Pleural fluid lysozyme as a diagnostic biomarker of pleural tuberculosis: A systematic review and meta-analysis.

Abstract

Pleural fluid lysozyme (LP) and its ratio to serum lysozyme (LP/LS) have been proposed as potential biomarkers for diagnosing tuberculous pleural effusion (TPE). We assessed the diagnostic accuracy of LP and LP/LS for TPE and evaluated their ability to differentiate TPE from other effusions. We queried the PubMed and Embase databases for studies indexed until October 2021. We included studies that (a) provided information regarding the sensitivity and specificity of LP or LP/LS for the diagnosis of TPE, or (b) compared LP or LP/LS between TPE and malignant or parapneumonic effusions. We used hierarchical summary receiver operating characteristic plots to model summary sensitivity and specificity. Random effects modeling was employed to pool standardized mean differences (SMD) across descriptive studies comparing TPE and other effusions. We included 11 publications in our review, most of which were small and of poor quality. The summary estimates for sensitivity, specificity, and diagnostic odds ratio (DOR) were 0.94 (95% confidence interval [CI] 0.53-1.00), 0.89 (95% CI 0.63-0.97), and 129.88 (95% CI 6.26-2695), and 0.98 (95% CI 0.58-1.00), 0.91 (95% CI 0.84-0.96), and 708.47 (95% CI 11.42-43946), respectively, for LP and LP/LS. Mean LP and LP/LS in TPE were significantly higher than in malignant effusions (summary SMD 1.51 [95% CI 1.04-1.98] and 1.77 [95% CI 1.31-2.22], respectively), and parapneumonic effusions (summary SMD 0.86 [95% CI 0.51-1.22] and 1.15 [95% CI 0.64-1.66], respectively). There is low-quality evidence of good diagnostic accuracy for both LP and LP/LS in identifying TPE, the latter being marginally superior.