Screening for Disease

www.thelancet.com/oncology Vol 11 August 2010 725

Articles

Lancet Oncol 2010; 11: 725–32

Published Online July 1, 2010 DOI:10.1016/S1470- 2045(10)70146-7

See Refl ection and Reaction page 702

Department of Urology (Prof J Hugosson MD, S Carlsson MD, G Aus MD, S Bergdahl MD, A Khatami MD, P Lodding MD, J Stranne MD), and Oncology (E Holmberg PhD), Institute of Clinical Sciences, Sahlgrenska Academy at University of Göteborg, Sweden; Department of Pathology, Institute of Biomedicine, Sahlgrenska Academy at University of Göteborg, Sweden (C G Pihl MD); Department of Laboratory Medicine, Lund University, University Hospital, Malmö, Sweden (Prof H Lilja MD); and Department of Clinical Laboratories, Urology, and Genitourinary Oncology (H Lilja), Memorial Sloan-Kettering Cancer Center, New York, NY, USA

Correspondence to: Prof Jonas Hugosson, Institute of Clinical Sciences, Department of Urology, Sahlgrenska University Hospital, Bruna stråket 11 B, 41345 Göteborg, Sweden [email protected]

Mortality results from the Göteborg randomised population-based prostate-cancer screening trial Jonas Hugosson, Sigrid Carlsson, Gunnar Aus, Svante Bergdahl, Ali Khatami, Pär Lodding, Carl-Gustaf Pihl, Johan Stranne, Erik Holmberg, Hans Lilja

Summary Background Prostate cancer is one of the leading causes of death from malignant disease among men in the developed world. One strategy to decrease the risk of death from this disease is screening with prostate-specifi c antigen (PSA); however, the extent of benefi t and harm with such screening is under continuous debate.

Methods In December, 1994, 20 000 men born between 1930 and 1944, randomly sampled from the population register, were randomised by computer in a 1:1 ratio to either a screening group invited for PSA testing every 2 years (n=10 000) or to a control group not invited (n=10 000). Men in the screening group were invited up to the upper age limit (median 69, range 67–71 years) and only men with raised PSA concentrations were off ered additional tests such as digital rectal examination and prostate biopsies. The primary endpoint was prostate-cancer specifi c mortality, analysed according to the intention-to-screen principle. The study is ongoing, with men who have not reached the upper age limit invited for PSA testing. This is the fi rst planned report on cumulative prostate-cancer incidence and mortality calculated up to Dec 31, 2008. This study is registered as an International Standard Randomised Controlled Trial ISRCTN54449243.

Findings In each group, 48 men were excluded from the analysis because of death or emigration before the randomisation date, or prevalent prostate cancer. In men randomised to screening, 7578 (76%) of 9952 attended at least once. During a median follow-up of 14 years, 1138 men in the screening group and 718 in the control group were diagnosed with prostate cancer, resulting in a cumulative prostate-cancer incidence of 12·7% in the screening group and 8·2% in the control group (hazard ratio 1·64; 95% CI 1·50–1·80; p<0·0001). The absolute cumulative risk reduction of death from prostate cancer at 14 years was 0·40% (95% CI 0·17–0·64), from 0·90% in the control group to 0·50% in the screening group. The rate ratio for death from prostate cancer was 0·56 (95% CI 0·39–0·82; p=0·002) in the screening compared with the control group. The rate ratio of death from prostate cancer for attendees compared with the control group was 0·44 (95% CI 0·28–0·68; p=0·0002). Overall, 293 (95% CI 177–799) men needed to be invited for screening and 12 to be diagnosed to prevent one prostate cancer death.

Interpretation This study shows that prostate cancer mortality was reduced almost by half over 14 years. However, the risk of over-diagnosis is substantial and the number needed to treat is at least as high as in breast-cancer screening programmes. The benefi t of prostate-cancer screening compares favourably to other cancer screening programs.

Funding The Swedish Cancer Society, the Swedish Research Council, and the National Cancer Institute.

Introduction The European Randomised Study of Screening for Prostate Cancer (ERSPC) compared a group of men invited for prostate-cancer screening based on prostate- specifi c antigen (PSA) with a control group without any active intervention. Interim analyses, based on a median follow-up of 9 years,1,2 showed that men randomised to active screening had a signifi cant reduction in prostate- cancer mortality; rate ratio (RR) 0·80 (95% CI 0·65–0·98, adjusted p=0·04).1 The number of men needed to be screened (NNS) to prevent one death from prostate cancer was 1410 (or 1068 in men who were actually screened1), which is similar to breast and colorectal cancer screening.3–6 However, the number of men needed to treat (NNT) to prevent one death was high (48 men), which might be explained by only 9 years of follow-up or by screening that resulted in the detection of a large proportion of indolent cancers.

These reports provide the fi rst level one evidence that PSA-based prostate-cancer screening can reduce prostate- cancer mortality. An open question, however, is whether the modest benefi t in reduced cancer mortality documented thus far outweighs the harms of over- detection. This issue is emphasised by the report from another large screening trial, the US-based Prostate, Lung, Colorectal, and Ovarian Cancer (PLCO) screening trial, which found no diff erence in prostate-cancer mortality between men randomised to screening and those in the control group at 11·5 years of follow-up.7 Other randomised studies have either been too small8,9 or criticised for methodological problems.10,11

The Göteborg randomised population-based prostate- cancer screening trial is a prospective randomised trial, planned and started in 1995, assessing the eff ects of PSA-based screening every 2 years. The trial is truly population-based, as individuals from the population

Articles

726 www.thelancet.com/oncology Vol 11 August 2010

register were randomised to screening or control groups without prior information, which results in a more representative study than randomisation after informed consent. The study design allows the analysis of both how a screening programme will be accepted by the population and its eff ectiveness in terms of prostate- cancer mortality reduction at a population level. The trial was designed and initiated independently from the ERSPC, although it was subsequently agreed to include a subset of participants in the ERSPC. According to the ethical committee approval from 1994, an analysis of this study was planned for after 15 years. The present report is the fi rst publication from the Göteborg trial assessing prostate-cancer mortality.

