Diagnostic accuracy of pleural fluid carbohydrate antigen 72-4 for malignant pleural effusion: a systematic review and meta-analysis

Introduction

Malignant pleural effusion (MPE) is a common sign in patients with advanced cancer. It has a poor prognosis, with a median survival of less than one year (1). Pleural effusion is not a specific sign of advanced cancer. It can also be caused by various diseases, including tuberculous pleurisy, heart failure, and pneumonia (2). Therefore, timely and accurate diagnosis of MPE in patients with undiagnosed pleural effusion is essential. The gold standards for MPE are cytology and pleural biopsy. Although the specificity of effusion cytological is 1.00, its sensitivity is only around 0.50 (3). For patients with negative effusion cytology, pleural biopsy is needed. However, pleural biopsy is invasive, hindering its implication in some patients, such as those with end-stage liver disease or coagulation disorders (4,5). In addition, the operation-related complications caused by biopsy are problematic (6). Tumor markers in the pleural fluid are alternative diagnostic tools for MPE due to their advantages of less invasiveness, low cost, rapidity and objectiveness (7-9). Numerous tumor markers found in pleural fluid can aid in diagnosing MPE. Notable markers include carcinoembryonic antigen (CEA) (10) and carbohydrate antigen 50 (CA50) (11). However, when used individually, these markers are insufficient for definitively confirming or ruling out MPE (8). Utilizing multiple tumor markers may enhance the diagnostic accuracy for MPE (8,12). Evaluating the diagnostic accuracy of a single tumor marker is essential for developing algorithms that incorporate multiple tumor markers for improved diagnosis.

Carbohydrate antigen 72-4 (CA72-4) is a tumor-associated epithelial mucin highly expressed in gastric, colon, and breast cancers (13). Many studies have investigated the value of serum CA72-4 in diagnosing cancers. Systematic reviews and meta-analyses revealed that it had moderate diagnostic accuracy for gastric cancer (14) and colon cancer (15). Several studies have evaluated its value in distinguishing MPE from benign pleural effusion (BPE) (16-19), but the results varied. The varied findings may be attributed to the participants enrolled, the CA72-4 assay, or the reference standard for MPE. No systematic review or meta-analysis has been performed to pool the findings of published studies.

This systematic review and meta-analysis aimed to evaluate the accuracy of pleural fluid CA72-4 for diagnosing MPE. We present this article in accordance with the PRISMA-DTA reporting checklist (20) (available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1664/rc).

Methods

Search strategy

The search algorithm in the PubMed database and Web of Science database was: (“CA 72-4” OR “CA 724” OR “CA-72-4” OR “CA-724” OR “CA72-4” OR “CA724” OR “carbohydrate antigen 724” OR “carbohydrate antigen 72-4” OR “carbohydrate antigen-72-4” OR “carbohydrate antigen-724” OR “cancer antigen 724” OR “cancer antigen 72-4” OR “cancer antigen-72-4” OR “cancer antigen-724” OR “CA-72-4 antigen”[nm]) AND (pleur* OR effusion*). The last search date was August 2024.

Study selection

The inclusion criteria of this systematic review and meta-analysis follow the PIDTA framework (21). P, patients with undiagnosed pleural effusion; I, the index test should be pleural fluid CA72-4; D, the study design can be prospective or retrospective; T, the target disease is MPE. Usually, the BPE should encompass tuberculous pleurisy, heart failure, or pneumonia; A, the study aimed to evaluate the diagnostic accuracy of pleural CA72-4. Studies meeting one of the following criteria were excluded: (I) animal studies; (II) non-English publications; (III) conference abstracts, literature reviews, commentaries, and case reports; (IV) a two-by-two table cannot be constructed due to the insufficient details reported. All retrieved literature was screened independently by two systematic reviewers. In the first round of literature screening, the reviewers read the title and abstract to exclude irrelevant studies. In the second round, the reviewers read the full text of the literature to determine the eligibility of the remaining studies.

Data extraction and risk of bias assessment

The data extracted from the eligible studies were the name of the first author, nationality of the participants, year of publication, the sample sizes of MPE and BPE, type of data collection (prospective or retrospective), CA72-4 assay, the reference standards for MPE and BPE, sensitivity, specificity, the area under the curve (AUC), and the threshold used to define positive CA72-4. According to the sample sizes of MPE and BPE, sensitivity, and specificity, a two-by-two table was constructed for each eligible study. The two-by-two table includes the numbers of patients with true positive (TP), false positive (FP), false negative (FN), and true negative (TN) CA72-4.

