Sage Journals: Discover world-class research

Abstract

Background and Objectives

Body cavity fluid examination sometimes presents a diagnostic challenge in cytology practice. This meta-analysis was undertaken to comprehensively assess the diagnostic potential of epithelial membrane antigen (EMA) in malignant effusions.

Materials and Methods

All relevant original articles about EMA in the diagnosis of malignant effusions published up to July 1, 2014 were retrieved. The overall sensitivity, specificity, positive and negative likelihood ratio, diagnostic odds ratio, and summary receiver operating characteristic (SROC) curve were pooled to evaluate the diagnostic value of EMA for malignant effusions using the Meta-Disc 1.4 and STATA 12.0 statistical software.

Results

Eleven studies met the inclusion criteria for the meta-analysis and the summary estimates for EMA in the diagnosis of malignant effusions were as follows: sensitivity 0.9 (95% CI 0.83-0.87), specificity 0.87 (95% CI 0.96-0.99), positive likelihood ratio 5.8 (95% CI 15.59-36.37), negative likelihood ratio 0.15 (95% CI 0.07-0.20) and diagnostic odds ratio 52.63 (95% CI 20.91-132.49). The SROC curve indicated that the maximum joint sensitivity and specificity (Q-value) was 0.88; the area under the curve was 0.94.

Conclusion

The present meta-analysis indicated that EMA may be a useful diagnostic tool with good sensitivity and specificity for differentiating malignant effusions from benign effusions.

Keywords

Diagnosis EMA Malignant effusion Meta-analysis

Introduction

Body cavity effusions may be the first manifestation of an underlying malignancy such as lung, breast, gastrointestinal or female genital adenocarcinoma and malignant mesothelioma (1-2-3). Distinguishing malignant effusions presents a common and sometimes difficult diagnostic challenge in day-to-day cytology practice due to wide morphological variation of mesothelial and malignant cells in serous effusions (4). In recent years, with the development of newer techniques, the diagnostic accuracy of effusion cytology has been enhanced (5). Immunocytochemical identification of carcinoma cells in effusions has been increasingly recognized especially when malignant cells are few in number and likely to be missed with conventional cytology. The epithelial membrane antigen (EMA) belongs to a group of proteins known as human milk fat globule membrane proteins, which are present in a wide variety of epithelia under normal and neoplastic conditions (6). An increasing number of studies are examining the ability of EMA antibody as a marker for the diagnosis of malignant effusions (3, 7-8-9-10-11-12-13-14-15-16). Nevertheless, the role of EMA in the differential diagnosis between malignant and benign effusions remains controversial. We therefore performed the present meta-analysis to establish the overall accuracy of EMA for differentiating malignant cells in serous effusions.

Materials and Methods

Search Strategy and Study Selection

PubMed, Embase and the Cochrane Library were searched to identify suitable studies up to July 1, 2014. We used the terms “EMA”, “epithelial membrane antigen”, “body fluids”, “effusions”, “sensitivity and specificity” and “accuracy”, while the syntax used for the search was ((“EMA”[All Fields] OR “epithelial membrane antigen”[All Fields]) AND (“effusions”[All Fields] OR “body fluids”[All Fields]) AND (“sensitivity”[All Fields] OR “specificity”[All Fields]) OR “accuracy”[All Fields])). The language was limited to English. Any other relevant studies cited in the reference lists of the selected papers were also evaluated to identify additional eligible studies. Conference abstracts and letters to journal editors were excluded because of the limited data presented in them.

To be eligible for inclusion in the meta-analysis, studies must evaluate EMA in the diagnosis of malignant effusions and provide sufficient data to calculate both sensitivity and specificity. Only studies including at least 10 body cavity fluid specimens were selected since very small studies may be vulnerable to selection bias. Two authors independently screened the articles for inclusion. Disagreements between the reviewers were resolved by consensus.

Data Extraction and Quality Assessment

Two authors independently extracted the data and reached a consensus on all items. Any discrepancies were resolved by discussion with a third author to reach a final consensus. The following data were collected from each study: (i) basic characteristics including the first author's name, year of publication, country, test method, cutoff value; (ii) diagnostic outcomes including sensitivity and specificity data. The methodological quality of the studies was assessed using guidelines published by the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies) tool, an evidence-based quality assessment tool for use in systematic reviews of diagnostic accuracy studies. The QUADAS-2 tool consists of 4 key domains that discuss patient selection, index test, reference standard and flow of patients through the study and timing of the index tests and reference standard (flow and timing) (17).