Methods Participants As of Dec 31, 1994, the population register documented 32 298 men born between 1930 and 1944 (age 50–64, median 56 years) living in the city of Göteborg, Sweden. By computer randomisation 20 000 of these men were identifi ed and allocated to either the intervention arm (screening group) or to a control group. The number of men in each birth cohort (1930–34, 1935–39, and 1940–44) was calculated to be proportional to the distribution in the original cohort. This resulted in larger birth cohorts from the 1940s than those from the early 1930s. The ethical review committee at the University of Göteborg approved this study in 1994.

Randomisation and masking The randomisation procedure was done at the Department of Statistics at the University of Göteborg. 10-digit personal identifi ers were the only available personal data for those doing the computer randomisation. No informed consent was needed from those in the control group. Masking of the group assignment was only done for the cause of death committee. However, possible discrepancies, caused by group assignments, were analysed for diff erences in the treatments given.

Procedures Invitations to screening began in January, 1995, and in 1996 the study became associated with the ERSPC without any changes in the protocol. Results from the men born between 1930 and 1939 have been published within the previous ERSPC report.1

Men allocated to the screening group were invited for PSA testing every second year, until they reached the upper age limit;12 the mean age at last invitation to screening was 69 years (67–71). The written invitation informed men about the study design, the complexity of PSA screening, and the voluntary nature of participation. Blood was processed within 3 h of venipuncture, frozen, and shipped frozen on dry ice for analyses within 2 weeks of the blood draw. Total PSA was measured using dual- label DELFIA Prostatus total/free PSA-assay (Perkin- Elmer, Turku, Finland). Calibration of this assay changed in 2004 to refl ect the WHO 96/670 calibrator;13,14 a correction factor was applied to the earlier measurements, and all fi gures given in this paper are in accordance with this calibration.

The PSA threshold needed to invite men to further urological work-up was 3·4 ng/mL (WHO corrected value; the nominal value was 3·0 ng/mL) between 1995 and 1998; in 1999 the threshold was changed to 2·9 ng/mL (nominal value 2·5 ng/mL) for consistency with other ERSPC sites. Due to the change of assay-calibrator, the threshold changed again to 2·5 ng/mL at the start of 2005.

Men with PSA below the threshold did not have further assessment, but were invited again after 2 years. Only men with PSA at or above the threshold were invited for further urological work-up, which included digital rectal examination (DRE), trans-rectal ultrasound (TRUS) examination, and laterally directed sextant biopsies. For men diagnosed with prostate cancer, the protocol did not specify any particular treatment; further evaluation and treatment was at the discretion of their physicians. Men with a benign fi nding at biopsy were invited for screening again after 2 years. Men with persistently raised PSA concentrations were recommended to have a new prostate biopsy at each visit at which PSA was raised. Seven screening rounds were completed by the end of 2008. Minor changes in the screening algorithm have been made during the study period.15

In both arms of the study, the incidence of prostate cancer was checked by linking with the West Swedish

Figure 1: Trial profi le PSA=prostate-specifi c antigen. PC=prostate cancer.

32 298 men in Göteborg on Dec 31, 1994, aged 50–64 years

20 000 randomised in a 1:1 ratio

9952 invited every 2 years for PSA testing 1995–2008

(screening group)

7578 attendees 2374 non-attendees

1046 with PC 27 died from PC

92 with PC 17 died from PC

9952 not invited (control group)

48 excluded 19 deceased or emigrated before randomisation date 29 men with prevalent prostate cancer

48 excluded 21 deceased or emigrated before randomisation date 27 men with prevalent prostate cancer

718 with PC 78 died from PC

Articles

www.thelancet.com/oncology Vol 11 August 2010 727

Regional Cancer Registry every third month from the start of the study. In 2009, we linked with all six regional cancer registries in Sweden and obtained data for prostate cancers diagnosed from Jan 1, 1995, through to Dec 31, 2008. For every man with prostate cancer, all available medical documentation was retrieved to establish tumour stage, treatment, and disease course. Additionally, for all deceased men we obtained a copy of the cause of death (COD) certifi cate. Two cases of prostate cancer, not registered in the regional cancer registries, were detected from COD certifi cates. Linkage with the population register was done every third month to identify all men who died or emigrated. The last date of follow-up was the date of death, date of emigration, or Dec 31, 2008.

COD for men diagnosed with prostate cancer was determined by an independent COD committee. The committee did a blinded review of all cases diagnosed with prostate cancer, including all medical records, pathology reports, and autopsy protocols, according to a standard algorithm used in the ERSPC.16 The COD certifi cates were not available to the COD committee. Deaths classifi ed as defi nitive prostate-cancer deaths, intervention-related deaths (ie, deaths from diagnostic procedures or treatment), or probable prostate-cancer deaths were regarded as deaths caused by prostate cancer, whereas other classifi cations were regarded as non- prostate cancer deaths.

