INTRODUCTION: Previous studies have assessed the diagnostic ability of pleural fluid adenosine deaminase (pfADA) in detecting tuberculous pleural effusions, with good specificity and sensitivity reported. However, in North Western Europe pfADA is not routinely used in the investigation of a patient with an undiagnosed pleural effusion, mainly due to a lack of evidence as to its utility in populations with low mycobacterium tuberculosis (mTB) incidence.
METHODS: Patients presenting with an undiagnosed pleural effusion to a tertiary pleural centre in South-West England over a 3 year period, were prospectively recruited to a pleural biomarker study. Pleural fluid from consecutive patients with robust 12-month follow up data and confirmed diagnosis were sent for pfADA analysis.
RESULTS: Of 338 patients enrolled, 7 had confirmed tuberculous pleural effusion (2%). All mTB effusions were lymphocyte predominant with a median pfADA of 72.0 IU/L (range- 26.7 to 91.5) compared to a population median of 12.0 IU/L (range- 0.3 to 568.4). The optimal pfADA cut off was 35 IU/L, which had a negative predictive value (NPV) of 99.7% (95% CI; 98.2-99.9%) for the exclusion of mTB, and sensitivity of 85.7% (95% CI; 42.2-97.6%) with an area under the curve of 0.88 (95% CI; 0.732-1.000).
DISCUSSION: This is the first study examining the diagnostic utility of pfADA in a low mTB incidence area. The chance of an effusion with a pfADA under 35 IU/L being of tuberculous aetiology was negligible. A pfADA of over 35 IU/L in lymphocyte-predominant pleural fluid gives a strong suspicion of mTB.