Statistical Analyses

The standard methods recommended for the diagnostic accuracy of meta-analyses were used in the present study (18). The following measures of test accuracy were computed for each study: sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR). The diagnostic threshold identified for each study was used to plot a summary receiver operating characteristic (SROC) curve (19). Spearman's rank correlation was performed as a test for threshold effect. The interstudy heterogeneity was calculated by the chi-square-based Q-test and the inconsistency index I². When a significant Q-test (p<0.05 or I²>50%) indicated heterogeneity among studies, the random-effect model (DerSimonian-Laird method) was conducted for the meta-analysis to calculate the pooled sensitivity, specificity and other related indexes of the studies; otherwise, the fixed-effect model (Mantel-Haenszel method) was used. Meta-regression was performed to investigate the source of heterogeneity within the included studies (inverse variance weighted) (20). The publication bias was tested by Deeks’ funnel plots (21). Analyses were performed using the following statistical software programs: STATA, version 12.0 (Stata Corporation) and Meta-Disc 1.4 for Windows (XI Cochrane Colloquium, Barcelona, Spain) (22, 23). All statistical tests were 2-sided, and significance was set at p<0.05.

Results

Characteristics and Quality of the Included Studies

The initial database search yielded 107 articles. From these 107 articles, 66 were excluded because they were not relevant to the study question, 10 were excluded because they were not human studies, 4 were excluded because they were reviews, letters and conference abstracts, leaving 27 potentially relevant studies. Full articles of potentially eligible studies were reviewed in depth and 11 individual studies met our inclusion criteria. The article selection process used in this study is summarized in Figure 1. The main clinical characteristics of the included studies are presented in Table I. Overall, the 11 selected studies included 609 individuals and had an average size of 55 individuals. The studies were performed in 7 countries (United States, Iran, Spain, Korea, Greece, Netherlands and China) and published between 1994 and 2011. As shown in Table II, the validity of the included trials was assessed using the QUADAS-2 tool. Based on the methods reported in each trial, each of the 14 components according to the QUADAS-2 criteria was graded as ‘‘yes’’, ‘‘unclear’’ or ‘‘no’’, which meant ‘‘low risk of bias’’, ‘‘uncertain of bias’’ and ‘‘high risk of bias’’, respectively (17).

Table I

Summary of the studies included in the meta-analysis

Study (ref.)	Country	Sample size	Diagnostic standard	Method	Cutoff	TP	FP	FN	TN
Illingworth et al 1994 (7)	UK	42	Histopathological or clinical	Smears	≥1-3 cells per field (×40) and/or 3-5 groups per slide staining	25	3	1	13
Kortsik et al 1995 (8)	Germany	54	Histological	Smears	Membranous or whole cell staining	27	9	7	11
Saad et al 2005 (9)	US	40	Histopathological or clinical	Cell blocks	Membrane staining	15	0	5	20
Ueda et al 2006 (10)	Japan	29	Histopathological or clinical	Smears	Cytoplasmic and/or membranous staining	20	5	1	3
Palaoro et al 2007 (11)	Argentina	36	Clinical	Smears	Cytoplasmic staining	16	2	3	15
Grefte et al 2008 (12)	Netherlands	34	Histological	Cell blocks	Cytoplasmic and/or membranous staining	23	1	0	10
Murugan et al 2009 (13)	India	77	Histopathological or clinical	Smears	Cytoplasmic and/or membranous staining	39	1	0	37
Shen et al 2009 (14)	US	73	Histological	Cell blocks	≥25% of cells staining	30	5	5	33
Su et al 2011 (15)	China	53	Histopathological or clinical	Cell blocks	Cytoplasmic staining	17	8	1	27
Kundu et al 2012 (3)	India	100	Histopathological or clinical	Smears	Membranous/cytoplasmic staining	21	3	7	69
Ensani et al 2011 (16)	Iran	71	Histological	Cell blocks	Cytoplasmic and membranous staining	59	1	4	7

TP = true positive; FP = false positive; FN = false negative; TN = true negative.