Statistical analysis The main outcome measures were absolute and relative- risk reduction in cumulative prostate-cancer mortality between study arms. Secondary measures were the cumulative prostate-cancer incidence and the proportion of screening attendees. A pre-study power calculation (two-sided test; p<0·05 and 80% power) was done with the assumption of a 70% participation rate. A 40% mortality diff erence between the study arms was calculated to become signifi cant 15 years after the study began (Dec 31, 2009). A new power calculation in 2009 incorporated the observed 76% participation rate in the Swedish branch of the published ERSPC results;1 the new calculation suggested that the study has suffi cient power to allow analyses to be done a year early.

Cumulative incidences of prostate cancer in the screening and control groups were plotted as 1 minus the Kaplan- Meier estimator. The corresponding hazard ratio (HR) for the incidence of prostate cancer between the groups was estimated by Cox regression and the proportional hazard assumption was tested with Schoenfeld residuals.17 A time- dependent covariate approach was used to estimate the HR at diff erent time periods after the start of screening to avoid violation of the proportional hazard assumption. The Nelson-Aalen method was used to calculate the cumulative hazard for prostate-cancer mortality.18 Poisson-regression analysis was used to estimate the mortality-rate ratio in the screening group versus the control group. All p values

were two-sided. NNS was calculated as 1 divided by absolute reduction in prostate-cancer mortality. As this study is an intention-to-screen analysis, we refer to NNS as the number needed to invite for screening. The NNT was calculated as 1 divided by (absolute reduction in prostate- cancer mortality multiplied by excess prostate-cancer incidence); we renamed this measure as number needed

Screening visit Total

1st 2nd 3rd 4th 5th 6th 7th

1st invitation round (1995–96)

Number of men invited .. .. .. .. .. .. .. 9890

Number of men participating 5855 .. .. .. .. .. .. 5855

Number of men with raised PSA 661 .. .. .. .. .. .. 661

Number of men with PC 144 .. .. .. .. .. .. 144

2nd invitation round (1997–99)

Number of men invited .. .. .. .. .. .. .. 9525

Number of men participating 580 4680 .. .. .. .. .. 5260

Number of men with raised PSA 66 543 .. .. .. .. .. 609

Number of men with PC 15 98 .. .. .. .. .. 113

3rd invitation round (1999–2000)*

Number of men invited .. .. .. .. .. .. .. 6920

Number of men participating 460 632 2283 .. .. .. .. 3375

Number of men with raised PSA 79 130 621 .. .. .. .. 830

Number of men with PC 29 23 108 .. .. .. .. 160

4th invitation round (2001–02)

Number of men invited .. .. .. .. .. .. .. 7873

Number of men participating 291 549 2251 1531 .. .. .. 4622

Number of men with raised PSA 49 63 125 497 .. .. .. 734

Number of men with PC 13 13 19 87 .. .. .. 132

5th invitation round (2003–04)

Number of men invited .. .. .. .. .. .. .. 6598

Number of men participating 207 342 547 1880 1138 .. .. 4114

Number of men with raised PSA 38 62 54 110 351 .. .. 615

Number of men with PC 9 11 6 20 65 .. .. 111

6th invitation round (2005–06)

Number of men invited .. .. .. .. .. .. .. 5733

Number of men participating 117 188 296 468 1556 850 .. 3475

Number of men with raised PSA 34 34 51 61 104 418 .. 702

Number of men with PC 13 6 14 11 20 81 145

7th invitation round (2007–08)

Number of men invited .. .. .. .. .. .. .. 4148

Number of men participating 68 94 145 241 374 1157 535 2614

Number of men with raised PSA 20 11 24 42 64 87 294 542

Number of men with PC 8 3 3 11 10 11 45 91

Total (1995–2008)

Total number of invitations in the study .. .. .. .. .. .. .. 50 687

Number of men participating 7578 6334 3794 4393 3325 2452 1439 29 315†

Number of men with raised PSA 947 843 875 710 519 505 294 4693‡

Number of men with PC 231 154 150 129 95 92 45 896

PSA=prostate-specifi c antigen. PC=prostate cancer. *The low attendance rate in the third invitation round was because men with total PSA<1 ng/mL in the second invitation round were not invited (except those born 1930–31). †The total number of PSA tests done in the study. ‡The total number of PSA tests exceeding the PSA cutoff during the study.

Table 1: Number and outcome of participants in relation to screening visit

Articles

728 www.thelancet.com/oncology Vol 11 August 2010

to diagnose, because many patients were not actually treated. The analyses were done using Stata, release 11. This study is registered with controlled-trials.com, number ISRCTN54449243.

Role of the funding source The funding sources had no role in the study design and conduct, collection, management, analysis, and interpretation of the data, or writing of the report. The funding sources had no access to the database, which is kept at the Sahlgrenska University Hospital. All authors of this manuscript have had full access to the database. JH had the fi nal responsibility to submit the paper for publication.

Results The trial profi le is shown in fi gure 1. Subsequent to randomisation, we excluded from analysis 56 men with a prior diagnosis of prostate cancer, 34 who had died, and

six who had emigrated but had not been removed from the population register at the time of randomisation. Thus, the screening and control groups each consisted of 9952 evaluable men. In the screening group, 7578 (76%) of 9952 men participated in at least one screening round (attendees; table 1). These men received 29 315 PSA tests during the study period. In 2469 (33%) of 7578 attendees, PSA was raised above the threshold at least once and a total of 4693 elevated PSA tests were recorded during the study (table 1). In men with raised PSA, 2298 (93%) of 2469 had prostate biopsy at least once; 4153 biopsy procedures were done in the study. The maximum follow- up time of 14 years was reached by 15 501 (78%) of the randomised men.

