For Reseacher_D

Abdulghani Alhindi 1995

4.pdf

Home >Nursing homework help >For Reseacher_D

Bronchoalveolar Lavage and Lung Biopsy in Patients With Cancer and Hematopoietic Stem-Cell Transplantation Recipients: A Systematic Review and Meta-Analysis DeepakBabu Chellapandian, Thomas Lehrnbecher, Bob Phillips, Brian T. Fisher, Theoklis E. Zaoutis, William J. Steinbach, Joseph Beyene, and Lillian Sung

DeepakBabu Chellapandian, Joseph Beyene, and Lillian Sung, The Hospital for Sick Children, Toronto; Joseph Beyene, McMaster University, Hamil- ton, Ontario, Canada; Thomas Lehrn- becher, Johann Wolfgang Goethe University, Frankfurt, Germany; Bob Phillips, Centre for Reviews and Dissemination, University of York, York, United Kingdom; Brian T. Fisher and Theoklis E. Zaoutis, Children’s Hospital of Philadelphia, Philadelphia, PA; and William J. Steinbach, Duke University Medical Center, Durham, NC.

Published online ahead of print at www.jco.org on January 5, 2015.

Authors’ disclosures of potential conflicts of interest are found in the article online at www.jco.org. Author contributions are found at the end of this article.

Corresponding author: Lillian Sung, MD, PhD, Division of Haematology/Oncology, The Hospital for Sick Children, 555 Univer- sity Ave, Toronto, ON, Canada M5G 1X8; e-mail: lillian.sung@sickkids.ca.

0732-183X/15/3305w-501w/$20.00

DOI: 10.1200/JCO.2014.58.0480

A B S T R A C T

Purpose The objective of this study was to describe the diagnostic yield and complication rate of bronchoalveolar lavage (BAL) and lung biopsy in the evaluation of pulmonary lesions in patients with cancer and recipients of hematopoietic stem-cell transplantation (HSCT).

Methods We conducted a systematic literature review and performed electronic searches of Ovid MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials. Studies were included if patients had cancer or were recipients of HSCT, and if they underwent BAL or lung biopsy for the evaluation of pulmonary lesions. Only English language publications were included.

Results In all, 14,148 studies were screened; 72 studies of BAL and 31 of lung biopsy were included. The proportion of procedures leading to any diagnosis was similar by procedure type (0.53 v 0.54; P � .94) but an infectious diagnosis was more common with BAL compared with lung biopsy (0.49 v 0.34; P � .001). Lung biopsy more commonly led to a noninfectious diagnosis (0.43 v 0.07; P � .001) and was more likely to change how the patient was managed (0.48 v 0.31; P � .002) compared with BAL. However, complications were more common with lung biopsy (0.15 v 0.08; P � .006), and procedure-related mortality was four-fold higher for lung biopsy (0.0078) compared with BAL (0.0018).

Conclusion BAL may be the preferred diagnostic modality for the evaluation of potentially infectious pulmonary lesions because of lower complication and mortality rates; thus, choice of procedure depends on clinical suspicion of infection. Guidelines to promote consistency in the approach to the evaluation of lung infiltrates may improve clinical care of patients.

INTRODUCTION

Patients who receive chemotherapy or who undergo hematopoietic stem-cell transplantation (HSCT) experience considerable toxicities of therapy: fever and neutropenia (FN) are two of the most common complications of treatment. When FN occurs, ob- taining blood cultures from all lumens of central venous catheters and a careful clinical examination for a source of infection are recommended.1 In the setting of persistent fever without an identified source, it is essential to evaluate other potential sites of infection. The lungs are a common site of infection2 and thus should be considered in second- ary evaluations.

Radiographic investigation can reveal nonspe- cific findings such as pulmonary nodules or other

lung findings. These radiographic findings present a therapeutic dilemma because they can represent a pulmonary infection from bacteria, viruses, and/or fungi. Mold infections are of particular concern given the need for prolonged therapy and the high attributable mortality rate.3 Specimen sources for diagnostic procedures that may provide insight into the etiology of pulmonary lesions include sputum evaluation, bronchoscopy with bronchoalveolar la- vage (BAL), and lung biopsy in addition to blood and urine samples.

In addition to knowing whether pulmonary abnormalities are infectious or noninfectious, knowledge of the specific pathogen is helpful for making treatment decisions. The diagnostic capabil- ities for identifying the etiology of infectious pneu- monic processes are changing over time. There has

JOURNAL OF CLINICAL ONCOLOGY R E V I E W A R T I C L E

V O L U M E 3 3 � N U M B E R 5 � F E B R U A R Y 1 0 2 0 1 5

http://www.jco.org

mailto:lillian.sung@sickkids.ca

http://dx.doi.org/10.1200/JCO.2014.58.0480

been an increase in the use of molecular biomarkers such as galacto- mannan (GM) and polymerase chain reaction (PCR) for identifying Aspergillus and in the use of various PCR platforms for identifying non-Aspergillus fungi. Noninfectious causes of pulmonary abnormal- ities, including malignancy and hemorrhage, are also important.

We recently published a guideline for the management of FN in pediatric patients.1 In this guideline, we noted the absence of data that identify the procedure with the greatest yield and lowest procedure- related risk for the evaluation of pulmonary lesions. A comparison of the benefits and risks of BAL and lung biopsy may help guide decision making when a patient presents with a potentially infectious pulmo- nary process. A systematic literature review specifically of patients with cancer and recipients of HSCT is important because the presence of neutropenia and thrombocytopenia may change the risk-benefit profile for these procedures. Consequently, the primary objective of this systematic review was to describe the diagnostic yield of BAL and lung biopsy in the evaluation of pulmonary lesions in patients with cancer and in recipients of HSCT. The secondary objectives were to describe the rate of complications and procedure-related mortality of BAL and lung biopsy. An exploratory objective was to describe the diagnostic properties of BAL GM and Aspergillus PCR in diagnosing invasive aspergillosis.

METHODS

This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses recommendations for reporting.4

Data Sources and Searches

We performed electronic searches of Ovid MEDLINE from 1980 to March 14, 2014, EMBASE from 1980 to 2014 week 10, and Cochrane Central Register of Controlled Trials to January 2014. The search strategy included the Medical Subject Heading terms and text words that identified adult or pediat- ric patients with neoplasms and HSCT recipients combined with identifica- tion of BAL and lung biopsy procedures (the full search strategy can be found in Appendix Table A1 [online only]). A manual search of reference lists of identified articles was also conducted.

Study Selection

Inclusion and exclusion criteria were defined a priori. Studies were in- cluded if patients had cancer or were recipients of HSCT and if they underwent BAL or lung biopsy for the evaluation of a pulmonary lesion. We limited studies to full-text articles published in English after 1980. If a study met any of the following criteria, it was excluded: (1) was not a full-text publication; (2) contained fewer than 10 procedures; (3) was a case-control study (which does not allow calculation of prevalence of diagnoses); (4) had noncancer or non- HSCT patients; (5) had a lung procedure that was conducted for initial diag- nosis of cancer, surveillance (not for the purpose of diagnosis), or evaluation of

Potentially relevant references identified (N = 14,148)

Citations screened by title/abstract (n = 11,932)

Full-text references retrieved for detailed evaluation

(n = 266)

Studies included in meta-analysis (n = 95)

Duplicates removed (n = 2,216)

Articles excluded as did not meet eligibility criteria

(n = 11,666)

)171 = n( dedulcxE Not full-text publication (n = 45) Fewer than 10 procedures (n = 1)

)6 = n( yduts lortnoc esaC Noncancer/HSCT human population (n = 64) Procedure for initial diagnosis of cancer (n = 7) Procedure for surveillance (n = 8) Procedure for drug toxicity (n = 1) Studies focused only on PCP (n = 1) All patients with infection/disease (n = 12) Diagnostic test not validated (n = 5) Did report results of procedures (n = 16) Duplicate publications (n = 5)

Fig 1. Flow diagram illustrating flow of studies identified from the search strategy and reasons for exclusion. HSCT, hematopoietic stem-cell transplanta- tion; PCP, Pneumocystis jirovecii.

Table 1. Characteristics of Included Studies by BAL and Lung Biopsy Procedures

Characteristic and Stratum

BAL (n � 72)

Lung Biopsy (n � 31)

No. of Studies %

Study population age, years Pediatric (younger than 18) 12 17 6 19 Adult (� 18) 37 51 13 42 Both 20 28 10 32

Study population diagnosis Cancer 10 14 5 16 HSCT 32 44 17 55 Both 30 42 9 29

Year of publication 2002 or earlier 38 53 16 52 After 2002 34 47 15 48

Study design Retrospective 52 72 28 90 Prospective 20 28 3 10

EORTC/MSG criteria used� 16 22 1 3 Aspergillus testing

Galatomannan 12 17 NA PCR 10 14 1 3

Brushing used for BAL 14† 19 NA Biopsy type

Transbronchial NA 5 16 Transthoracic NA 26 84

Biopsy-image guided NA 7 23 Patient selection low-risk bias 54 75 24 77 Index test low-risk bias 16 22 2 6

Abbreviations: BAL, bronchoalveolar lavage; EORTC/MSG, European Organi- sation for Research and Treatment of Cancer/Mycoses Study Group; HSCT, hematopoietic stem-cell transplantation; NA, not applicable; PCR, polymerase chain reaction.