Table II

Summary of the methodological quality assessment of the included studies according to QUADAS-2 criteria

Study (ref.)	Risk of bias				Applicability concerns
Study (ref.)	Patient selection	Index test	Reference standard	Flow and timing	Patient selection	Index test	Reference standard
Illingworth et al 1994 (7)	LR	LR	UR	UR	LC	LC	LC
Kortsik et al 1995 (8)	UR	LR	LR	LR	LC	LC	LC
Saad et al 2005 (9)	UR	LR	LR	UR	LC	LC	LC
Ueda et al 2006 (10)	LR	LR	UR	UR	LC	LC	LC
Palaoro et al 2007 (11)	LR	LR	UR	UR	LC	LC	UC
Grefte et al 2008 (12)	LR	LR	UR	UR	LC	LC	LC
Murugan et al 2009 (13)	LR	LR	LR	UR	LC	LC	LC
Shen et al 2009 (14)	UR	LR	LR	LR	LC	LC	LC
Su et al 2011 (15)	UR	LR	LR	UR	LC	LC	LC
Kundu et al 2012 (3)	LR	UR	LR	LR	LC	LC	LC
Ensani et al 2011 (16)	LR	LR	LR	LR	LC	LC	LC

LR = low risk; HR = high risk; UR = unclear risk; LC = low concern; HC = high concern; UC = unclear concern.

Fig. 1

Flow chart of selection process for eligible articles.

Diagnostic Accuracy

The between-study heterogeneity was assessed by the I² index to choose the appropriate calculation model. Since significant heterogeneity between studies was observed in sensitivity, specificity, PLR, NLR and DOR data (I² = 66.8%, 76.9%, 80.5%, 49.8% and 55.6%, respectively; all p<0.05), the random-effect model was used in this meta-analysis.

Forest plots of the sensitivity and specificity of these 11 studies concerning EMA in the diagnosis of malignant effusions are shown in Figure 2. The pooled sensitivity and specificity were 0.9 (95% CI 0.86-0.93) and 0.87 (95% CI 0.82-0.9), respectively. The overall PLR and NLR were 5.8 (95% CI 3-11.22) and 0.15 (95% CI 0.09-0.25), respectively. The pooled DOR was 52.63 (95% CI 20.91-132.49). Figure 3 shows the SROC curves for EMA, which indicates the sensitivity versus 1-specificity of individual studies. We also used the Q-value, the intersection point of the SROC curve with a diagonal line from the left upper corner to the right lower corner of the ROC space, which corresponds to the highest common value of sensitivity and specificity for the test, for the overall measure of the discriminatory power of the test. Our data showed that the SROC curve for EMA was positioned near the desirable upper left corner and the Q-value was 0.88; the area under the curve (AUC) was 0.94, which suggested that the diagnostic accuracy of EMA reached a relatively high level.

Fig. 2

Forest plots of the sensitivity and specificity of EMA in the diagnosis of malignant effusions for all studies. The point estimates of sensitivity and specificity for each study are shown as solid circles and the size of each solid circle indicates the sample size of each study. Error bars are 95% confidence intervals (19).

Fig. 3

Summary receiver operating characteristic (SROC) curve for EMA in the diagnosis of malignant effusions for all studies. Solid circles represent each study included in the meta-analysis. The size of each solid circle indicates the size Applicability concernsch study. The regression SROC curve summarizes the overall diagnostic accuracy.

Meta-Regression and Publication Bias

As the I² test for the pooled sensitivity, specificity, NLR and DOR showed significant heterogeneity between studies, a meta-regression analysis was performed to explore the possible reasons for the heterogeneity. We used 4 covariates in the meta-regression: (i) test method (cell blocks or smears); (ii) sample size (≥50 or <50); (iii) risk of patient selection bias (low risk or not); (iv) risk of bias in reference standard (low risk or not). None of the above covariates were found to be a significant source of heterogeneity (p = 0.49, 0.73, 0.77 and 0.15, respectively).

The publication bias of the literature was assessed using Deeks’ funnel plots. The graphical funnel plots of the included studies were symmetrical (Fig. 4). The Deeks’ test also indicated that there were no publication biases (p = 0.41).

Fig. 4

Funnel graph for the assessment of potential publication bias of the 11 included studies. The funnel graph plots the log of the diagnostic odds ratio (DOR) against the standard error of the log of the DOR (an indicator of sample size). Solid circles represent each study in the meta-analysis. The line indicates the regression line.