Prostate cancer was diagnosed in 1138 (11·4%) men in the screening group and 718 (7·2%) in the control group (table 2). Of those men with detected prostate cancer in the screening group, 896 (78·7%) of 1138 were diagnosed as a result of an invitation to the study (table 1). Of these 896 men, 231 were detected at their fi rst screening visit and 665 during subsequent screening rounds. At the fi rst screening visit 3671 (48·4%) of 7578 had PSA below 1·00 ng/mL, 2960 (39·1%) of 7578 had a PSA between 1·00 and 2·99 ng/mL, and 947 (12·5%) of 7578 had a PSA of 3·00 ng/mL or greater; the risk of being diagnosed with prostate cancer during follow-up was 2·6%, 17·6%, and 45·5% respectively. The cumulative incidence of prostate cancer at 14 years was 12·7% in the screening group versus 8·2% in the control group (HR 1·64, 95% CI 1·50–1·80; p<0.0001; fi gure 2). During the fi rst year, after the start of screening, the HR was 5·2 (95% CI 3·1–8·6), which subsequently decreased to 3·7 (95% CI 2·2–6·2) at 1–2 years, 2·6 (95% CI 1·9–-3·.6) at 2–4 years, 2·1 (95% CI 1·7–2·7) at 4–6 years, 1·7 (95% CI 1·3–2·1) at 6–8 years, and 1·2 (95% CI 1·0–1·3) at 8 years or more (fi gure 2).

Most of the prostate cancers diagnosed in the screening group were early-stage disease (table 2). The number of men with advanced prostate cancer (metastases or PSA >100 ng/mL at diagnosis) was lower in the screening group than in the control group (46 men vs 87; p=0·0003; table 2). Notably, in non-attendees in the screening group, a high proportion of cancers were advanced at diagnosis (table 2).

The diff erence in stage distribution was mirrored by the treatment diff erence, with more hormonal therapy used in the control group than in the screening group and more surveillance or treatment with curative intent in the screening group than in the control group (table 3). However, in men with low- and moderate-risk tumours, the proportion having curative treatment was similar between groups: 476 (49·2%) of 967 in the screening group and 228 (50·8%) of 448 in the control group. In the men diagnosed with prostate cancer, the median follow-up after diagnosis was 6·7 (IQR 3·1–9·5) years in the screening group and 4·3 (2·1–7·1) years in the control group. The COD committee and COD certifi cates were highly concordant in assessing whether the deaths

Control group (n=9952)

Screening group (n=9952)

All (n=9952) Attendees (n=7578)

Non-attendees (n=2374)

Number of men with prostate cancers diagnosed (%)

718 (7·2%) 1138 (11·4%) 1046 (13·8%) 92 (3·9%)

Tumour grouping (%)

Low risk* 199 (2%) 604 (6·1%) 590 (7·8%) 14 (0·6%)

Moderate risk† 249 (2·5%) 363 (3·6%) 339 (4·5%) 24 (1%)

High risk‡ 126 (1·3%) 96 (1%) 76 (1%) 20 (0.8%)

Advanced disease§ 87 (0·9%) 46 (0·5%) 25 (0·3%) 21 (0·9%)

Unknown¶ 57 (0·6%) 29 (0·3%) 16 (0·2%) 13 (0·5%)

*T1, not N1 or M1, and Gleason score ≤6 and prostate-specifi c antigen <10 ng/mL.†T1–2, but not N1 or M1, with a Gleason score ≤7, prostate-specifi c antigen <20 ng/mL or both; and not meeting the criteria for low risk.‡T1–4, but not N1 or M1, with a Gleason score ≥8, prostate-specifi c antigen <100 ng/mL, or both; and not meeting the criteria for low or moderate risk.§N1 or M1, or prostate-specifi c antigen ≥100 ng/mL.¶Includes seven cases detected at autopsy.

Table 2: Prostate cancers diagnosed in the study groups

Figure 2: Cumulative incidence of prostate cancer in the screening group and in the control group

0

0·02

0·04

0·06

0·08

0·10

0·12

0·14

Pr ob

ab ili

ty o

f p ro

st at

e ca

nc er

d ia

gn os

is

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

9952 9952

8961 9214

7847 8185

6761 7168

Time from randomisation (years)

Screening group Control group

Number at risk Screening group

Control group

Articles

www.thelancet.com/oncology Vol 11 August 2010 729

were caused by prostate cancer. According to the COD committee review, 78 men in the control group died from prostate cancer (77 according to death certifi cates) compared with 44 in the screening group (45 according to death certifi cates). Within the screening group, 27 (0·4%) of 7578 prostate-cancer-specifi c deaths were registered among attendees versus 17 (0·7%) of 2374 non-attendees (fi gure 1; table 4). Of the attendees who died from prostate cancer, 13 were diagnosed with prostate cancer at fi rst screening (prevalence screen); the youngest of these men was 59 years of age at diagnosis. Attendees who were older than 60 years of age at study entry seemed to have a higher risk of dying from prostate cancer (19 prostate-cancer deaths in attendees in the screening group vs 35 prostate-cancer deaths in the control group) compared with men younger than 60 years of age at study entry (8 prostate cancer deaths in attendees in the screening group vs 43 prostate cancer deaths in the control group) (table 4).

The RR of dying from prostate cancer was 0·56 (95% CI 0·39–0·82; p=0·002) in the screening group compared with the control group (fi gure 3). The absolute cumulative-risk reduction (Kaplan-Meier estimates) of death from prostate cancer at 14 years was 0·40% (95% CI 0·17–0·64), from 0·90% in the control group to 0·50% in the screening group. A secondary analysis showed that the RR of death from prostate cancer for attendees compared with the control group was 0·44 (95% CI 0·28–0·68; p=0·0002) and the RR of death from prostate cancer for non-attendees compared with the control group was 1·05 (95% CI 0·62–1·78 p=0·84). The number of men with prostate cancer who died from unrelated causes was 109 (9·6%) of 1138 in the screening group and 54 (7·5%) of 718 in the control group. However, the follow-up time was longer for the men diagnosed with prostate cancer in the screening group than the control group (6·7 vs 4·3 years). Therefore, the cumulative risk (Kaplan-Meier estimates) of non- prostate cancer deaths measured from the date of prostate-cancer diagnosis was similar at 10 years, 13·1% in the screening group and 15·0% in the control group (log-rank test p=0·50).