�Used criteria from the European Organisation for Research and Treatment of Cancer to define invasive fungal infection. †Brushing done sometimes (n � 12) or always (n � 2).

Chellapandian et al

drug toxicity; (6) focused only on Pneumocystis jirovecii pneumonia (BAL is the standard approach for diagnosing suspected Pneumocystis jirovecii pneu- monia); (7) was a case series reporting only patients with positive diagnostic tests (a case series does not provide data on how many patients underwent the procedure without a positive test); (8) performed for the purpose of validating a diagnostic test; (9) did not report results of procedure; and (10) was a duplicate publication.

Two reviewers (D.C. and L.S.) independently evaluated the titles and abstracts of publications identified by the search strategy, and any potentially relevant publication was retrieved in full. Final inclusion of studies into the systematic review was by agreement of both reviewers. Agreement on study inclusion between the two reviewers was evaluated by using the � statistic, and agreement was defined as slight (0.00 to 0.20), fair (0.21 to 0.40), moderate (0.41 to 0.60), substantial (0.61 to 0.80) or almost perfect (0.81 to 1.00).5

Data Abstraction and Methodologic Approach

Two reviewers (D.C. and L.S.) abstracted all data in duplicate, and any discrepancies were resolved by consensus. The primary outcomes were the proportion of procedures with any diagnosis, infectious diagnoses, and non- infectious diagnoses. In addition, the proportion of procedures with bacterial, viral, and fungal etiologies was described. Fungal infection was defined by each study. The secondary outcomes were complications (defined by each study) and procedure-related mortality. In the evaluation of the diagnostic properties of BAL GM and Aspergillus PCR, the gold standard was proven or probable invasive aspergillosis according to the revised criteria from the European Organisation for Research and Treatment of Cancer/Invasive Fungal Infec- tions Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG).6

Each study described BAL, lung biopsy, or both. Studies were initially separated into those that evaluated and described any infectious or noninfec- tious diagnosis versus those that focused only on a specific diagnosis. For the latter group, we were interested in the diagnostic properties of BAL, GM, and PCR for Aspergillus infection.

Potential factors that could influence the diagnostic and complication rates were age (pediatric [age 18 years or younger], adult, or both), patient population (cancer, HSCT, or both), year of study (dichotomized at 2002), bronchial brushing during BAL (yes or no), and biopsy type (transbronchial or transthoracic). We also evaluated whether patient selection bias and index test bias influenced the infection diagnosis rate or whether patient selection bias influenced the complication rate. For the diagnosis of fungal infection from BAL, we also evaluated whether the study used EORTC/MSG criteria for the diagnosis of fungal infection and whether the study included GM and Asper- gillus PCR for the BAL specimen.

Assessment of Study Quality

Two reviewers assessed study quality, and any discrepancies were re- solved by consensus. Study quality was assessed by using QUADAS-2 for review of diagnostic tests.7 Elements were patient selection (could selection of

patients have introduced bias); index test (could the conduct or interpretation of the test have introduced bias); reference standard (used only for the evalu- ation of BAL GM and PCR for Aspergillus infection; could conduct or inter- pretation of the reference standard have introduced bias); and flow and timing (were all patients included in the analysis). These elements were rated at low, high, and unclear risk of bias.

Statistical Methods

This meta-analysis was performed by using Review Manager (RevMan, version 5.2, Copenhagen, Denmark: The Nordic Cochrane Centre, The Co- chrane Collaboration, 2011). Data were synthesized by using proportions as the measure of effect. Because proportions are not normally distributed, all analyses were conducted by using the natural logarithm of the proportion as the outcome. All estimates are presented as proportions with 95% CIs. The percentage of patients with an outcome may be derived by multiplying the proportion by 100 (eg, 0.53 is synonymous with 53% of patients experiencing the outcome). Heterogeneity was described by using the I2 value, which de- scribes the percentage of variability due to heterogeneity rather than to sam- pling error.8 To explore sources of heterogeneity, we performed stratified analyses by using RevMan; heterogeneity between subgroups was evaluated by using the �2 statistic. Statistical significance was defined as P � .05. RevMan was also used to calculate sensitivity and specificity of BAL GM and Aspergillus PCR. There are statistical concerns about pooling sensitivity and specificity because these values are correlated; thus, the ranges were displayed for each test.9,10 Publication bias was not investigated, given the nature of the outcome, namely proportions.

RESULTS

The flow of study identification and selection is illustrated in Figure 1. There were 14,148 studies identified by the search strategy, of which 266 were retrieved for full evaluation; 95 of those were included in the meta-analysis. Agreement on study inclusion between the two reviewers was almost perfect, with � statistic of 87.9% (95% CI, 82.0% to 93.9%).

Overall, there were 72 studies that evaluated BAL procedures. Characteristics of these studies are provided in Appendix Table A2 (online only). The number of studies at low risk of bias for patient selection was 54 (75%), index test was 16 (22%), and flow and timing was 64 (89%). Of these studies, 53 described the prevalence of any identified organism, whereas 21 focused on a specific organism (two studies described all organisms but also evaluated the diagnostic prop- erties of BAL GM or Aspergillus PCR).11,12 Three studies focused on specific radiologic abnormalities: two included pneumonitis13,14 and one included diffuse infiltrates.15

Table 2. Proportion of Procedures With Specific Diagnosis Stratified by BAL and Lung Biopsy

Diagnosis

BAL (n � 4,893; 6,203 procedures) Lung Biopsy

(n � 976; 1,306 procedures)

P for BAL v BiopsyProportion 95% CI No. of Studies I2 Proportion 95% CI No. of Studies I2

Any 0.53 0.49 to 0.58 46 90 0.54 0.47 to 0.61 29 73 .94 Infectious 0.49 0.45 to 0.54 50 88 0.34 0.28 to 0.42 30 84 � .001 Bacterial 0.19 0.16 to 0.23 49 89 0.06 0.04 to 0.07 28 0 � .001 Viral 0.13 0.09 to 0.18 49 96 0.09 0.06 to 0.14 28 84 .24 Fungal 0.19 0.16 to 0.23 50 86 0.20 0.15 to 0.28 28 86 .71 Noninfectious 0.07 0.05 to 0.09 42 75 0.43 0.35 to 0.52 27 88 � .001 Complications 0.08 0.06 to 0.11 35 82 0.15 0.11 to 0.21 23 77 .006 Results changed management 0.31 0.26 to 0.38 25 92 0.48 0.40 to 0.57 17 82 .002

Abbreviation: BAL, bronchoalveolar lavage.

Lung Procedures in Cancer and Stem-Cell Transplantation

There were 31 studies that evaluated lung biopsies; characteristics of those studies are included in Appendix Table A3 (online only). The number of studies at low risk of bias for patient selection was 24 (77%), index test was two (6%), and flow and timing was 31 (100%). Of those studies, 31 described the prevalence of any organism, whereas one focused on a specific organism. Twenty-six studies evaluated trans-

thoracic biopsy and five evaluated transbronchial biopsy. Seven studies16-22 included both BAL and lung biopsy, and one study16

reported two separate biopsy cohorts (image-guided and open-lung biopsy). Four studies focused on specific radiologic abnormalities: three83-85 included diffuse infiltrates and one86 focused on pulmonary nodules. Characteristics of BAL and lung biopsy studies are