The revised diagnostic accuracy study quality assessment tool-2 (QUADAS-2) was to assess the quality of eligible studies (22). The data extraction and quality assessment were performed independently by two systematic reviewers. Any disagreement was resolved by consensus.

Statistical analysis

A bivariate model was utilized to combine the sensitivity and specificity along with their 95% confidence intervals (CIs) from the eligible studies (23). A summary receiver operating characteristic (sROC) curve was used to summarize eligible studies, and the AUC of sROC was used to measure the overall diagnostic performance of CA72-4 (24). The Deeks’s test and a funnel plot were used to evaluate the degree of publication bias (25), and I² was used to determine the heterogeneity among studies (26). All data analyses were performed using Revman and Stata 16.0 (Stata Corp LP, College Station, TX, USA). A P value less than 0.05 was defined as statistically significant.

Results

Study selection process and characteristics of the eligible studies

Figure 1 is a flow chart depicting the study selection process. Eight studies with 828 MPEs and 963 BPEs were included in this meta-analysis. Five studies were prospective design, and one was retrospective (19). Participants in these studies were from China (n=1) (17), USA (n=1) (27), Brazil (n=1) (16), France (n=1) (19), Sweden (n=1) (28), and Spain (n=3) (18,29,30). CA72-4 levels in pleural fluid were measured by radioimmunoassay in two studies (27,29). One study used immunoradiometry to determine CA72-4 (30). Two studies used enzyme immunoassay to determine CA72-4 (19,28), and three used an electrochemiluminescence assay (16-18). The eligible studies are summarized in Table 1.

Figure 1 A flow chart of study selection.

Quality assessment

Table 2 depicts the quality assessment results of the eligible studies. The patient selection domain in two studies was labeled high because patients were not consecutively enrolled or enrolled retrospectively (17,19). The index test domain in four studies was labeled high because the authors used a data-driven threshold to define positive CA72-4 (16,19,29,30). The flow and timing in two studies were high because not all patients were included in the data analysis (27,30).

Main findings of the eligible studies

The diagnostic accuracy of CA72-4 for MPE is summarized in Table 3. The threshold values of CA72-4 ranged between 3.4 and 21 U/mL, with sensitivities between 0.33 and 0.68 and specificity between 0.83 and 1.00. The AUCs reported in eligible studies ranged between 0.60 and 0.80.

Meta-analysis

Figure 2 is the forest plots of CA72-4 for diagnosing MPE. The pooled sensitivity was 0.47 (95% CI: 0.39–0.55), and the pooled specificity was 0.98 (95% CI: 0.95–0.99). The I² of sensitivity and specificity were 0.88 and 0.89. The positive likelihood ratio (PLR) and negative likelihood ratio (NLR) were 26.6 (95% CI: 10.0–70.1) and 0.54 (95% CI: 0.47–0.63).

Figure 2 Forest map of CA72-4 in the pleural fluid to diagnose MPE. CI, confidence interval; CA72-4, carbohydrate antigen 72-4; MPE, malignant pleural effusion.

Figure 3 is the sROC curve of CA72-4. The AUC was 0.77 (95% CI: 0.73–0.80). Figure 4 is the funnel plot of CA72-4. Deeks’s test showed a significant publication bias across all eligible studies (P=0.03). Figure 5 shows a Fagan nomogram of CA72-4 for diagnosing MPE. Assuming the prevalence of MPE is 50% in a target population, the positive predictive value (PPV) of CA72-4 was 0.96, indicating that the possibility of MPE is 96% in patients with positive CA72-4. The negative predictive value (NPV) was about 35%, suggesting the possibility of BPE was 35% in patients with negative CA72-4.

Figure 3 The sROC curve of pleural fluid CA72-4 for diagnosing MPE. sROC, summary receiver operating characteristic; SENS, sensitivity; SPEC, specificity; AUC, area under the curve; CI, confidence interval; CA72-4, carbohydrate antigen 72-4; MPE, malignant pleural effusion.

Figure 4 A funnel plot assessing publication bias of CA 72-4 for diagnosing MPE. ESS, effective sample size; CA72-4, carbohydrate antigen 72-4; MPE, malignant pleural effusion.

Figure 5 A Fagan plot depicting the predictive values of CA72-4 under different prevalences. LR, likelihood ratio; CA72-4, carbohydrate antigen 72-4.