The NNS to prevent one prostate-cancer death was 293 (95% CI 177–799), whereas the NNT was 12. If the calculations were restricted to attendees, the respective numbers were 234 (95% CI 154–492) and 15.

Discussion The aim of this prospective, population-based randomised screening study was to assess the eff ectiveness of a screening programme in which men were fi rst randomised and then asked to participate. The design gives more representative results than does randomisation after informed consent, and mirrors the situation when screening is introduced in the population. The study yielded two major fi ndings. First, a PSA-based screening programme is acceptable to men aged 50 years

or older, with 76% attending at least once. Second, with such a participation rate, a screening programme will decrease prostate-cancer mortality by as much as half over 14 years’ follow-up.

Half of the attendees who died from prostate cancer were diagnosed at their fi rst screening visit and many of these men were 60 years of age or older at study entry. In a programme in which all men started screening at 50 years of age, some men could instead be diagnosed at a curable stage; therefore, potential for larger mortality reduction exists (table 4).12

Control group (n=718)

Screening group (n=1138)

All (n=1138) Attendees (n=1046)

Non-attendees (n=92)

Primary radical prostatectomy* 241 (33·6%) 468 (41·1%) 439 (42·0%) 29 (31·5%)

Primary radiation 75 (10·4%) 93 (8·2%) 81 (7·7%) 12 (13·0%)

Primary endocrine treatment 162 (22·6%) 80 (7·0%) 47 (4·5%) 33 (35·9%)

Primary surveillance followed by curative treatment†

36 (5·0%) 142 (12·5%) 141 (13·5%) 1 (1·1%)

Primary surveillance followed by endocrine treatment

20 (2·8%) 23 (2·0%) 21 (2·0%) 2 (2·2%)

Surveillance at last follow-up 152 (21·2%) 314 (27·6%) 301 (28·8%) 13 (14·1%)

Not treated‡ 32 (4·5%) 18 (1·6%) 16 (1·5%) 2 (2·2%)

Data are n (%). *Includes nine cryosurgeries and six cystoprostatectomies.†Includes two cystoprostatectomies. ‡Includes seven cases detected at autopsy.

Table 3: Treatments for prostate cancer, by study group

Total Control group Screening group

All Attendees Non-attendees

1930–34

Total number 5563 2789 2774 2064 710

Number with PC 615 259 356 318 38

Number of deaths 1689 853 836 488 348

Number of PC deaths 62 35 27 19 8

1935–39

Total number 6284 3161 3123 2420 703

Number with PC 654 252 402 372 30

Number of deaths 1284 650 634 360 274

Number of PC deaths 47 35 12 6 6

1940–44

Total number 8057 4002 4055 3094 961

Number with PC 587 207 380 356 24

Number of deaths 990 479 511 267 244

Number of PC deaths 13 8 5 2 3

Total

Total number 19 904 9952 9952 7578 2374

Number with PC 1856 718 1138 1046 92

Number of deaths 3963 1982 1981 1115 866

Number of PC deaths 122 78 44 27 17

PC=prostate cancer.

Table 4: Outcome of men in relation to birth cohort at entry to the study

Articles

730 www.thelancet.com/oncology Vol 11 August 2010

This study shows a much higher mortality reduction than in previous studies: an RR of 0·56 in this study compared with 0·80 in the ERSPC (0·84 if the Swedish cohort is withdrawn),1 and no reduction in the PLCO study.7 Several factors might account for this. First, the men in our study were younger (median age 56 years at baseline) than in both previous publications (median age >60 years). Younger men are less likely than older men to have incurable prostate cancer at the fi rst screening and are therefore more likely to gain the full benefi t of screening. Second, the PSA threshold for biopsy was lower in our study than in most other ERSPC branches and in the PLCO trial. However, DRE was never used as a screening tool in our study, but was used by most ERSPC centres at the fi rst screening round and in the design of PLCO trial. Addition of DRE in our study might have resulted in an even larger mortality reduction than seen in our study, although only a few incurable cancers were found in men who attended the programme, and some of these incurable cancers were still non-palpable at diagnosis.12 Third, the interval of screening in this study (every 2 years) was shorter than in the other ERSPC branches (every 4 years), although longer than in the PLCO trial (every year). Fourth, this study had a much higher rate of biopsy for men with a positive screening result (93% vs 30–40% in PLCO19), a much lower rate of PSA testing before the start of the study (estimated as 3% vs 44% in PLCO), and probably a lower rate of contamination in the control group than in the PLCO trial. Fifth, the present study has much longer follow-up than do the ERSPC and PLCO studies (median 14 years from randomisation vs 9 years for ERSPC,1 and 11·5 years for PLCO7).