)modnar( W IC %59 noitroporP latoT stnevE ydutS %2.2 55.0 ot 43.0 54.0 29 14 7002 ,nainemrA %2.2 94.0 ot 92.0 93.0 101 93 8002 ,yaluozA %0.2 24.0 ot 22.0 13.0 68 72 1002 ,irA−neB %4.2 37.0 ot 35.0 36.0 101 46 0102 ,noregreB %5.2 18.0 ot 36.0 37.0 59 96 5002 ,regnissiB %7.1 04.0 ot 71.0 72.0 36 71 7002 ,amsreoB %7.1 86.0 ot 92.0 84.0 72 31 7002 ,regruB %6.1 76.0 ot 62.0 64.0 42 11 8002 ,inadroC %3.2 57.0 ot 15.0 46.0 66 24 6891 ,reinnodroC %9.1 75.0 ot 92.0 24.0 25 22 5891 ,reinnodroC %1.2 47.0 ot 24.0 95.0 14 42 7891 ,reinnodroC %1.2 16.0 ot 63.0 84.0 46 13 7991 ,naganuD %2.2 75.0 ot 63.0 74.0 09 24 5002 ,yrrebnekiE %8.1 25.0 ot 32.0 73.0 94 81 8991 ,giwE %3.2 57.0 ot 15.0 36.0 86 34 0102 ,wolsroF %2.2 89.0 ot 55.0 58.0 31 11 8002 ,ayuruF %0.1 25.0 ot 11.0 92.0 12 6 3102 ,sassaG %4.2 76.0 ot 15.0 95.0 541 68 3102 ,trebliG %3.2 36.0 ot 04.0 25.0 97 14 8991 ,rezalG %8.1 95.0 ot 62.0 24.0 83 61 9991 ,nosurG %4.2 94.0 ot 63.0 34.0 322 59 7002 ,atpuG %0.1 26.0 ot 21.0 33.0 51 5 8991 ,yssenneH %0.2 48.0 ot 34.0 56.0 32 51 6991 ,nilrueH %0.2 39.0 ot 84.0 57.0 61 21 9891 ,nilrueH %3.2 77.0 ot 05.0 46.0 35 43 9991 ,lessueH %2.2 15.0 ot 03.0 14.0 19 73 6002 ,retsiemfoH %2.2 44.0 ot 82.0 63.0 531 84 5002 ,lahtnehoH %0.2 63.0 ot 02.0 72.0 421 43 0002 ,agnirauH %5.2 45.0 ot 24.0 84.0 642 811 8002 ,lemmuH %4.2 87.0 ot 65.0 86.0 87 35 7002 ,wosaK %9.1 07.0 ot 33.0 25.0 13 61 6991 ,oninaL %8.1 76.0 ot 92.0 84.0 92 41 2991 ,nibbuCcM %4.2 98.0 ot 26.0 87.0 04 13 7891 ,nrubliM %6.1 74.0 ot 81.0 13.0 54 41 4002 ,civoricebaluM %4.1 75.0 ot 81.0 63.0 52 9 1002 ,yarruM %1.1 18.0 ot 91.0 05.0 01 5 2002 ,nnamueN %4.2 26.0 ot 44.0 45.0 721 86 7991 ,onagaP %6.1 24.0 ot 71.0 82.0 35 51 2002 ,kraP %4.2 65.0 ot 14.0 94.0 961 28 5002 ,letaP %1.2 24.0 ot 52.0 33.0 121 04 8002 ,tabbaR %7.1 67.0 ot 23.0 55.0 22 21 0002 ,alimaR %1.2 07.0 ot 93.0 55.0 44 42 3102 ,oaR %5.2 00.1 ot 77.0 59.0 22 12 8891 ,otiaS %9.1 58.0 ot 14.0 56.0 02 31 0002 ,nenolaS %6.2 95.0 ot 15.0 55.0 895 923 0102 ,nonnahS %0.1 24.0 ot 90.0 22.0 72 6 1002 ,inabuoS %4.1 83.0 ot 31.0 42.0 05 21 9891 ,sekotS %2.2 97.0 ot 94.0 56.0 34 82 2002 ,nedlE nav %4.2 57.0 ot 65.0 66.0 301 86 5991 ,ffiE nov %7.1 04.0 ot 81.0 82.0 86 91 7991 ,etihW

%001 45.0 ot 54.0 94.0 669,3 ledom stceffe modnaR Heterogeneity: I2 = 87.7%, Τ2 = 0.0769, P < .001

0.80.60.40.2

Fig 2. Forest plot of proportion of infec- tious diagnosis among studies. Squares indicate proportion of procedures with an in- fectious diagnosis. Horizontal lines through the squares represent 95% CIs. W, weight.

Chellapandian et al

summarized in Table 1; 86% of BAL studies and 84% of lung biopsy studies included patients undergoing HSCT.

Table 2 provides the main outcomes of the study. The proportion of procedures leading to any diagnosis was similar for BAL and lung biopsy (0.53 v 0.54; P � .94). However, an infectious diagnosis was more common with BAL compared with lung biopsy (0.49 v 0.34; P � .001; Fig 2). In contrast, noninfectious diagnosis was more common with lung biopsy compared with BAL (0.43 v 0.07; P � .001). Change in management occurred more often with lung biopsy compared with BAL (0.48 v 0.31; P � .002). However, complica- tions were more common with lung biopsy compared with BAL (0.15 v 0.08; P � .006). Among the 30 BAL studies that reported procedure-related mortality, there were five BAL-related deaths among 2,792 procedures (0.0018). Among the 20 biopsy studies that reported procedure-related mortality, there were five biopsy- related deaths among 637 procedures (0.0078).

Table 3 provides the results of the stratified analyses that evaluate the proportion of procedures leading to an infectious diagnosis for BAL and lung biopsy. Study characteristics did not explain heterogeneity in the proportion of BAL procedures yield- ing an infectious diagnosis. However, transthoracic lung biopsy was more likely to yield an infectious diagnosis compared with transbronchial biopsy (0.39 v 0.12; P � .001).

There were 50 BAL studies that reported the rate of invasive fungal infection in addition to any infection. There was no difference in the rate of fungal diagnosis among the six studies27-32 that used

EORTC/MSG criteria to define invasive fungal infection (0.19; 95% CI, 0.11 to 0.31) compared with the 44 studies that did not use these criteria (0.19; 95% CI, 0.16 to 0.23; P � .95). Invasive fungal infection detection among the four studies27,32-34 that used BAL GM (0.31; 95% CI, 0.23 to 0.43) was significantly higher compared with the 46 studies that did not use the test (0.18; 95% CI, 0.15 to 0.22; P � .005). Only two studies35,36 used BAL Aspergillus PCR in the setting of evaluation of any infection; invasive fungal infection was diagnosed in 0.24 (95% CI, 0.14 to 0.40).

Table 4 provides the results of the stratified analyses that evalu- ated complications. When stratified by procedure type, the use of brushings during BAL was associated with more complications com- pared with BAL procedures without brushings (P � .03). Among lung biopsy procedures, children were more likely to experience complica- tions compared with adults or mixed-age populations (P � .003). Transthoracic procedures were also more commonly associated with complications compared with transbronchial procedures (0.18 v 0.05; P � .005). When stratified by study characteristics, lung biopsy, when compared with BAL, was more commonly associated with complica- tions in children (0.37 v 0.08; P � .001).

For BAL studies that evaluated a specific organism, the most common organism was Aspergillus in 13 studies11,12,32,37-46; GM only was evaluated in seven studies,32,37-42 Aspergillus PCR only was evaluated in four studies,43-46 and both were evaluated in two studies.11,12 Diagnostic properties were evaluated in seven GM and four PCR studies. Among these 11 studies, the number at low risk

Table 3. Proportion of Procedures With Infection Diagnosis Stratified by Study Characteristics

Characteristic

BAL Lung Biopsy

P for BAL v Biopsy No. of

Procedures Infection Diagnosis 95% CI P

No. of Procedures

Infection Diagnosis 95% CI P

Age group Children 542 0.45 0.37 to 0.56 .10 155 0.32 0.21 to 0.49 .02 .15 Adult 2,238 0.48 0.42 to 0.55 651 0.24 0.15 to 0.36 .002 Both 1,062 0.56 0.50 to 0.63 402 0.45 0.37 to 0.56 .08

Underlying diagnosis Cancer 393 0.49 0.38 to 0.64 .97 263 0.46 0.23 to 0.90 .47 .84 HSCT 2,309 0.50 0.45 to 0.55 645 0.30 0.23 to 0.40 � .001 Both 1,264 0.48 0.40 to 0.58 364 0.36 0.26 to 0.50 .13

Year published 2002 or earlier 1,376 0.49 0.42 to 0.57 .97 580 0.41 0.34 to 0.50 .06 .16 After 2002 2,590 0.49 0.44 to 0.54 692 0.28 0.19 to 0.40 .002

Study design Retrospective 3,455 0.50 0.45 to 0.55 .50 1,237 0.33 0.27 to 0.41 .02 � .001 Prospective 511 0.46 0.36 to 0.58 35 0.52 0.38 to 0.72 .51

Brushing used for BAL Yes 799 0.52 0.42 to 0.63 .56 NA NA NA NA No 3,167 0.48 0.44 to 0.53 NA

Biopsy type Transbronchial NA NA NA 211 0.12 0.06 to 0.24 .001 NA Transthoracic 1,061 0.39 0.32 to 0.47 NA

Patient selection bias Low risk 3,135 0.50 0.46 to 0.55 .46 1,130 0.33 0.27 to 0.42 .36 � .001 Not low risk 831 0.46 0.35 to 0.59 142 0.39 0.30 to 0.52 .45

Index test bias Low risk 275 0.43 0.29 to 0.65 .51 76 0.31 0.12 to 0.76 .79 .50 Not low risk 3,691 0.50 0.45 to 0.54 1,196 0.35 0.28 to 0.42 .001

Abbreviations: BAL, bronchoalveolar lavage; HSCT, hematopoietic stem-cell transplantation; NA, not applicable.

Lung Procedures in Cancer and Stem-Cell Transplantation

of reference standard bias was five (45%). Against the EORTC/ MSG gold standard, the sensitivity of BAL GM ranged from 0.58 (95% CI, 0.39 to 0.75) to 1.00 (95% CI, 0.85 to 1.00), and the specificity ranged from 0.78 (95% CI, 0.71 to 0.84) to 1.00 (95% CI, 0.85 to 1.00). The sensitivity of Aspergillus PCR ranged from 0.69 (95% CI, 0.50 to 0.84) to 0.94 (95% CI, 0.80 to 0.99) and the specificity ranged from 0.75 (95% CI, 0.43 to 0.95) to 0.94 (95% CI, 0.88 to 0.98).

DISCUSSION

In this study, which focuses on lung procedures for the diagnosis of infectious and noninfectious causes of pulmonary infiltrates among patients with cancer and HSCT recipients, we found that BAL resulted in a diagnosis in 0.53 procedures and lung biopsy resulted in a diag- nosis in 0.54 procedures. An infectious diagnosis was made more frequently for BAL (0.49 v 0.34) whereas a noninfectious diagnosis was made more frequently for lung biopsy (0.43 v 0.07). However, the rate of complications was almost twice as high for lung biopsies compared with BAL (0.15 v 0.08), and children were at higher risk of complica- tions. Procedure-related mortality was four-fold higher for lung bi- opsy compared with BAL (0.0078 v 0.0018).