Discussion

This is the first systematic review and meta-analysis to investigate the diagnostic accuracy of pleural fluid CA72-4 for MPE. The present meta-analysis included eight studies with 828 MPEs and 963 BPEs. The pooled sensitivity and specificity of CA72-4 for MPE were 0.47 (95% CI: 0.39–0.55) and 0.98 (95% CI: 0.95–0.99), respectively. The AUC of sROC was 0.77 (95% CI: 0.73–0.80). The eligible studies have significant publication bias. These findings indicate pleural fluid CA72-4 is an auxiliary diagnostic marker for MPE.

The pooled sensitivity of CA72-4 was 0.47 (95% CI: 0.39–0.55), indicating that 53% of MPE would be missed if CA72-4 was used alone to confirm MPE. On the other hand, the pooled specificity of CA72-4 was 0.98 (95% CI: 0.95–0.99), indicating that only 2% of BPE would be misdiagnosed as MPE if CA72-4 was used alone to exclude MPE. Therefore, CA72-4 has moderate diagnostic accuracy for MPE. It may be used to confirm MPE rather than exclude.

Sensitivity and specificity are two basic diagnostic metrics of a diagnostic marker. A limitation of these two metrics is threshold-dependent. In other words, they only reflect the accuracy of a marker at a specified threshold. For example, most eligible studies set their thresholds between 3.4 and 21 U/mL, and the corresponding sensitivity and specificity were 0.40 and 1.00. These findings suggest that the sensitivity and specificity of CA72-4 were 0.40 and 1.00, with a threshold between 3.4 and 21 U/mL. What sensitivity and specificity will be obtained if a higher threshold is used remains unclear. Therefore, sensitivity and specificity are not suitable metrics to measure the global diagnostic accuracy of a marker (31). In contrast, the AUC of sROC is not affected by the threshold used and thus represents a global indicator of diagnostic accuracy (32). The AUC of the sROC curve is between 0.5 and 1.0, and a higher AUC indicates higher diagnostic accuracy (24). The AUC of CA72-4 was 0.77 (95% CI: 0.73–0.80), which also supports that CA72-4 is an auxiliary diagnostic marker for MPE.

NLR and PLR are two prevalence-independent metrics that exclude or confirm target disease. It is generally believed that PLR >10 strongly suggests the presence of disease, and NLR <0.1 strongly supports the exclusion of target disease (33). The PLR of CA72-4 was 26.6 (95% CI: 10.0–70.1), indicating that the risk of MPE was extremely high in patients with positive CA72-4. By contrast, the NLR of CA72-4 was 0.54 (95% CI: 0.47–0.63), which does not support the implication of CA72-4 in excluding MPE. Therefore, from a statistical perspective, patients with elevated CA72-4 can be used to confirm the presence of MPE.

The QUADAS-2 tool was used to assess the quality of the included studies. The available studies had some design weaknesses. Some studies did not enroll participants consecutively, which may impair the participants’ representativeness of the available studies. The generalizability of these studies is thus problematic. The thresholds used in some studies were data-driven. This threshold selection method may overestimate the diagnostic accuracy of a biomarker (34). Further studies with rigorous designs are needed to validate the findings of available studies.

Many conventional tumor markers have been reported as useful diagnostic markers for MPE, such as CEA, neuron-specific enolase (NSE), carbohydrate antigen 125 (CA125), carbohydrate antigen 153 (CA153), carbohydrate antigen 199 (CA199), and cytokeratin-19-fragment (CYFRA21-1) (35). Two meta-analyses reported these tumor markers had a pooled sensitivity of 0.50 and a specificity of 0.90 for diagnosing MPE (36,37). The pooled sensitivity and specificity of CA72-4 were 0.47 (95% CI: 0.39–0.55) and 0.98 (95% CI: 0.95–0.99), respectively, indicating that the accuracy of CA72-4 in diagnosing MPE is comparable to that of conventional tumor markers. One study reported that CA72-4 had a specificity of 1 at the threshold of 21 U/mL (18). Therefore, CA72-4 >21 U/mL can be used to confirm MPE. This hypothesis is also supported by our previous study (17).

Publication bias is not uncommon in clinical research. Positive findings have a high probability of being published in academic journals. One aim of meta-analysis is to verify the presence of publication bias. The present study revealed a significant publication bias, suggesting that previous studies may have overestimated the diagnostic accuracy of CA72-4.

