Date Published: March 26, 2019
Publisher: Public Library of Science
Author(s): Ashutosh Nath Aggarwal, Ritesh Agarwal, Inderpaul Singh Sehgal, Sahajal Dhooria, George A. Eapen.
Pleural fluid adenosine deaminase (ADA) is a useful diagnostic test for tuberculous pleural effusion (TPE), but its exact threshold and accuracy in clinical decision-making is unclear. We aimed to assess diagnostic performance of ADA in TPE and to clarify its optimal diagnostic threshold.
We searched PubMed, Embase, and Cochrane Library databases for articles indexed up to October 2018. We included English language studies that provided both sensitivity and specificity of ADA in TPE diagnosis. Summary estimates for sensitivity and specificity were obtained through bivariate random effects model, both overall and at prespecified threshold ranges of <36, 40±4, 45–65 and >65 IU/L.
We retrieved 2162 citations, and included 174 publications with 27009 patients. All studies showed high risk of bias. Summary sensitivity, specificity and diagnostic odds ratio estimates were 0.92 (95% CI 0.90–0.93), 0.90 (95% CI 0.88–0.91) and 97.42 (95% CI 74.90–126.72) respectively. 65 studies with ADA threshold of 40±4 IU/L showed summary sensitivity and specificity of 0.93 (95% CI 0.90–0.95) and 0.90 (95% CI 0.87–0.91) respectively. Four studies with ADA threshold >65 IU/L showed summary sensitivity and specificity of 0.86 (95% CI 0.61–0.96) and 0.94 (95% CI 0.80–0.99) respectively.
ADA levels in pleural fluid show good diagnostic accuracy in diagnosis of TPE; however, all included studies showed high risk of bias. It was not possible to derive any firm inference on relative clinical utility of different diagnostic thresholds.
Tuberculosis (TB) remains a major cause of morbidity and mortality throughout the world. The World Health Organization (WHO) estimated nearly 10.4 million new cases globally in 2015, with maximum burden in the developing world . Pulmonary TB is the commonest form of TB, and a definite microbiological diagnosis is possible in most instances through sputum microscopy, culture and the use of cartridge-based nucleic acid amplification tests (CBNAAT). Extrapulmonary TB, representing nearly 15% of the global TB burden, is more difficult to diagnose. Tuberculous pleural effusion (TPE), one of the commonest forms of extrapulmonary TB, is a diagnostic challenge with rather poor microbiologic confirmation rates from pleural fluid analysis [2, 3]. Even diagnostic tools like CBNAAT and interferon-gamma release assays have shown suboptimal diagnostic accuracy [4, 5].
We performed this review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [16, 17]. An Institutional Review Board approval was not necessary as this work involved systematic review and meta-analysis of previously published studies. We searched the PubMed, Embase, and Cochrane Library databases, for articles indexed up to December 2016. We updated our search on October 31, 2018. Our search strategy for the PubMed database was (“adenosine deaminase”[MeSH Terms] OR “adenosine deaminase”[All Fields] OR “ADA”[All Fields]) AND (“pleura”[MeSH Terms] OR “pleura”[All Fields] OR “pleural”[All Fields] OR “pleuritis”[All Fields] OR “pleurisy”[MeSH Terms] OR “pleurisy”[All Fields] OR “nonrespiratory”[All Fields] OR “non-respiratory”[All Fields] OR “extrapulmonary”[All Fields] OR “extra-pulmonary”[All Fields]). A similar search strategy was used for the other two databases.
We retrieved 2162 citations from our database search, and assessed 657 full-text articles after initial title and abstract review (Fig 1). We finally selected 174 studies for data synthesis (S1 Table of online supplement). Details on these individual publications are provided in S2 and S3 Tables of the online supplement. The studies finally included for review yielded information on 10696 TPE patients and 16313 patients with pleural effusions of other etiologies, giving an overall TPE prevalence of 39.6%. 84 (48.3%) studies were conducted in countries currently designated by the WHO as ‘high burden countries’. 120 (69.0%) studies had a prospective design. Blinding was ensured in 26 (14.9%) studies only. Only 16 (9.1%) studies reported on inclusion of human immunodeficiency seropositive subjects. 85 (48.9%) studies included patients with exudative effusions only. 80 (46.0%) studies used a ‘definite TB’ reference standard. 68 (39.1%) studies used composite clinical criteria for diagnosing TPE whereas 26 (14.9%) studies provided no such information. Most studies (105, 60.3%) reported ADA measurements as per the colorimetric procedure proposed by Guisti; however, the diagnostic thresholds varied widely (S3 Table, online supplement). 18 (10.3%) studies reported diagnostic accuracy data using multiple ADA thresholds. In all, we analyzed 65 (37.4%) studies using a ADA threshold from 36–44 IU/L, and another 56 (32.2%) using 45–65 IU/L as the ADA threshold. The most common threshold used in our analysis was 40 IU/L (43 studies having 2718 TPE patients and 3306 patients of other effusions). Only four (2.3%) studies provided data from thresholds of 70 IU/L or more.
The previous largest meta-analysis on this topic included 63 studies indexed till March 2007 . We have updated this by 111 additional studies, both published during the last decade as well as older studies not included in the earlier review. In addition, we also address the clinically important issue of choosing an appropriate ADA threshold during patient evaluation.