These results suggest that BAL may be the preferred initial diagnostic modality for evaluating potentially infectious pulmo- nary lesions among patients with cancer and recipients of HSCT. BAL procedures were at lower risk of complications and procedure-related deaths with similar yield compared with trans- thoracic biopsies. Furthermore, with the advent of improvements

in diagnostic capabilities such as GM and PCR, the diagnostic yield of BAL should improve. If invasive fungal infection is suspected, we suggest that BAL GM should be included in the evaluation because BAL GM significantly improves the detection of fungi and has adequate diagnostic properties in this population.

The procedure-related mortality rate of lung biopsy is not negli- gible. This is not surprising, given the nature of these patients, many of whom are thrombocytopenic. This review suggests that lung biopsies are more likely to be useful when noninfectious diagnoses are sus- pected. Indeed, a BAL is by design a diagnostic test focused on captur- ing infectious complications, whereas lung biopsy is a more comprehensive approach and can obtain much needed information regarding histopathologic changes, including the presence of malig- nancy. Lung biopsies did lead to a change in management in almost half the patients. However, the risk of complications and procedure- related deaths should be considered when deciding whether the pro- cedure is worthwhile as an initial diagnostic procedure. Transthoracic biopsies were more likely to yield a diagnosis compared with trans- bronchial biopsies and may be particularly useful when a noninfec- tious diagnosis is suspected.

Our interpretation that BAL may be the preferred initial diagnostic modality in the setting of a potentially infectious pathology is consistent with a review among immunocompromised children in general, which also recommended BAL as the initial diagnostic tool.90 Our recommen- dation to include BAL GM in the evaluation of potential pulmonary fungus is also consistent with guideline statements from other groups.1,91

What is striking from our review is the percentage of times that any diagnosis, approximately 50% using either BAL or lung biopsy,

Table 4. Proportion of Procedures With Complications Stratified by Study Characteristics

Characteristic

BAL Lung Biopsy

P BAL v Biopsy No. of

Procedures Complications 95% CI P No. of

Procedures Complications 95% CI P

Age group Children 408 0.08 0.04 to 0.17 .10 115 0.37 0.22 to 0.62 .003 .001 Adult 2102 0.11 0.07 to 0.16 621 0.12 0.07 to 0.21 .72 Both 784 0.05 0.02 to 0.09 246 0.13 0.08 to 0.21 .008

Underlying diagnosis Cancer 346 0.03 0.00 to 0.22 .64 297 0.11 0.04 to 0.34 .88 .27 HSCT 1926 0.08 0.05 to 0.13 356 0.15 0.09 to 0.25 .11 Both 1022 0.08 0.05 to 0.14 352 0.16 0.09 to 0.27 .10

Year 2002 or earlier 1159 0.04 0.02 to 0.08 .008 334 0.15 0.10 to 0.25 .91 � .001 After 2002 2135 0.11 0.08 to 0.16 671 0.15 0.09 to 0.24 .33

Study design Retrospective 2714 0.07 0.05 to 0.10 .22 948 0.16 0.11 to 0.22 .37 .002 Prospective 580 0.11 0.06 to 0.19 57 0.11 0.05 to 0.23 .99

Brushing used for BAL Yes 929 0.04 0.02 to 0.09 .03 NA NA NA NA No 2365 0.10 0.07 to 0.14 NA

Biopsy type Transbronchial NA NA NA 147 0.05 0.02 to 0.11 .005 NA Transthoracic 858 0.18 0.13 to 0.24 NA

Patient selection bias Low risk 2468 0.08 0.05 to 0.11 .98 908 0.16 0.11 to 0.23 .66 .006 Not low risk 826 0.08 0.04 to 0.17 97 0.14 0.08 to 0.24 .24

Abbreviations: BAL, bronchoalveolar lavage; HSCT, hematopoietic stem-cell transplantation; NA, not applicable.

Chellapandian et al

can be made when radiographic lung nodules or other findings are directly pursued via either approach. Irrespective of the initial diagnostic approach, we suggest that investigation be conducted expeditiously because treatment may influence diagnostic yield.92

If a BAL is the initial evaluation and is negative or yields a pathogen of questionable significance, it is then important to consider whether transthoracic biopsy results may alter patient manage- ment. It is also important to stress that our results were heteroge- neous and thus, yield and complication rates of BAL and biopsies may vary at different institutions on the basis of operator expertise and indications for the procedure.

The strength of this systematic review is the pooling of clinically meaningful end points and restriction to patients with cancer and HSCT. However, there are several limitations. First, we relied on the definition of diagnostic etiologies provided in individual studies. For example, BAL is not a sterile procedure, and some studies identified viruses and bacteria which may not have been causally related to the pulmonary process. Similarly, biopsy studies sometimes focused only on specific diagnoses such as invasive fungal infection and malig- nancy. However, some biopsy studies included entities such as diffuse alveolar damage and fibrosis. Without access to the primary data, we were not able to further classify the different types of lung biopsy diagnoses, although change in management based on the lung biopsy is probably more meaningful to clinicians. Second, studies were con- ducted over a long period of time, although our stratified analyses did not suggest that the year of publication had an impact on the propor- tion of procedures with an infectious diagnosis or complication. Third, our results are confounded by indication for the procedure. In addition, there were too few studies that stratified results by specific radiologic patterns to permit analysis, and thus, we are uncertain whether the specific radiologic patterns should influence the choice of procedure. Fourth, there is the potential for reporting bias, both at the study and the outcome level because no registry of

such studies exists, and study protocols are not available for inspec- tion. Furthermore, it was not possible to undertake mathematical estimates of publication bias because, to the best of our knowledge, robust methods for evaluating publication bias are not available when the outcome is a proportion.

Future studies should focus on the development of clinical practice guidelines and algorithms for the evaluation of patients with lung infiltrates in the oncology and HSCT setting. In addition, the use of biomarkers should be further explored. In conclusion, BAL may be the preferred initial diagnostic modality for the eval- uation of potentially infectious pulmonary lesions among patients with cancer and HSCT recipients. Guidelines for promoting con- sistency in the approach to the evaluation of lung infiltrates may improve clinical care of patients.

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Disclosures provided by the authors are available with this article at www.jco.org.

AUTHOR CONTRIBUTIONS

Conception and design: DeepakBabu Chellapandian, Thomas Lehrnbecher, Lillian Sung Financial support: Lillian Sung Collection and assembly of data: DeepakBabu Chellapandian, Lillian Sung Data analysis and interpretation: Thomas Lehrnbecher, Bob Phillips, Brian T. Fisher, Theoklis E. Zaoutis, William J. Steinbach, Joseph Beyene, Lillian Sung Manuscript writing: All authors Final approval of manuscript: All authors

REFERENCES

1. Lehrnbecher T, Phillips R, Alexander S, et al: Guideline for the management of fever and neutro- penia in children with cancer and/or undergoing hematopoietic stem-cell transplantation. J Clin On- col 30:4427-4438, 2012

2. Sung L, Lange BJ, Gerbing RB, et al: Micro- biologically documented infections and infection- related mortality in children with acute myeloid leukemia. Blood 110:3532-3539, 2007

3. Nosari AM, Caira M, Pioltelli ML, et al: Hema e-Chart registry of invasive fungal infections in haematological patients: Improved outcome in re- cent years in mould infections. Clin Microbiol Infect 19:757-762, 2013

4. Moher D, Liberati A, Tetzlaff J, et al: Pre- ferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 6:e1000097, 2009

5. Landis JR, Koch GG: The measurement of observer agreement for categorical data. Biometrics 33:159-174, 1977

6. De Pauw B, Walsh TJ, Donnelly JP, et al: Revised definitions of invasive fungal disease from the European Organization for Research and Treat- ment of Cancer/Invasive Fungal Infections Cooper- ative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/

MSG) Consensus Group. Clin Infect Dis 46:1813- 1821, 2008

7. Whiting PF, Rutjes AW, Westwood ME, et al: QUADAS-2: A revised tool for the quality assess- ment of diagnostic accuracy studies. Ann Intern Med 155:529-536, 2011

8. Higgins JP, Green S (eds): Cochrane Hand- book for Systematic Reviews of Interventions Ver- sion 5.1.0 (updated March 2011). The Cochrane Collaboration. www.cochrane-handbook.org

9. Shapiro DE: Issues in combining indepen- dent estimates of the sensitivity and specificity of a diagnostic test. Acad Radiol 1:S37-S47, 1995

10. Reitsma JB, Glas AS, Rutjes AW, et al: Bivariate analysis of sensitivity and specificity produces informative summary measures in diag- nostic reviews. J Clin Epidemiol 58:982-990, 2005

11. Verweij PE, Latgé JP, Rijs AJ, et al: Compar- ison of antigen detection and PCR assay using bronchoalveolar lavage fluid for diagnosing invasive pulmonary aspergillosis in patients receiving treat- ment for hematological malignancies. J Clin Micro- biol 33:3150-3153, 1995

12. Reinwald M, Spiess B, Heinz WJ, et al: Diagnosing pulmonary aspergillosis in patients with hematological malignancies: A multicenter prospec- tive evaluation of an Aspergillus PCR assay and a galactomannan ELISA in bronchoalveolar lavage samples. Eur J Haematol 89:120-127, 2012