There are several limitations. First, the number of included studies, as well as the total sample size, is small. Therefore, the 95% CIs of sensitivity and specificity were wide. Further studies with large sample sizes are needed to improve the precision of sensitivity and specificity. Second, significant heterogeneity was observed across eligible studies. Still, due to the small number of included studies, subgroup analysis or meta-regression cannot be performed to explore the source of heterogeneity. Third, due to the limited resources, we did not search non-English databases; thus, the language bias cannot be excluded.

Conclusions

In summary, pleural fluid CA72-4 is an auxiliary diagnostic marker for MPE. With thresholds between 3.4 and 21 U/mL, CA72-4 helps confirm MPE, but its value is limited in ruling out MPE. Patients with CA72-4 >21 U/mL had an extremely high risk of MPE. Considering the small sample size, the design weaknesses of the eligible studies, and publication bias, further studies with large sample sizes and rigorous designs remain needed to evaluate the diagnostic accuracy of CA72-4 for MPE.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the PRISMA-DTA reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1664/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1664/prf

Funding: This study was supported by funding from the Program for Young Talents of Science and Technology in Universities of Inner Mongolia Autonomous Region (No. NJYT22018, to Z.D.H.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1664/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Roberts ME, Neville E, Berrisford RG, et al. Management of a malignant pleural effusion: British Thoracic Society Pleural Disease Guideline 2010. Thorax 2010;65:ii32-40. [Crossref] [PubMed]
2. Light RW. Pleural effusions. Med Clin North Am 2011;95:1055-70. [Crossref] [PubMed]
3. Kassirian S, Hinton SN, Cuninghame S, et al. Diagnostic sensitivity of pleural fluid cytology in malignant pleural effusions: systematic review and meta-analysis. Thorax 2023;78:32-40. [Crossref] [PubMed]
4. Wei Y, Shen K, Lv T, et al. Comparison between closed pleural biopsy and medical thoracoscopy for the diagnosis of undiagnosed exudative pleural effusions: a systematic review and meta-analysis. Transl Lung Cancer Res 2020;9:446-58. [Crossref] [PubMed]
5. Dixon G, de Fonseka D, Maskell N. Pleural controversies: image guided biopsy vs. thoracoscopy for undiagnosed pleural effusions? J Thorac Dis 2015;7:1041-51. [Crossref] [PubMed]
6. Goh KJ, Leong CK, Young SL, et al. Diagnostic value and safety of medical thoracoscopy in undiagnosed pleural effusions-a prospective observational cohort study. J Thorac Dis 2024;16:3142-51. [Crossref] [PubMed]
7. Zheng WQ, Hu ZD. Pleural fluid biochemical analysis: the past, present and future. Clin Chem Lab Med 2023;61:921-34. [Crossref] [PubMed]
8. Zheng WQ, Porcel JM, Hu ZD. Tumor markers determination in malignant pleural effusion: pearls and pitfalls. Clin Chem Lab Med 2025;63:515-20. [Crossref] [PubMed]
9. Chan C, Chan KKP. Pleural fluid biomarkers: a narrative review. J Thorac Dis 2024;16:4764-71. [Crossref] [PubMed]
10. Shi HZ, Liang QL, Jiang J, et al. Diagnostic value of carcinoembryonic antigen in malignant pleural effusion: a meta-analysis. Respirology 2008;13:518-27. [Crossref] [PubMed]
11. Cha SN, Niu Y, Wen JX, et al. Pleural carbohydrate antigen 50 and malignant pleural effusion: a prospective, double-blind diagnostic accuracy test. Transl Lung Cancer Res 2024;13:1061-8. [Crossref] [PubMed]
12. Yang Y, Liu YL, Shi HZ. Diagnostic Accuracy of Combinations of Tumor Markers for Malignant Pleural Effusion: An Updated Meta-Analysis. Respiration 2017;94:62-9. [Crossref] [PubMed]
13. Li M, Men X, Zhang X. Diagnostic value of carbohydrate antigen 72-4 combined with carbohydrate antigen 15.3 in ovarian cancer, cervical cancer and endometrial cancer. J BUON 2020;25:1918-27.