Up to 10 years of follow-up, the Nelson-Aalen plot in our study resembles that which was published in the ERSPC study, suggesting that most of the benefi t from screening occurs after 10 years (fi gure 3). This is to be expected from a disease with long a lead-time and a long natural course.20,21

Although the median follow-up from randomisation is long, the follow-up time measured from prostate-cancer diagnosis is rather short; 6·7 years for attendees versus 4·3 years for controls in this study compared with 6·3 versus 5·2 years in the PLCO study.7

The reasons as to why our study shows an important mortality reduction and the PLCO trial did not, despite a similar follow-up after diagnosis in the two studies, might in part be explained by the absence of pre-screening in our study, which meant many aggressive cancers were still detectable. Furthermore, contamination in the control group was low—at least during the fi rst 5 years of our study. An indication of these important diff erences is that despite the randomisation of 76 693 men in the PLCO trial versus 19 904 in our study, only 174 prostate-cancer deaths were recorded in the PLCO trial7 compared with 122 in our study. The men in the PLCO trial were also older.

The RR of 0·56 within 14 years corresponds to an absolute risk reduction of 0·40% and no eff ect on overall mortality (similar number of men at risk at 14 years, [fi gure 3] and similar number of total deaths in the study group [table 4]). These low mortality fi gures are related to the young age of participants at the start of the study and the comparatively short follow-up after prostate-cancer diagnosis. Because about 5% of deaths among Swedish men are caused by prostate cancer,22 it is obvious that we have so far studied only the early eff ects of screening. If the relative-risk reduction is sustained over time the mortality reduction, even measured in terms of absolute- risk reduction, might become important. An indication of this is the large diff erence between the arms in the number of men needing endocrine treatment—182 (1·8%) in the control group versus 103 (1·0%) in the screening group. The fact that 79 more men in the control group were treated with endocrine treatment than in the screening group might also be regarded as an important advantage. The increased ratio of unrelated deaths reported in the screening group compared with the control group (9·6% vs 7·5%) is explained by the longer follow-up of patients with prostate cancer in the screening group than in the control group, because there is no diff erence in non-prostate-cancer mortality if Kaplan- Meier estimates are calculated from diagnosis.

The high rate of attendance to this PSA-based screening programme is corroborated by fi ndings from several uncontrolled trials. Bartsch and co-workers23 reported that 86·6% of men accepted an off er of a free PSA test and that 85·0% of those with raised PSA concentrations consented to additional urological assessment with prostate biopsies.23 Moreover, all centres in the ERSPC study reported a high acceptance rate for screening.1 The screening procedures with PSA testing and prostate biopsy are seldom associated with severe psychological distress, even for men with repeatedly raised PSA concentrations.24,25 We therefore conclude that acceptance is not an obstacle for a population-based prostate-cancer- screening programme.

Figure 3: Cumulative risk of death from prostate cancer using Nelson-Aalen cumulative hazard estimates

0

0·002

0·004

0·006

0·008

0·010

N el

so n-

A al

en c

um ul

at iv

e ha

za rd

e st

im at

es

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

9952 9952

9333 9345

8585 8580

7746 7755

Time from randomisation (years)

Screening group Control group

Number at risk Screening group

Control group

Articles

www.thelancet.com/oncology Vol 11 August 2010 731

Diff erences between screening and control groups in cancer stage and grade show the cancer stage migration introduced by a PSA-based screening programme. Although 1·6 times as many prostate cancers were diagnosed in the screening group, the absolute number of patients with advanced disease was lower in the screening group than in the control group. Therefore, screening caused a true stage migration and resulted in a diff erent distribution of treatments between the two groups. However, in men with early cancer (low- and moderate-risk cancer), treatment with curative intent was as common in the control and screening groups (51% vs 49%), suggesting that the mortality diff erence resulted from screening and not from diff erent treatments.

At 14 years of follow-up, the number who needed to be invited to screening (corresponding to NNS) to prevent one prostate cancer death was 293, and the number who needed to be diagnosed (corresponding to NNT) was 12. These fi gures, and the RR of 0·56 in our study, can be compared with those of the commonly recommended practices of screening for breast and colon cancer. Because these fi gures are time-dependent, we focus this comparison on studies with similar follow-up periods. For mammography, a 2009 meta-analysis of randomised trials showed a number needed to invite to screening of 377 (credible interval 230–1050) for women aged 60–69 years and 1339 (credible interval 322–7455) for women aged 50–59 years, and RRs of 0·68 and 0·86 respectively at 11–20 years of follow-up.5 Individual studies included in the meta-analysis, as well as other mammography studies, have shown similar numbers.3,26–33 In a 2009 Cochrane review, the NNT for mammography was 10 over 10 years.3 For colorectal-cancer screening by faecal occult-blood test, the RRs varied between 0·67 and 0·87 in four randomised trials34–37 and was 0·84 overall in both a 2008 Cochrane review4 (after 11·7–18·4 years) and a meta-analysis by Towler and colleagues6 (after 7·8–13·0 years). Towler and colleagues6 estimated the NNS after 10 years to be 1173 (95% CI 741–2807). Moreover, a multicentre study has reported an RR of 0·69 for colorectal-cancer mortality, with fl exible-sigmoidoscopy screening for colorectal cancer, and an NNS of 489 at a median follow-up of 11·2 years.38 The NNT cannot be calculated for comparison because screening for colorectal cancer is associated with a reduced colorectal-cancer incidence.

The NNT in our study is substantially lower than that in the ERSPC study,1 which suggests that the NNT is very dependent on the length of follow-up. It is not easy to predict at which follow-up period the NNT will stabilise. Furthermore, since NNT in prostate-cancer screening mainly refl ects the risk of over-diagnosis, it is not easy at this point to make estimates of this risk but it is probably not as high as some have feared,39 at least if screening is restricted to the age groups included in this study. As many as 314 (30%) screening attendees were in active surveillance at last follow-up in this study. The strategy

of active surveillance will at least reduce the risk of over- treatment and the risk associated seems low.40

Since the benefi t from prostate-cancer screening takes a long time to achieve—only marginal benefi ts are gained within the fi rst 10 years of starting prostate-cancer screening—one should be cautious to recommend that all elderly men have PSA screening. As the risk of over- diagnosis and over-treatment are still the major concerns in prostate-cancer screening, inviting men over the age of 70 for PSA screening seems questionable.