13. Cathomas G, Morris P, Pekle K, et al: Rapid diagnosis of cytomegalovirus pneumonia in marrow transplant recipients by bronchoalveolar lavage us- ing the polymerase chain reaction, virus culture, and the direct immunostaining of alveolar cells. Blood 81:1909-1914, 1993

14. Cordonnier C, Bernaudin JF, Bierling P, et al: Pulmonary complications occurring after allogeneic bone marrow transplantation: A study of 130 con- secutive transplanted patients. Cancer 58:1047- 1054, 1986

15. Gupta S, Jain A, Warneke CL, et al: Outcome of alveolar hemorrhage in hematopoietic stem cell transplant recipients. Bone Marrow Transplant 40: 71-78, 2007

16. Armenian SH, La Via WV, Siegel SE, et al: Evaluation of persistent pulmonary infiltrates in pe- diatric oncology patients. Pediatr Blood Cancer 48: 165-172, 2007

17. Gassas A, Craig-Barnes H, Dell SD, et al: Severe lung injury and lung biopsy in children post-hematopoietic stem cell transplantation: The differences between allo- geneic and autologous transplantation. Pediatr Transplant 17:278-284, 2013

18. Gilbert CR, Lerner A, Baram M, et al: Utility of flexible bronchoscopy in the evaluation of pulmo- nary infiltrates in the hematopoietic stem cell trans- plant population: A single center fourteen year experience. Arch Bronconeumol 49:189-195, 2013

Lung Procedures in Cancer and Stem-Cell Transplantation

http://www.jco.org

http://www.cochrane-handbook.org

19. Mulabecirovic A, Gaulhofer P, Auner HW, et al: Pulmonary infiltrates in patients with haemato- logic malignancies: Transbronchial lung biopsy in- creases the diagnostic yield with respect to neoplastic infiltrates and toxic pneumonitis. Ann Hematol 83:420-422, 2004

20. Patel NR, Lee PS, Kim JH, et al: The influ- ence of diagnostic bronchoscopy on clinical out- comes comparing adult autologous and allogeneic bone marrow transplant patients. Chest 127:1388- 1396, 2005

21. Soubani AO, Qureshi MA, Baynes RD, et al: Flexible bronchoscopy in the diagnosis of pulmonary infiltrates following autologous peripheral stem cell transplantation for advanced breast cancer. Bone Marrow Transplant 28:981-985, 2001

22. White P, Bonacum JT, Miller CB: Utility of fiberoptic bronchoscopy in bone marrow transplant patients. Bone Marrow Transplant 20:681-687, 1997

23. Boersma WG, Erjavec Z, van der Werf TS, et al: Bronchoscopic diagnosis of pulmonary infiltrates in granulocytopenic patients with hematologic ma- lignancies: BAL versus PSB and PBAL. Respir Med 101:317-325, 2007

24. Kasow KA, King E, Rochester R, et al: Diagnostic yield of bronchoalveolar lavage is low in allogeneic hema- topoietic stem cell recipients receiving immunosuppres- sive therapy or with acute graft-versus-host disease: The St. Jude experience, 1990-2002. Biol Blood Marrow Transplant 13:831-837, 2007

25. Rabbat A, Chaoui D, Lefebvre A, et al: Is BAL useful in patients with acute myeloid leukemia ad- mitted in ICU for severe respiratory complications? Leukemia 22:1361-1367, 2008

26. Furuya ME, González-Martínez F, Vargas MH, et al: Guidelines for diagnosing and treating pulmonary infiltrates in children with acute leukae- mia: Impact of timely decisions. Acta Paediatr 97: 928-934, 2008

27. Hohenthal U, Itälä M, Salonen J, et al: Bron- choalveolar lavage in immunocompromised patients with haematological malignancy: Value of new mi- crobiological methods. Eur J Haematol 74:203-211, 2005

28. Shannon VR, Andersson BS, Lei X, et al: Utility of early versus late fiberoptic bronchoscopy in the evaluation of new pulmonary infiltrates following hematopoietic stem cell transplantation. Bone Mar- row Transplant 45:647-655, 2010

29. Forslöw U, Remberger M, Nordlander A, et al: The clinical importance of bronchoalveolar lavage in allogeneic SCT patients with pneumonia. Bone Marrow Transplant 45:945-950, 2010

30. Hofmeister CC, Czerlanis C, Forsythe S, et al: Retrospective utility of bronchoscopy after hema- topoietic stem cell transplant. Bone Marrow Trans- plant 38:693-698, 2006

31. Azoulay E, Mokart D, Rabbat A, et al: Diag- nostic bronchoscopy in hematology and oncology patients with acute respiratory failure: Prospective multicenter data. Crit Care Med 36:100-107, 2008

32. Bergeron A, Belle A, Sulahian A, et al: Con- tribution of galactomannan antigen detection in BAL to the diagnosis of invasive pulmonary aspergillosis in patients with hematologic malignancies. Chest 137:410-415, 2010

33. Cordani S, Manna A, Vignali M, et al: Bron- choalveolar lavage as a diagnostic tool in patients with hematological malignancies and pneumonia. Infez Med 16:209-213, 2008

34. Salonen J, Lehtonen OP, Teräsjärvi MR, et al: Aspergillus antigen in serum, urine and bron- choalveolar lavage specimens of neutropenic pa-

tients in relation to clinical outcome. Scand J Infect Dis 32:485-490, 2000

35. Bissinger AL, Einsele H, Hamprecht K, et al: Infectious pulmonary complications after stem cell transplantation or chemotherapy: Diagnostic yield of bronchoalveolar lavage. Diagn Microbiol Infect Dis 52:275-280, 2005

36. Hummel M, Rudert S, Hof H, et al: Diagnostic yield of bronchoscopy with bronchoalveolar lavage in febrile patients with hematologic malignancies and pulmonary infiltrates. Ann Hematol 87:291-297, 2008

37. Becker MJ, Lugtenburg EJ, Cornelissen JJ, et al: Galactomannan detection in computerized tomography-based broncho-alveolar lavage fluid and serum in haematological patients at risk for invasive pulmonary aspergillosis. Br J Haematol 121:448- 457, 2003

38. Fréalle E, Decrucq K, Botterel F, et al: Diag- nosis of invasive aspergillosis using bronchoalveolar lavage in haematology patients: Influence of bron- choalveolar lavage human DNA content on real-time PCR performance. Eur J Clin Microbiol Infect Dis 28:223-232, 2009

39. Luong ML, Filion C, Labbé AC, et al: Clinical utility and prognostic value of bronchoalveolar la- vage galactomannan in patients with hematologic malignancies. Diagn Microbiol Infect Dis 68:132- 139, 2010

40. Nguyen MH, Leather H, Clancy CJ, et al: Galactomannan testing in bronchoalveolar lavage fluid facilitates the diagnosis of invasive pulmonary aspergillosis in patients with hematologic malignan- cies and stem cell transplant recipients. Biol Blood Marrow Transplant 17:1043-1050, 2011

41. Penack O, Rempf P, Graf B, et al: Aspergillus galactomannan testing in patients with long-term neutropenia: Implications for clinical management. Ann Oncol 19:984-989, 2008

42. Racil Z, Kocmanova I, Toskova M, et al: Galactomannan detection in bronchoalveolar lavage fluid for the diagnosis of invasive aspergillosis in patients with hematological diseases: The role of factors affecting assay performance. Int J Infect Dis 15:e874-e881, 2011

43. Buchheidt D, Baust C, Skladny H, et al: Clinical evaluation of a polymerase chain reaction assay to detect Aspergillus species in bronchoalveo- lar lavage samples of neutropenic patients. Br J Haematol 116:803-811, 2002

44. Jones ME, Fox AJ, Barnes AJ, et al: PCR- ELISA for the early diagnosis of invasive pulmonary aspergillus infection in neutropenic patients. J Clin Pathol 51:652-656, 1998

45. Khot PD, Ko DL, Hackman RC, et al: Devel- opment and optimization of quantitative PCR for the diagnosis of invasive aspergillosis with bronchoal- veolar lavage fluid. BMC Infect Dis 8:73, 2008

46. Raad I, Hanna H, Huaringa A, et al: Diagnosis of invasive pulmonary aspergillosis using polymer- ase chain reaction-based detection of aspergillus in BAL. Chest 121:1171-1176, 2002

47. Englund JA, Piedra PA, Jewell A, et al: Rapid diagnosis of respiratory syncytial virus infections in immunocompromised adults. J Clin Microbiol 34: 1649-1653, 1996

48. Kuehnhardt D, Hannemann M, Schmidt B, et al: Therapeutic implication of BAL in patients with neutropenia. Ann Hematol 88:1249-1256, 2009

49. Azoulay E, Mokart D, Lambert J, et al: Diag- nostic strategy for hematology and oncology pa- tients with acute respiratory failure: Randomized controlled trial. Am J Respir Crit Care Med 182: 1038-1046, 2010

50. Rao U, Piccin A, Malone A, et al: Utility of bronchoalveolar lavage in the diagnosis of pulmo- nary infection in children with haematological malig- nancies. Ir J Med Sci 182:177-183, 2013

51. Cordonnier C, Bernaudin JF, Fleury J, et al: Diagnostic yield of bronchoalveolar lavage in pneu- monitis occurring after allogeneic bone marrow transplantation. Am Rev Respir Dis 132:1118-1123, 1985