14. Wang H, Jin W, Wan C, et al. Diagnostic value of combined detection of CA72-4, CA19-9, and carcinoembryonic antigen comparing to CA72-4 alone in gastric cancer: a systematic review and meta-analysis. Transl Cancer Res 2022;11:848-56. [Crossref] [PubMed]
15. Yanqing H, Cheng D, Ling X. Serum CA72-4 as a Biomarker in the Diagnosis of Colorectal Cancer: A Meta-analysis. Open Med (Wars) 2018;13:164-71. [Crossref] [PubMed]
16. Antonangelo L, Sales RK, Corá AP, et al. Pleural fluid tumour markers in malignant pleural effusion with inconclusive cytologic results. Curr Oncol 2015;22:e336-41. [Crossref] [PubMed]
17. Cao XS, Yan L, Jiang TW, et al. Pleural fluid carbohydrate antigen 72-4 and malignant pleural effusion: a diagnostic test accuracy study. Ther Adv Respir Dis 2024;18:17534666231222333. [Crossref] [PubMed]
18. Trapé J, Sant F, Franquesa J, et al. Evaluation of two strategies for the interpretation of tumour markers in pleural effusions. Respir Res 2017;18:103. [Crossref] [PubMed]
19. Miédougé M, Rouzaud P, Salama G, et al. Evaluation of seven tumour markers in pleural fluid for the diagnosis of malignant effusions. Br J Cancer 1999;81:1059-65. [Crossref] [PubMed]
20. Salameh JP, Bossuyt PM, McGrath TA, et al. Preferred reporting items for systematic review and meta-analysis of diagnostic test accuracy studies (PRISMA-DTA): explanation, elaboration, and checklist. BMJ 2020;370:m2632. [Crossref] [PubMed]
21. Zhang M, Hu ZD. Suggestions for designing studies investigating diagnostic accuracy of biomarkers. Ann Transl Med 2019;7:788. [Crossref] [PubMed]
22. Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529-36. [Crossref] [PubMed]
23. Reitsma JB, Glas AS, Rutjes AW, et al. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005;58:982-90. [Crossref] [PubMed]
24. Walter SD. Properties of the summary receiver operating characteristic (SROC) curve for diagnostic test data. Stat Med 2002;21:1237-56. [Crossref] [PubMed]
25. Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol 2005;58:882-93. [Crossref] [PubMed]
26. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-60. [Crossref] [PubMed]
27. Ferroni P, Szpak C, Greiner JW, et al. CA 72-4 radioimmunoassay in the diagnosis of malignant effusions. Comparison of various tumor markers. Int J Cancer 1990;46:445-51. [Crossref] [PubMed]
28. Ustün H, Borazan A, Bilgiçli N, et al. Diagnostic value of tumoural markers in pleural effusions. Int J Clin Pract 2004;58:22-5. [Crossref] [PubMed]
29. Villena V, López-Encuentra A, Echave-Sustaeta J, et al. Diagnostic value of CA 549 in pleural fluid. Comparison with CEA, CA 15.3 and CA 72.4. Lung Cancer 2003;40:289-94. [Crossref] [PubMed]
30. Villena V, López-Encuentra A, Echave-Sustaeta J, et al. Diagnostic value of CA 72-4, carcinoembryonic antigen, CA 15-3, and CA 19-9 assay in pleural fluid. A study of 207 patients. Cancer 1996;78:736-40. [Crossref] [PubMed]
31. Altman DG, Bland JM. Diagnostic tests. 1: Sensitivity and specificity. BMJ 1994;308:1552. [Crossref] [PubMed]
32. Reitsma JB, Moons KG, Bossuyt PM, et al. Systematic reviews of studies quantifying the accuracy of diagnostic tests and markers. Clin Chem 2012;58:1534-45. [Crossref] [PubMed]
33. Deeks JJ, Altman DG. Diagnostic tests 4: likelihood ratios. BMJ 2004;329:168-9. [Crossref] [PubMed]
34. Leeflang MM, Moons KG, Reitsma JB, et al. Bias in sensitivity and specificity caused by data-driven selection of optimal cutoff values: mechanisms, magnitude, and solutions. Clin Chem 2008;54:729-37. [Crossref] [PubMed]
35. Zhang M, Yan L, Lippi G, et al. Pleural biomarkers in diagnostics of malignant pleural effusion: a narrative review. Transl Lung Cancer Res 2021;10:1557-70. [Crossref] [PubMed]
36. Nguyen AH, Miller EJ, Wichman CS, et al. Diagnostic value of tumor antigens in malignant pleural effusion: a meta-analysis. Transl Res 2015;166:432-9. [Crossref] [PubMed]
37. Wu Q, Li M, Zhang S, et al. Clinical diagnostic utility of CA 15-3 for the diagnosis of malignant pleural effusion: A meta-analysis. Exp Ther Med 2015;9:232-8. [Crossref] [PubMed]