In summary, in this trial prostate-cancer screening was well accepted by the general population and can result in a relevant reduction in cancer mortality, greater than that reported in screening for breast or colorectal cancer. Nevertheless, PSA screening is associated with a long and varying lead time, resulting in a risk of over-diagnosis that is substantial but still of a largely unknown magnitude.

Contributors JH had full access to all the data in the study and takes responsibility for

the integrity of the data and the accuracy of the data analysis. JH is the

principal investigator of the study and was responsible for planning of

the study. HL participated in the conception and design of the study and

contributed with supervision and administrative support. JH and HL

were responsible for funding. JH, AK, PL, JS, SB, and C-GP collected

data. JH, GA, SC, AK, PL, JS, and SB did the biopsy procedures. C-GP

did the pathological examination of all specimens in the study. EH, JH,

SC, HL, and GA analysed the data (extraction of results and the statistical

analysis). JH, JS, SC, HL, and GA interpreted the data. SC performed the

literature search. JH, SC, GA, HL, JS wrote, JH, SC, GA, HL, PL, AK, SB

revised, and EH and C-GP reviewed the paper.

Confl ict of interest HL holds patents for free PSA and hK2 assays, and has received

honoraria from GlaxoSmithKline. JH has received lecture fees from

GlaxoSmithKline and Abbott Pharmaceuticals. All other authors

declared no confl icts of interest.

Acknowledgments This study was supported by the Swedish Cancer Society (Contract

numbers 09 0107, 080315, and 083455), the Swedish Research Council

(Medicine) (20095) and the National Cancer Institute grant number

(R21-CA127768-01A1). Grants were also received from the Stichting af

Jochnick Foundation, Catarina and Sven Hagstroms family foundation,

Gunvor and Ivan Svensson’s foundation, Johanniterorden, King Gustav

V Jubilée Clinic Cancer Research Foundation, Sahlgrenska University

Hospital, Abbott Pharmaceuticals (Sweden), and Schering Plough

(Sweden). We also acknowledge the Sidney Kimmel Centre for Prostate

and Urologic Cancers, David H Koch through the Prostate Cancer

Foundation, and Fundación Federico SA for their funding support.

We thank the COD committee (Bo Johan Norlén, Silas Pettersson, and

Eberhard Varenhorst); Helén Ahlgren; Maria Nyberg; Charlotte Becker,

Gun-Britt Eriksson, Kerstin Håkansson, and Mona Hassan Al-Battat for

expert assistance with immunoassays; Björn Zackrisson, Erik Pileblad,

Rebecka Godtman, and Anna Grenabo Bergdahl for various aspects in

this screening project and especially for helping with all biopsies

performed; Janet Novak for her assistance with editing the manuscript,

which was paid for by Memorial Sloan-Kettering Cancer Center; and

Andrew Vickers for valuable criticism and suggestions.

References 1 Schroder FH, Hugosson J, Roobol MJ, et al. Screening and

prostate-cancer mortality in a randomized European study. N Engl J Med 2009; 360: 1320–28.

2 Roobol MJ, Kerkhof M, Schroder FH, et al. Prostate cancer mortality reduction by prostate-specifi c antigen-based screening adjusted for nonattendance and contamination in the European Randomised Study of Screening for Prostate Cancer (ERSPC). Eur Urol 2009; 56: 584–91.

Articles

732 www.thelancet.com/oncology Vol 11 August 2010

3 Gotzsche PC, Nielsen M. Screening for breast cancer with mammography. Cochrane Database Syst Rev 2009; 4: CD001877.

4 Hewitson P, Glasziou P, Watson E, Towler B, Irwig L. Cochrane systematic review of colorectal cancer screening using the fecal occult blood test (hemoccult): an update. Am J Gastroenterol 2008; 103: 1541–49.

5 Nelson HD, Tyne K, Naik A, , et al. Screening for breast cancer: an update for the U.S. Preventive Services Task Force. Ann Intern Med 2009; 151: 727–37.

6 Towler B, Irwig L, Glasziou P, Kewenter J, Weller D, Silagy C. A systematic review of the eff ects of screening for colorectal cancer using the faecal occult blood test, hemoccult. BMJ 1998; 317: 559–65.

7 Andriole GL, Grubb RL, 3rd, Buys SS, Chia D, Church TR, Fouad MN, et al. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med 2009; 360: 1310–19.

8 Sandblom G, Varenhorst E, Lofman O, Rosell J, Carlsson P. Clinical consequences of screening for prostate cancer: 15 years follow-up of a randomised controlled trial in Sweden. Eur Urol 2004; 46: 717–23.

9 Kjellman A, Akre O, Norming U, Tornblom M, Gustafsson O. 15-year followup of a population based prostate cancer screening study. J Urol 2009; 181: 1615–21.

10 Labrie F, Candas B, Cusan L, et al. Screening decreases prostate cancer mortality: 11-year follow-up of the 1988 Quebec prospective randomized controlled trial. Prostate 2004; 59: 311–18.

11 Ilic D, O’Connor D, Green S, Wilt TA. Screening for prostate cancer: A Cochrane systematic review. Cancer Causes Control 2007; 18: 279–85.

12 Bergdahl AG, Aus G, Lilja H, Hugosson J. Risk of dying from prostate cancer in men randomized to screening: diff erences between attendees and nonattendees. Cancer 2009; 115: 5672–79.