52. Cordonnier C, Escudier E, Nicolas JC, et al: Evaluation of three assays on alveolar lavage fluid in the diagnosis of cytomegalovirus pneumonitis after bone marrow transplantation. J Infect Dis 155:495- 500, 1987

53. Milburn HJ, Prentice HG, du Bois RM: Role of bronchoalveolar lavage in the evaluation of inter- stitial pneumonitis in recipients of bone marrow transplants. Thorax 42:766-772, 1987

54. Saito H, Anaissie EJ, Morice RC, et al: Bron- choalveolar lavage in the diagnosis of pulmonary infiltrates in patients with acute leukemia. Chest 94:745-749, 1988

55. Crawford SW, Bowden RA, Hackman RC, et al: Rapid detection of cytomegalovirus pulmonary infection by bronchoalveolar lavage and centrifuga- tion culture. Ann Intern Med 108:180-185, 1988

56. Heurlin N, Lönnqvist B, Tollemar J, et al: Fiberoptic bronchoscopy for diagnosis of opportu- nistic pulmonary infections after bone marrow trans- plantation. Scand J Infect Dis 21:359-366, 1989

57. Gleaves CA, Meyers JD: Rapid detection of cytomegalovirus in bronchoalveolar lavage speci- mens from marrow transplant patients: Evaluation of a direct fluorescein-conjugated monoclonal anti- body reagent. J Virol Methods 26:345-349, 1989

58. Stokes DC, Shenep JL, Parham D, et al: Role of flexible bronchoscopy in the diagnosis of pulmonary infiltrates in pediatric patients with cancer. J Pediatr 115:561-567, 1989

59. Levy H, Horak DA, Tegtmeier BR, et al: The value of bronchoalveolar lavage and bronchial wash- ings in the diagnosis of invasive pulmonary aspergil- losis. Respir Med 86:243-248, 1992

60. McCubbin MM, Trigg ME, Hendricker CM, et al: Bronchoscopy with bronchoalveolar lavage in the evaluation of pulmonary complications of bone marrow transplantation in children. Pediatr Pulmonol 12:43-47, 1992

61. Weiss SM, Hert RC, Gianola FJ, et al: Com- plications of fiberoptic bronchoscopy in thrombocy- topenic patients. Chest 104:1025-1028, 1993

62. von Eiff M, Zühlsdorf M, Roos N, et al: Pulmonary infiltrates in patients with haematologic malignancies: Clinical usefulness of non-invasive bronchoscopic procedures. Eur J Haematol 54:157- 162, 1995

63. Lanino E, Sacco O, Kotitsa Z, et al: Fiberoptic bronchoscopy and bronchoalveolar lavage for the evaluation of pulmonary infiltrates after BMT in children. Bone Marrow Transplant 18:117-120, 1996

64. Heurlin N, Bergström SE, Winiarski J, et al: Fungal pneumonia: The predominant lung infection causing death in children undergoing bone marrow transplantation. Acta Paediatr 85:168-172, 1996

65. Dunagan DP, Baker AM, Hurd DD, et al: Bronchoscopic evaluation of pulmonary infiltrates following bone marrow transplantation. Chest 111: 135-141, 1997

66. Pagano L, Pagliari G, Basso A, et al: The role of bronchoalveolar lavage in the microbiological di- agnosis of pneumonia in patients with haematologi- cal malignancies. Ann Med 29:535-540, 1997

Chellapandian et al

67. Hennessy BJ, Dowd N, Crotty GM, et al: Respiratory failure and intensive care support fol- lowing bone marrow transplantation. Hematology 3:315-319, 1998

68. Glazer M, Breuer R, Berkman N, et al: Use of fiberoptic bronchoscopy in bone marrow transplant recipients. Acta Haematol 99:22-26, 1998

69. Ewig S, Torres A, Riquelme R, et al: Pulmo- nary complications in patients with haematological malignancies treated at a respiratory ICU. Eur Respir J 12:116-122, 1998

70. Gruson D, Hilbert G, Portel L, et al: Severe respiratory failure requiring ICU admission in bone marrow transplant recipients. Eur Respir J 13:883- 887, 1999

71. Heussel CP, Kauczor HU, Heussel GE, et al: Pneumonia in febrile neutropenic patients and in bone marrow and blood stem-cell transplant recipi- ents: Use of high-resolution computed tomography. J Clin Oncol 17:796-805, 1999

72. Rámila E, Sureda A, Martino R, et al: Bron- choscopy guided by high-resolution computed to- mography for the diagnosis of pulmonary infections in patients with hematologic malignancies and nor- mal plain chest X-ray. Haematologica 85:961-966, 2000

73. Huaringa AJ, Leyva FJ, Signes-Costa J, et al: Bronchoalveolar lavage in the diagnosis of pulmo- nary complications of bone marrow transplant pa- tients. Bone Marrow Transplant 25:975-979, 2000

74. Murray PV, O’Brien ME, Padhani AR, et al: Use of first line bronchoalveolar lavage in the immu- nosuppressed oncology patient. Bone Marrow Transplant 27:967-971, 2001

75. Ben-Ari J, Yaniv I, Nahum E, et al: Yield of bronchoalveolar lavage in ventilated and non- ventilated children after bone marrow transplanta- tion. Bone Marrow Transplant 27:191-194, 2001

76. van Elden LJ, van Kraaij MG, Nijhuis M, et al: Polymerase chain reaction is more sensitive than viral culture and antigen testing for the detection of respiratory viruses in adults with hematological can- cer and pneumonia. Clin Infect Dis 34:177-183, 2002

77. Park JR, Fogarty S, Brogan TV: Clinical utility of bronchoalveolar lavage in pediatric cancer pa- tients. Med Pediatr Oncol 39:175-180, 2002

78. Neumann M, von Bredow C, Ratjen F, et al: Bronchoalveolar lavage protein patterns in children with malignancies, immunosuppression, fever and pulmonary infiltrates. Proteomics 2:683-689, 2002

79. Ison MG, Hayden FG, Kaiser L, et al: Rhino- virus infections in hematopoietic stem cell trans- plant recipients with pneumonia. Clin Infect Dis 36:1139-1143, 2003

80. Martino R, Porras RP, Rabella N, et al: Pro- spective study of the incidence, clinical features, and outcome of symptomatic upper and lower re- spiratory tract infections by respiratory viruses in adult recipients of hematopoietic stem cell trans- plants for hematologic malignancies. Biol Blood Marrow Transplant 11:781-796, 2005

81. Eikenberry M, Bartakova H, Defor T, et al: Natural history of pulmonary complications in chil- dren after bone marrow transplantation. Biol Blood Marrow Transplant 11:56-64, 2005

82. Burger CD: Utility of positive bronchoalveolar lavage in predicting respiratory failure after hemato- poietic stem cell transplantation: A retrospective analysis. Transplant Proc 39:1623-1625, 2007

83. Crawford SW, Hackman RC, Clark JG: Open lung biopsy diagnosis of diffuse pulmonary infil- trates after marrow transplantation. Chest 94:949- 953, 1988

84. Springmeyer SC, Silvestri RC, Sale GE, et al: The role of transbronchial biopsy for the diagnosis of diffuse pneumonias in immunocompromised mar- row transplant recipients. Am Rev Respir Dis 126: 763-765, 1982

85. Wang JY, Chang YL, Lee LN, et al: Diffuse pulmonary infiltrates after bone marrow transplanta- tion: The role of open lung biopsy. Ann Thorac Surg 78:267-272, 2004

86. Gulbahce HE, Pambuccian SE, Jessurun J, et al: Pulmonary nodular lesions in bone marrow transplant recipients: Impact of histologic diagnosis on patient management and prognosis. Am J Clin Pathol 121:205-210, 2004

87. Kallenberg MH, Gill RR, Factor RE, et al: Diagnostic efficacy and safety of computed tomography-guided transthoracic needle biopsy in patients with hematologic malignancies. Acad Ra- diol 16:1408-1415, 2009

88. Gupta S, Sultenfuss M, Romaguera JE, et al: CT-guided percutaneous lung biopsies in patients with haematologic malignancies and undiagnosed pulmonary lesions. Hematol Oncol 28:75-81, 2010

89. Sharma SK, Kumar S, Singh AK, et al: Feasi- bility and outcome of CT-guided lung biopsy in patients with hematological diseases and suspected fungal pneumonia. J Infect Dev Ctries 7:748-752, 2013

90. Efrati O, Sadeh-Gornik U, Modan-Moses D, et al: Flexible bronchoscopy and bronchoalveolar lavage in pediatric patients with lung disease. Pedi- atr Crit Care Med 10:80-84, 2009

91. Marchetti O, Lamoth F, Mikulska M, et al: ECIL recommendations for the use of biological markers for the diagnosis of invasive fungal diseases in leukemic patients and hematopoietic SCT recipients. Bone Mar- row Transplant 47:846-854, 2012

92. Reinwald M, Hummel M, Kovalevskaya E, et al: Therapy with antifungals decreases the diagnos- tic performance of PCR for diagnosing invasive aspergillosis in bronchoalveolar lavage samples of patients with haematological malignancies. J Anti- microb Chemother 67:2260-2267, 2012