Discussion

Serous effusion is a common complication of a variety of benign and malignant disorders. Although cytomorphology remains the principal method to detect malignant cells in effusions, the morphological changes exhibited by reactive mesothelial cells often confound the diagnosis in the cytological evaluation of body cavity effusions (24). Therefore, complementary methods such as immunocytochemistry are often used to help the cytological diagnosis. EMA is an antibody to a human milk fat globule membrane immunogen (25). EMA reactivity was seen in a variety of simple and glandular epithelia including the respiratory tract, pancreas, intestine and endometrium. It has been suggested that EMA positivity is a sign of malignancy in epithelial mesothelial proliferations (26, 27). In recent years, a number of diagnostic tests have focused on the value of EMA in the differential diagnosis of benign and malignant effusions, but these studies reported a range of specificities and sensitivities. As meta-analysis is an essential tool for accurately and reliably summarizing evidence (28), we performed this meta-analysis to assess the overall diagnostic accuracy of EMA for malignant effusions.

The results of the analysis, which included 11 independent clinical diagnostic tests with 609 patients, suggest that EMA has good overall sensitivity (90%) and specificity (87%) in the diagnosis of malignant effusions. The DOR of a diagnostic test is the ratio of the odds of positivity in disease relative to the odds of positivity in the non-diseased, which is a single indicator of test performance that combines the data from sensitivity and specificity into a single number (29). The value of a DOR ranges from 0 to infinity, with higher values indicating better discriminatory test performance. In this meta-analysis, the pooled DOR was 52.63, also suggesting a high level of overall accuracy. The SROC curve presents a global summary of the test performance, and shows the trade-off between sensitivity and specificity. The analysis-based SROC curve showed that the maximum joint sensitivity and specificity was 0.88; the AUC was 0.94, also confirming that EMA was very helpful in the diagnosis of malignant effusions, although not perfect. The PLR was 5.8 and the NLR was 0.15. Likelihood ratios >10 and <0.1 are considered strong indicators to rule in or rule out a diagnosis, respectively (30). In clinical practice, patients with malignant effusions have an about 6-fold higher chance of being EMA-positive than patients with benign effusions, and this was high enough. On the other hand, the NLR was 0.15, which means that the probability of having malignant effusions in EMA-negative patients is 15% in theory; this indicates that, although not as good as expected, measurement of this marker could be helpful in the diagnosis of malignant effusions in clinical use.

An exploration of the reasons for heterogeneity rather than computation of a single summary measure is an important purpose of meta-analysis (31). In our meta-analysis, the effects of different methods, sample sizes and QUADAS results on relative diagnostic odds ratios (RDOR) were assessed. However, we did not find that the variables included in the meta-regression analysis substantially affected the diagnostic accuracy of EMA for malignant effusions.

Our data suggest that EMA may, to a certain extent, play a role in the diagnosis of malignant effusions. However, based on the findings in our meta-analysis, we cannot recommend using a single marker alone for the diagnosis of malignant effusions, and combination of EMA with other tumor markers seems to increase the diagnostic accuracy (32). In addition to the antibody analyzed in the present meta-analysis, some other antibodies such as MOC-31, Ber-EP4, CA19-9 and CEA have also been evaluated for their use in the diagnosis of malignant effusions. A recent meta-analysis showed that the summary estimates for MOC-31 in the diagnosis of malignant effusions were: sensitivity 0.85 (95% CI 0.83-0.87) and specificity 0.97 (95% CI 0.96-0.99) (33). Some investigators reported that calretinin and Ber-EP4 had high sensitivity (90% and 98%, respectively) and high specificity (100% and 86%, respectively) in differentiating malignant from benign effusions (34). However, it should be pointed out that there are insufficient related studies to evaluate the combination of EMA and other tumor markers in the diagnostic accuracy for malignant effusions, and due to the varying diagnostic accuracy of identical markers between studies it is unclear which markers perform best. Therefore, more immunomarkers should be evaluated for their diagnostic accuracy in larger sample sizes to find the optimum panel of antibodies for the diagnosis of malignant effusions.

Our study has several limitations. First, the exclusion of conference abstracts, letters to journal editors and non-English language studies may have resulted in publication bias, and the accuracy estimates may have been inflated due to the preferential acceptance of papers reporting favorable results. Secondly, studies with small sample sizes appeared to overestimate the diagnostic accuracy of EMA for the diagnosis of malignant effusions. Thirdly, because of a lack of required data reported in the original publications, we could not analyze the effect of factors such as laboratory infrastructure, immunotechnology expertise, and patient setting on the accuracy of the EMA measurements. For the same reason we were unable to ascertain whether the study design (e.g., blinded, cross-sectional, consecutive/random, prospective) affected the diagnostic accuracy. Further study is therefore needed to evaluate the diagnostic accuracy of EMA in clinical applications.