13 World Health Organization. WHO Expert Committee on Biological Standardization. Forty-eighth report. WHO technical report series, no. 889. Geneva, Switzerland: World Health Organization, 1999.

14 Raff erty B, Rigsby P, Rose M, Stamey T, Gaines Das R. Reference reagents for prostate-specifi c antigen (PSA): establishment of the First International Standards for free PSA and PSA (90:10). Clin Chem 2000; 46: 1310–17.

15 ERSPC website. http://media.erspc-media.org/sweden/ (accessed May 5, 2010).

16 De Koning HJ, Blom J, Merkelbach JW, et al. Determining the cause of death in randomized screening trial(s) for prostate cancer. BJU Int 2003; 92 (suppl 2): 71–78.

17 Grambsch P, Therneau T. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 1994; 81: 515–26.

18 Aalen O. Nonparametric inference for a family of counting processes. Ann Stat 1978; 6: 701–26.

19 Grubb RL, 3rd, Pinsky PF, Greenlee RT, et al. Prostate cancer screening in the Prostate, Lung, Colorectal and Ovarian cancer screening trial: update on fi ndings from the initial four rounds of screening in a randomized trial. BJU Int 2008; 102: 1524–30.

20 Tornblom M, Eriksson H, Franzen S, et al. Lead time associated with screening for prostate cancer. Int J Cancer 2004; 108: 122–29.

21 Albertsen PC, Hanley JA, Fine J. 20-year outcomes following conservative management of clinically localized prostate cancer. JAMA 2005; 293: 2095–101.

22 The National Board of Health and Welfare, 2009. Offi cial Statistics of Sweden. Causes of death 2007. http://www.socialstyrelsen.se/ publikationer2009/dodsorsaker2007 (accessed June 25, 2010).

23 Bartsch G, Horninger W, Klocker H, et al. Tyrol prostate cancer demonstration project: early detection, treatment, outcome, incidence and mortality. BJU Int 2008; 101: 809–16.

24 Brindle LA, Oliver SE, Dedman D, et al. Measuring the psychosocial impact of population-based prostate-specifi c antigen testing for prostate cancer in the UK. BJU Int 2006; 98: 777–82.

25 Carlsson S, Aus G, Wessman C, Hugosson J. Anxiety associated with prostate cancer screening with special reference to men with a positive screening test (elevated PSA) - results from a prospective, population-based, randomised study. Eur J Cancer 2007; 43: 2109–16.

26 Miller AB, To T, Baines CJ, Wall C. The Canadian National Breast Screening Study-1: breast cancer mortality after 11 to 16 years of follow-up. A randomized screening trial of mammography in women age 40 to 49 years. Ann Intern Med 2002; 137: 305–12.

27 Miller AB, To T, Baines CJ, Wall C. Canadian National Breast Screening Study-2: 13-year results of a randomized trial in women aged 50-59 years. J Natl Cancer Inst 2000; 92: 1490–99.

28 Nystrom L, Andersson I, Bjurstam N, Frisell J, Nordenskjold B, Rutqvist LE. Long-term eff ects of mammography screening: updated overview of the Swedish randomised trials. Lancet 2002; 359: 909–19.

29 Moss SM, Cuckle H, Evans A, et al. Eff ect of mammographic screening from age 40 years on breast cancer mortality at 10 years’ follow-up: a randomised controlled trial. Lancet 2006; 368: 2053–60.

30 Tabar L, Fagerberg G, Chen HH, et al. Effi cacy of breast cancer screening by age. New results from the Swedish Two-County Trial. Cancer 1995; 75: 2507–17.

31 Shapiro S, Venet W, Strax P, Venet L, Roeser R. Ten- to fourteen- year eff ect of screening on breast cancer mortality. J Natl Cancer Inst 1982; 69: 349–55.

32 Frisell J, Lidbrink E, Hellstrom L, Rutqvist LE. Followup after 11 years--update of mortality results in the Stockholm mammographic screening trial. Breast Cancer Res Treat 1997; 45: 263–70.

33 Bjurstam N, Bjorneld L, Warwick J, et al. The Gothenburg Breast Screening Trial. Cancer 2003; 97: 2387–96.

34 Kronborg O, Jorgensen OD, Fenger C, Rasmussen M. Randomized study of biennial screening with a faecal occult blood test: results after nine screening rounds. Scand J Gastroenterol 2004; 39: 846–51.

35 Lindholm E, Brevinge H, Haglind E. Survival benefi t in a randomized clinical trial of faecal occult blood screening for colorectal cancer. Br J Surg 2008; 95: 1029–36.

36 Mandel JS, Church TR, Ederer F, Bond JH. Colorectal cancer mortality: eff ectiveness of biennial screening for fecal occult blood. J Natl Cancer Inst 1999; 91: 434–37.

37 Scholefi eld JH, Moss S, Sufi F, Mangham CM, Hardcastle JD. Eff ect of faecal occult blood screening on mortality from colorectal cancer: results from a randomised controlled trial. Gut 2002; 50: 840–44.

38 Atkin WS, Edwards R, Kralj-Hans I, et al. Once-only fl exible sigmoidoscopy screening in prevention of colorectal cancer: a multicentre randomised controlled trial. Lancet 2010; 375: 1624–33.

39 Albertsen PC. The unintended burden of increased prostate cancer detection associated with prostate cancer screening and diagnosis. Urology 2009; 75: 399–405.

40 Khatami A, Aus G, Damber JE, Lilja H, Lodding P, Hugosson J. PSA doubling time predicts the outcome after active surveillance in screening-detected prostate cancer: results from the European randomized study of screening for prostate cancer, Sweden section. Int J Cancer 2007; 120: 170–74.