93. Shulman HM, Hackman RC, Sale GE, et al: Rapid cytologic diagnosis of cytomegalovirus inter- stitial pneumonia on touch imprints from open-lung biopsy. Am J Clin Pathol 77:90-94, 1982

94. McCabe RE, Brooks RG, Mark JB, et al: Open lung biopsy in patients with acute leukemia. Am J Med 78:609-616, 1985

95. Hall TS, Hutchins GM, Baker RR: A critical review of the use of open lung biopsy in the management of the oncologic patient with acute pulmonary infiltrates. Am J Clin Oncol 10:249-252, 1987

96. Shorter NA, Ross AJ 3rd, August C, et al: The usefulness of open-lung biopsy in the pediatric bone marrow transplant population. J Pediatr Surg 23: 533-537, 1988

97. Snyder CL, Ramsay NK, McGlave PB, et al: Diagnostic open-lung biopsy after bone marrow transplantation. J Pediatr Surg 25:871-876, 1990

98. Won HJ, Lee KS, Cheon JE, et al: Invasive pulmonary aspergillosis: Prediction at thin-section CT in patients with neutropenia—A prospective study. Radiology 208:777-782, 1998

99. White DA, Wong PW, Downey R: The utility of open lung biopsy in patients with hematologic malignancies. Am J Respir Crit Care Med 161:723- 729, 2000

100. Dunn JC, West KW, Rescorla FJ, et al: The utility of lung biopsy in recipients of stem cell transplantation. J Pediatr Surg 36:1302-1303, 2001

101. Hayes-Jordan A, Benaim E, Richardson S, et al: Open lung biopsy in pediatric bone marrow transplant patients. J Pediatr Surg 37:446-452, 2002

102. Kim K, Lee MH, Kim J, et al: Importance of open lung biopsy in the diagnosis of invasive pulmo- nary aspergillosis in patients with hematologic ma- lignancies. Am J Hematol 71:75-79, 2002

103. Shaikh ZH, Torres HA, Walsh GL, et al: Open lung biopsy in bone marrow transplant recipients has a poor diagnostic yield for a specific diagnosis. Transpl Infect Dis 4:80-84, 2002

104. Jantunen E, Piilonen A, Volin L, et al: Radio- logically guided fine needle lung biopsies in the evaluation of focal pulmonary lesions in allogeneic stem cell transplant recipients. Bone Marrow Trans- plant 29:353-356, 2002

105. Nosari A, Anghilieri M, Carrafiello G, et al: Utility of percutaneous lung biopsy for diagnosing filamentous fungal infections in hematologic malig- nancies. Haematologica 88:1405-1409, 2003

106. Zihlif M, Khanchandani G, Ahmed HP, et al: Surgical lung biopsy in patients with hematological malignancy or hematopoietic stem cell transplanta- tion and unexplained pulmonary infiltrates: Im- proved outcome with specific diagnosis. Am J Hematol 78:94-99, 2005

107. Carrafiello G, Laganà D, Nosari AM, et al: Utility of computed tomography (CT) and of fine needle aspiration biopsy (FNAB) in early diagnosis of fungal pulmonary infections. Study of infections from filamentous fungi in haematologically immuno- deficient patients. Radiol Med 111:33-41, 2006

108. Yang TM, Wang PN, Kao KC, et al: Out- come of hematopoietic stem cell recipients who were mechanically ventilated and admitted to intensive care units. J Formos Med Assoc 106: 295-301, 2007

■ ■ ■

Lung Procedures in Cancer and Stem-Cell Transplantation

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Bronchoalveolar Lavage and Lung Biopsy in Patients With Cancer and Hematopoietic Stem-Cell Transplantation Recipients: A Systematic Review and Meta-Analysis

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I � Immediate Family Member, Inst � My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO’s conflict of interest policy, please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.

DeepakBabu Chellapandian No relationship to disclose

Thomas Lehrnbecher Honoraria: Gilead Sciences, Astellas Pharma, Pfizer, MSD, Merck, Bristol-Myers Squibb Consulting or Advisory Role: Gilead Sciences, MSD, Merck Speakers’ Bureau: Gilead Sciences, Merck, MSD, Astellas Pharma, Pfizer Research Funding: Gilead Sciences Travel, Accommodations, Expenses: Gilead Sciences, Astellas Pharma, Merck, MSD, Pfizer, Baxter International

Bob Phillips No relationship to disclose

Brian T. Fisher Research Funding: Pfizer (Inst), Enzon Pharmaceuticals (Inst), Wyeth (Inst)

Theoklis E. Zaoutis Consulting or Advisory Role: Merck, Pfizer, Cubist Pharmaceuticals Research Funding: Merck (Inst), Cubist Pharmaceuticals (Inst)

William J. Steinbach Consulting or Advisory Role: Merck, Astellas Pharma Research Funding: Merck, Astellas Pharma

Joseph Beyene No relationship to disclose

Lillian Sung No relationship to disclose

Chellapandian et al

http://www.asco.org/rwc

http://jco.ascopubs.org/site/ifc

Acknowledgment

We thank Elizabeth Uleryk for her support in conducting the literature search and Amanda Celis and Cathy O’Sullivan for their administrative assistance.

Appendix

Table A1. Search Strategies

Database Inclusive Dates Search Set History

MEDLINE 1946 to March 13, 2014 1 43 exp �neoplasms/or exp �Stem Cell Transplantation/or �Bone Marrow Transplantation/or �Cryptogenic Organizing Pneumonia/or exp �Lung Diseases, Fungal/or exp �Aspergillus/ or exp �Aspergillosis/or �Opportunistic Infections/

2 bronchoalveolar lavage/or bronchoalveolar lavage fluid/or (bronchoalveolar adj2 lavage).mp.

3 biopsy/or biopsy, needle/or biopsy, fine-needle/or endoscopic ultrasound-guided fine needle aspiration/or biopsy, large-core needle/or image-guided biopsy/

4 exp Lung/ 5 3 and 4 6 2 or 5 7 1 and 6 8 limit 10 to yr � “1980-Current”

EMBASE 1947 to 2014 Week 10 1 exp �neoplasm/or exp �stem cell transplantation/or exp �bone marrow transplantation/or �bronchiolitis obliterans organizing pneumonia/or exp �aspergillosis/or exp �Aspergillus/ or exp �lung mycosis/or �opportunistic infection/

2 65 lung lavage/or ((pulmonary or lung or bronchoalveolar) adj2 (lavage� or wash�)).ti,ab. or bal.ti,ab

3 lung biopsy/or open lung biopsy/or transthoracic biopsy/or (transthoracic adj2 biops�).ti,ab. 4 2 or 3 5 1 and 4 6 limit 5 to yr � “1980-Current”

EBM Reviews, Cochrane Central Register of Controlled Trials

January 2014 1 exp �neoplasms/or exp �Stem Cell Transplantation/or �Bone Marrow Transplantation/or �Cryptogenic Organizing Pneumonia/or exp �Lung Diseases, Fungal/or exp �Aspergillus/ or exp �Aspergillosis/or �Opportunistic Infections/or exp �Neoplasm/or exp �stem cell transplantation/or exp �bone marrow transplantation/or �bronchiolitis obliterans organizing pneumonia/or exp �lung mycosis/or �opportunistic infection/

2 lung lavage/or ((pulmonary or lung or bronchoalveolar) adj2 (lavage� or wash�)).ti,ab. or bal.ti,ab. or bronchoalveolar lavage/or bronchoalveolar lavage fluid/

3 lung biopsy/or open lung biopsy/or transthoracic biopsy/or (transthoracic adj2 biops�).ti,ab. or ((exp lung/and biopsy/) or biopsy, needle/or biopsy, fine-needle/or endoscopic ultrasound-guided fine needle aspiration/or biopsy, large-core needle/or image-guided biopsy/)

4 2 or 3 5 1 and 4 6 limit 5 to yr � “1980-Current”