It should also be noted that the diagnosis of malignant effusions by immunocytochemistry is still flawed by technical problems. Cytological examination of malignant effusions in routine practice can contribute to diagnosis only in cases with adequate cytological preparations (35). Moreover, in effusion cytology, specific staining may not be obtained because of disrupted cells and membrane fragments sticking to slides, and there is a lack of parallel samples of the same cells for additional tests or to run the control (36). Experimental conditions and methods of different laboratories may also affect the diagnostic results. Therefore, in the era of precision medicine, obtaining histological material still plays a crucial role that cytological examination cannot replace for investigations in targeted therapies. Future work should focus on increasing the homogeneity and overall performance of cytological diagnosis for malignant effusions to minimize the current weaknesses.

In conclusion, our meta-analysis demonstrated that EMA may be a useful adjunct to conventional diagnostic tools with good sensitivity and specificity for differentiating malignant from benign effusions, while the results of immunostaining should be interpreted in parallel with the gold standard of morphology and clinical findings.

Footnotes

Abbreviations

Acknowledgment

We thank all authors of the primary studies included in our meta-analysis.

Financial support: None.

Conflict of interest: None.

References

McGrath

Anderson

Diagnosis of pleural effusion: a systematic approach. Am J Crit Care 2011 20 2 119–127 quiz 128.

Davidson

The diagnostic and molecular characteristics of malignant mesothelioma and ovarian/peritoneal serous carcinoma. Cytopathology 2011 22 1 5–21

Kundu

Handa

Mohan

Role of DNA flow cytometry and immunocytochemical analysis in diagnosis of malignant effusions. Diagn Cytopathol 2012 40 10 887–892

Davidson

Malignant effusions: from diagnosis to biology. Diagn Cytopathol 2004 31 4 246–254

Saha

Dey

Vhora

Nijhawan

Role of DNA flow cytometry and image cytometry on effusion fluid. Diagn Cytopathol 2000 22 2 81–85

Dion

Williams

Herlyn

Major

Human milk fat globule membrane glycoproteins express blood group-related determinants primarily on mucin-like epithelial membrane antigens and gp70. Biochem Int 1990 22 2 295–302

Illingworth

Young

Johnson

Immunofluorescent staining of metastatic carcinoma cells in serious fluid with carcinoembryonic antibody, epithelial membrane antibody, AUA-1 and Ber-EP4. Cytopathology 1994 5 5 270–281

Kortsik

Werner

Freudenberg

. Immunocytochemical characterization of malignant mesothelioma and carcinoma metastatic to the pleura: IOB3—a new tumor marker. Lung 1995 173 2 79–87

Saad

Cho

Liu

Silverman

The value of epithelial membrane antigen expression in separating benign mesothelial proliferation from malignant mesothelioma: a comparative study. Diagn Cytopathol 2005 32 3 156–159

10.

Ueda

Iwata

Ono

Takahashi

Comparison of three cytologic preparation methods and immunocytochemistries to distinguish adenocarcinoma cells from reactive mesothelial cells in serous effusion. Diagn Cytopathol 2006 34 1 6–10

11.

Palaoro

Blanco

Gamboni

Rocher

Rotenberg

Usefulness of ploidy, AgNOR and immunocytochemistry for differentiating benign and malignant cells in serous effusions. Cytopathology 2007 18 1 33–39

12.

Grefte

de Wilde

Salet-van de Pol

Tomassen

Raaymakers-van Geloof

Bulten

Improved identification of malignant cells in serous effusions using a small, robust panel of antibodies on paraffin-embedded cell suspensions. Acta Cytol 2008 52 1 35–44

13.

Murugan

Siddaraju

Habeebullah

Basu

Immunohistochemical distinction between mesothelial and adenocarcinoma cells in serous effusions: a combination panel-based approach with a brief review of the literature. Indian J Pathol Microbiol 2009 52 2 175–181

14.

Shen

Pinkus

Deshpande

Cibas

Usefulness of EMA, GLUT-1, and XIAP for the cytologic diagnosis of malignant mesothelioma in body cavity fluids. Am J Clin Pathol 2009 131 4 516–523

15.

Liu

Xie

Jiang

Cytological differential diagnosis among adenocarcinoma, epithelial mesothelioma, and reactive mesothelial cells in serous effusions by immunocytochemistry. Diagn Cytopathol 2011 39 12 900–908

16.

