable diseases
- The global economics of chronic and noncommunicable diseases
- 4.5 Legislative and regulatory initiatives
- Environmental pollution: old and new
- Taxing sugar-sweetened beverages: the Brazilian case
- 4.6 Vaccination
- Early implementation and monitoring of HPV vaccination in Bhutan
- Introduction and monitoring of a national HPV vaccination programme in Rwanda
- 4.7 Screening – principles
- 4.8 Screening – implementation
- 5 Cancer by organ site
- A guide to the epidemiology data in World Cancer Report
- 5.1 Lung cancer
- Earliest molecular evidence of tobacco-induced injury in the airway
- Biomarkers of air pollution
- 5.2 Breast cancer
- Biological mechanisms mediating reduced breast cancer risk through physical activity
- Treatment determined by BRCA1/2 mutation
- 5.3 Oesophageal cancer
- Risk of malignant progression in Barrett oesophagus patients
- 5.4 Stomach cancer
- Gastric cancer prevention
- 5.5 Colorectal cancer
- Lgr5 stem cells in self-renewal and intestinal cancer
- Tumour-elicited inflammation and malignant progression in colorectal cancer
- 5.6 Liver cancer
- Treating chronic hepatitis with antiviral drugs to prevent liver cancer
- 5.7 Pancreatic cancer
- 5.8 Head and neck cancers
- 5.9 Kidney cancer
- 5.10 Bladder cancer
- 5.11 Cancers of the male reproductive organs
- 5.12 Cancers of the female reproductive organs
- 5.13 Haematopoietic and lymphoid malignancies
- Molecular detection of minimal residual disease in childhood leukaemia
- De novo DNA methyltransferases in normal and malignant haematopoiesis
- 5.14 Skin cancer
- Evaluating sunscreen as a measure to reduce the risk of melanoma
- 5.15 Thyroid cancer
- 5.16 Tumours of the nervous system
- Glioma genomics and its implications in neuro-oncology
- 6 Cancer control
- 6.1 National cancer control plans
- 6.2 Current global national cancer control capacity
- Cervical cancer prevention in six African countries
- Traditional household cooking: risks and prevention
- 6.3 Health systems strengthening for cancer control
- Towards understanding the economics of cancer: priority areas for research
- 6.4 Research infrastructure – biobanks, cohorts, registries, and data linkage
- 6.5 Advocacy for cancer control
- 6.6 Law in cancer control
- 6.7 The global economic burden of cancer
- Cancer control experiences
- Cancer control in Africa: options for a vulnerable continent
- Cancer control in Canada: challenges and strategies in a high-income country
- Cancer control in China: preventive policies and accessible health care
- Cancer control in France: towards patient-centred precision medicine
- Cancer control in India: cancer care through a four-tier system
- Cancer control in Jordan: goals for low- and middle-income countries
- Cancer control in Morocco: action in harmony with the socioeconomic and cultural context
- Cancer control in Peru: “El Plan Esperanza” – the Hope Plan
- Cancer control in Turkey: an encouraging national cancer control plan for the future
- Contributors
- Disclosures of interests
- Sources
- Subject index
- A
- B
- C
- D
- E
- F
- G
- H
- I
- J
- K
- L
- M
- N
- O
- P
- Q
- R
- S
- T
- U
- V
- W
- X
- Y