Lung Procedures in Cancer and Stem-Cell Transplantation

Table A2. Characteristics of Included Studies of BAL in the Meta-Analysis

Reference Year

Published Year Study

Ended Diagnosis

Type Study Design Population Brushing� Indication Protocol†

Risk of Bias

Patient Selection

Index Test

Flow and Timing

Cordonnier et al51 1985 1983 HSCT Retrospective Both No No High High Low Cordonnier et al14 1986 1984 HSCT Retrospective Both No No High High Low Cordonnier et al52 1987 1985 HSCT Retrospective Both No Yes Low High Low Milburn et al53 1987 NA HSCT Retrospective Both No No Low High Low Saito et al54 1988 1987 Both Retrospective Adult Sometimes No High High Low Crawford et al55 1988 1987 HSCT Prospective Both No No Low Low Low Heurlin et al56 1989 1987 HSCT Retrospective Adult Sometimes No Low High Low Gleaves et al57 1989 NA HSCT Retrospective NA No No Low High Low Stokes et al58 1989 1988 Both Retrospective Pediatric Sometimes No Low High Low Levy et al59 1992 1988 Both Retrospective Adult Sometimes No High High Low McCubbin et al60 1992 1990 HSCT Retrospective Pediatric No No Low High High Weiss et al61 1993 1991 HSCT Prospective Both No No Low High Low Cathomas et al13 1993 1991 HSCT Prospective NA No No High Low Low Verweij et al11 1995 NA Both Prospective Adult No No Low Low Low von Eiff et al62 1995 1992 Cancer Prospective Both Yes No Low High Low Englund et al47 1996 1994 Both Prospective Adult No No Low Low Low Lanino et al63 1996 1996 HSCT Retrospective Pediatric No No Low High Low Heurlin et al64 1996 NA HSCT Retrospective Pediatric Sometimes No Low High Low White et al22 1997 1995 HSCT Retrospective Adult No No Low High Low Dunagan et al65 1997 1994 HSCT Retrospective Adult Sometimes No Low High Low Pagano et al66 1997 1996 Cancer Retrospective Both Sometimes No Low High Low Hennessy et al67 1998 1997 HSCT Retrospective Adult No No High High Low Glazer et al68 1998 1995 HSCT Retrospective Both No No Low High Low Jones et al44 1998 NA Cancer Retrospective Both No No Low High Low Ewig et al69 1998 1993 Both Retrospective Both Sometimes No High High Low Gruson et al70 1999 1997 HSCT Prospective Adult No Yes High High Low Heussel et al71 1999 NA Both Prospective Adult No Yes High High Low Salonen et al34 2000 1999 Both Retrospective Adult No No Low High Low Ramila et al72 2000 NA Both Prospective Both No Yes Low High Low Huaringa et al73 2000 NA HSCT Retrospective NA No No Low High Low Soubani et al21 2001 NA HSCT Retrospective Adult No No High High Low Murray et al74 2001 1999 Both Retrospective Adult No No Low High Low Ben-Ari et al75 2001 1999 HSCT Retrospective Pediatric No No Low High Low van Elden et al76 2002 2000 Both Retrospective Adult No No High High Low Buchheidt et al43 2002 2000 Both Prospective Adult No Yes Low Low Low Raad et al46 2002 1997 Cancer Prospective Both No No Low Low Low Park et al77 2002 1998 Cancer Retrospective Pediatric No No Low High Low Neumann et al78 2002 NA Cancer Prospective Pediatric No Yes Low Low Low Becker et al37 2003 2001 Both Prospective Adult No Yes Low Low Low Ison et al79 2003 NA HSCT Retrospective Both No No High High High Mulabecirovic et al19 2004 2002 Both Retrospective Adult Yes No Low High Low Patel et al20 2005 2001 HSCT Retrospective Adult No No Low High Low Hohenthal et al27 2005 2002 Both Retrospective Adult No No Low High Low Bissinger et al35 2005 NA Both Retrospective Adult No No Low High Low Martino et al80 2005 2003 HSCT Prospective Adult No Yes Low Low Low Eikenberry et al81 2005 1999 HSCT Retrospective Pediatric No Yes Low High Low Hofmeister et al30 2006 2004 HSCT Retrospective Both No No Low High High Gupta et al15 2007 2004 HSCT Retrospective Adult No No High High Low Burger et al82 2007 1998 HSCT Retrospective Adult No No Low High Low Boersma et al23 2007 NA Cancer Prospective Adult Sometimes Yes Low Low High Kasow et al24 2007 2002 HSCT Retrospective Both No No Low High Low Armenian et al16 2007 2003 Both Retrospective Pediatric No No Low High Low Rabbat et al25 2008 2002 Both Prospective Adult No No High High Low Penack et al41 2008 NA Both Prospective Adult No No Low High Low Khot et al45 2008 2003 Both Retrospective Adult No No Low High Low Azoulay et al31 2008 NA Both Prospective Adult No No High Low Low Cordani et al33 2008 2008 Cancer Retrospective Adult Sometimes No Low High Low Hummel et al36 2008 2003 Both Retrospective Both No Yes Low High High

(continued on following page)

Chellapandian et al

Table A2. Characteristics of Included Studies of BAL in the Meta-Analysis (continued)

Reference Year

Published Year Study

Ended Diagnosis

Type Study Design Population Brushing� Indication Protocol†

Risk of Bias

Patient Selection

Index Test

Flow and Timing

Furuya et al26 2008 2005 Cancer Retrospective Pediatric No Yes Low High Low Frealle et al38 2009 2004 Both Retrospective Adult No No Low Low Low Kuehnhardt et al48 2009 2004 Both Retrospective Both No No Low High Low Shannon et al28 2010 1999 HSCT Retrospective Adult No No Low High Low Azoulay et al49 2010 2007 Both Prospective Adult No Yes High Low Low Luong et al39 2010 2008 Both Retrospective Adult No No High Low High Bergeron et al32 2010 2006 Both Retrospective Adult No Yes Low Low Low Forslow et al29 2010 2004 HSCT Retrospective Both Sometimes No Low High Low Nguyen et al40 2011 2006 Both Retrospective Adult No No Low High Low Racil et al42 2011 2009 Cancer Retrospective Adult No No Low High High Reinwald et al12 2012 2011 Both Prospective Both No No Low Low High Gilbert et al18 2013 2009 HSCT Retrospective Adult Sometimes No Low High Low Gassas et al17 2013 2010 HSCT Retrospective Pediatric No No High High Low Rao et al50 2013 2009 Both Retrospective Pediatric No No Low High Low

NOTE. Under Diagnosis Type, “both” means the study included both HSCT and patients with cancer. Under Population, “both” means the study included both adult and pediatric patients. Abbreviations: BAL, bronchoalveolar lavage; HSCT, hematopoietic stem-cell transplantation; NA, not applicable. �Brushing used during the BAL process. †Indication for BAL standardized.

Lung Procedures in Cancer and Stem-Cell Transplantation

Table A3. Characteristics of Included Studies of Lung Biopsy in the Meta-Analysis

Reference Year

Published Year Study

Ended Diagnosis

Type Study Design Population Type of Biopsy

Indication Protocol�

Risk of Bias

Patient Selection

Index Test

Flow and Timing

Shulman et al93 1982 1980 HSCT Retrospective NA Transthoracic No Low High Low Springmeyer et al84 1982 1978 HSCT Prospective NA Transthoracic No High Low Low McCabe et al94 1985 1982 Cancer Retrospective Both Transthoracic No High High Low Hall et al95 1987 1986 Both Retrospective Both Transthoracic No Low High Low Crawford et al83 1988 1987 HSCT Retrospective Both Transthoracic No Low High Low Shorter et al96 1988 1986 HSCT Retrospective Pediatric Transthoracic No Low High Low Snyder et al97 1990 1988 HSCT Retrospective Both Transthoracic No Low High Low White et al22 1997 1995 HSCT Retrospective Adult Transbronchial No Low High Low Won et al98 1998 1997 Both Prospective Both Transthoracic No High High Low White et al99 2000 1998 Both Retrospective Adult Transthoracic No Low High Low Soubani et al21 2001 NA HSCT Retrospective Adult Transbronchial No High High Low Dunn et al100 2001 1998 HSCT Retrospective Pediatric Transthoracic No Low High Low Hayes-Jordan et al101 2002 1998 HSCT Retrospective Pediatric Transthoracic No Low High Low Kim et al102 2002 2001 Both Retrospective Both Transthoracic No Low High Low Shaikh et al103 2002 1999 HSCT Retrospective Both Transthoracic No Low High Low Jantunen et al104 2002 1998 HSCT Retrospective Both Transthoracic No Low High Low Nosari et al105 2003 2002 Cancer Retrospective Adult Transthoracic No Low High Low Gulbahce et al86 2004 1998 HSCT Retrospective Both Transthoracic No High High Low Wang et al85 2004 2001 HSCT Retrospective Both Transthoracic No High High Low Mulabecirovic et al19 2004 2002 Both Retrospective Adult Transbronchial No Low High Low Patel et al20 2005 2001 HSCT Retrospective Adult Transbronchial No Low High Low Zihlif et al106 2005 2003 Both Retrospective Adult Transthoracic No Low High Low Carrafiello et al107 2006 2003 Cancer Retrospective Adult Transthoracic No Low High Low Yang et al108 2007 2005 HSCT Retrospective Adult Transthoracic No High High Low Armenian et al16 2007 2003 Both Retrospective Pediatric Transthoracic† No Low High Low Armenian et al16 2007 2003 Both Retrospective Pediatric Transthoracic† No Low High Low Kallenberg et al87 2009 2007 Both Retrospective Adult Transthoracic No Low Low Low Gupta et al88 2010 2005 Cancer Retrospective Adult Transthoracic No Low High Low Gilbert et al18 2013 2009 HSCT Retrospective Adult Transbronchial No Low High Low Gassas et al17 2013 2010 HSCT Retrospective Pediatric Transthoracic No Low High Low Sharma et al89 2013 2011 Cancer Prospective Adult Transthoracic Yes Low High Low

NOTE. Under Diagnosis Type, “both” means the study included both HSCT and patients with cancer. Under Population, “both” means the study included both adult and pediatric patients. Abbreviations: HSCT, hematopoietic stem-cell transplantation; NA, not available. �Indication for bronchoalveolar lavage standardized. †Study described open and image-guided lung biopsies as two separate cohorts.

Chellapandian et al

Bronchoalveolar Lavage and Lung Biopsy in Patients With Cancer and Hematopoietic Stem-Cell Trans ...

INTRODUCTION
METHODS

Data Sources and Searches
Study Selection
Data Abstraction and Methodologic Approach
Assessment of Study Quality
Statistical Methods

RESULTS
DISCUSSION
REFERENCES
Acknowledgment
Appendix