Ensani

Nematizadeh

Irvanlou

Accuracy of immunohistochemistry in evaluation of malignant pleural and peritoneal effusions. Pol J Pathol 2011 62 2 95–100

17.

Whiting

Rutjes

Westwood

et al; QUADAS-2 Group. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011 155 8 529–536

18.

Jones

Ashrafian

Skapinakis

. Diagnostic accuracy meta-analysis: a review of the basic principles of interpretation and application. Int J Cardiol 2010 140 2 138–144

19.

Midgette

Stukel

Littenberg

A meta-analytic method for summarizing diagnostic test performances: receiver-operating-characteristic-summary point estimates. Med Decis Making 1993 13 3 253–257

20.

Baker

White

Cappelleri

Kluger

Coleman

Health Outcomes, Policy, and Economics (HOPE) Collaborative Group. Understanding heterogeneity in meta-analysis: the role of meta-regression. Int J Clin Pract 2009 63 10 1426–1434

21.

van Enst

Ochodo

Scholten

Hooft

Leeflang

Investigation of publication bias in meta-analyses of diagnostic test accuracy: a meta-epidemiological study. BMC Med Res Methodol 2014 14 1 70

22.

Zamora

Abraira

Muriel

Khan

Coomarasamy

Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006 6 1 31

23.

Whiting

Harbord

de Salis

Egger

Sterne

Evidence-based diagnosis. J Health Serv Res Policy 2008 13 Suppl 3 57–63

24.

Bisht

Handa

Mohan

Lehl

Complementary value of DNA flow cytometry and image morphometry in detection of malignant cells in effusion fluids. Malays J Pathol 2014 36 2 83–90

25.

Pinkus

Kurtin

Epithelial membrane antigen—a diagnostic discriminant in surgical pathology: immunohistochemical profile in epithelial, mesenchymal, and hematopoietic neoplasms using paraffin sections and monoclonal antibodies. Hum Pathol 1985 16 9 929–940

26.

Whitaker

The cytology of malignant mesothelioma. Cytopathology 2000 11 3 139–151

27.

Shield

Callan

Devine

Markers for metastatic adenocarcinoma in serous effusion specimens. Diagn Cytopathol 1994 11 3 237–245

28.

Jain

Sharma

Singh

Doing meta-analysis in research: a systematic approach. Indian J Dermatol Venereol Leprol 2012 78 3 242–250

29.

Zhang

Tian

Zhang

Hemoglobin A1c for diagnosis of postpartum abnormal glucose tolerance among women with gestational diabetes mellitus: diagnostic meta-analysis. PLoS ONE 2014 9 7 e102144

30.

Deeks

Altman

Diagnostic tests 4: likelihood ratios. BMJ 2004 329 7458 168–169

31.

Petitti

Approaches to heterogeneity in meta-analysis. Stat Med 2001 20 23 3625–3633

32.

Liang

Shi

Qin

Liang

Jiang

Yang

Diagnostic accuracy of tumour markers for malignant pleural effusion: a meta-analysis. Thorax 2008 63 1 35–41

33.

Wang

. Diagnostic accuracy of MOC-31 for malignant effusions: a meta-analysis. Tumour Biol 2014 35 6 6003–6009

34.

van der Bij

Schaake

Koffijberg

Burgers

de Mol

Moons

Markers for the non-invasive diagnosis of mesothelioma: a systematic review. Br J Cancer 2011 104 8 1325–1333

35.

Rakha

Patil

Abdulla

Abdulkader

Chaudry

Soomro

The sensitivity of cytologic evaluation of pleural fluid in the diagnosis of malignant mesothelioma. Diagn Cytopathol 2010 38 12 874–879

36.

Kulkarni

Desai

Ajit

Chinoy

Utility of the thromboplastin-plasma cell-block technique for fine-needle aspiration and serous effusions. Diagn Cytopathol 2009 37 2 86–90

Diagnostic Accuracy of Epithelial Membrane Antigen for Malignant Effusions: A Meta-Analysis

Abstract

Background and Objectives

Materials and Methods

Results

Conclusion

Keywords

Introduction

Materials and Methods

Search Strategy and Study Selection

Data Extraction and Quality Assessment

Statistical Analyses

Results

Characteristics and Quality of the Included Studies

Diagnostic Accuracy

Meta-Regression and Publication Bias

Discussion

Footnotes

Abbreviations

Acknowledgment

References