Tumor necrosis factor (TNF) is an important local host response mediator in tuberculous pleural effusion (TPE) and is proposed as a potential biomarker for diagnosing TPE. We assessed the performance of pleural fluid TNF in the diagnosis of TPE, and evaluated its ability to distinguish TPE from parapneumonic or malignant effusions.
We queried the PubMed and Embase databases for studies indexed till August 2020. We included studies that (a) provided data on sensitivity and specificity of pleural fluid TNF for the diagnosis of TPE, or (b) compared pleural fluid TNF levels between TPE and malignant or parapneumonic effusions. We used a hierarchical summary receiver operating characteristic plot to model summary sensitivity and specificity. A random effects model was used to pool standardized mean differences (SMD) across studies comparing TPE and other effusions. We explored heterogeneity using subgroup analysis. We also performed meta-regression to identify factors significantly influencing results.
We retrieved 1090 citations, and included 38 publications, in our review. The summary estimates for sensitivity, specificity, and diagnostic odds ratio were 0.79 (95% CI 0.72-0.84), 0.82 (95% CI 0.76-0.87), and 16.84 (95% CI 9.47-29.95) respectively. Pleural fluid TNF levels were significantly higher in TPE than in malignant effusions (summary SMD 1.50, 95% CI 1.13-1.87), but not parapneumonic effusions (summary SMD 0.61, 95% CI -0.14 to 1.35). None of the prespecified subgroup variables significantly influenced summary estimates.
Pleural fluid TNF has poor diagnostic accuracy for diagnosing TPE and imperfectly discriminates TPE from parapneumonic pleural effusions.