The sensitivity and specificity of pleural fluid MTB-specific elispot assay among adults with suspected tuberculous pleurisy compared to pleural biopsy in Mulago Hospital

Abstract

Introduction and rationale: Pleural effusions are a common reason for admission in Mulago hospital and tuberculosis is the most common infectious cause. The diagnosis of tuberculous pleural effusion with standard analysis of pleural fluid is inaccurate. Pleural tissue histopathology and culture for Mycobacterium tuberculosis is considered the gold standard but the procedure is invasive, requires skilled clinicians and does not provide immediate results. To address these limitations, we evaluated whether an MTB-specific Enzyme Linked Immunospot (Elispot) assay performed on pleural fluid mononuclear cells accurate and feasible for diagnosing tubercuous pleurisy. Objectives: To evaluate the diagnostic accuracy of MTB-specific Elispot assay performed on pleural fluid mononuclear cells among adults with suspected tuberculous pleural effusion at Mulago hospital pulmonology unit. Methods: This cross-sectional study was conducted in Mulago Hospital, pulmonology unit. The study was done in two phases, initial phase between October 2008 and April 2009 and the latter phase in June 2009. Patients above 18 years with clinical and radiological suspicion of tuberculous pleural effusion following chart review with pleural effusion greater than 25% of hemithorax, and provided written informed consent were consecutively screened. All screened patients underwent diagnostic thoracocentesis. Those with exudative pleural effusion by Light’s criteria and lymphocyte predominant effusions were enrolled. Standardized questionnaire was administered and patients underwent pleural biopsy. Biopsy specimens were cultured on Lowenstein-Jensen media and submitted for histopathology. At the time of biopsy, 50 mL of pleural fluid was collected for isolation of pleural fluid mononuclear cell. In the first phase the cells were stored in liquid nitrogen and Elispot run as a batch whereas in the second phase Elispot was run on fresh cells. Also 15 mL of pleural fluid was collected during biopsy for cytology. Elispot assays were performed if the pleural fluid mononuclear cell count was greater than 1000, 000 and viability greater than 67% after storage or on fresh samples. Characteristics of patients were compared using Fisher’s exact test for dichotomous variables and Wilcoxon rank-sum test for continuous variables. Sensitivity and specificity of Elispot were calculated using combined results of pleural tissue culture and histopathology as a reference standard. Results: Of the 51 patients who underwent pleural biopsy, 25 patients had Elispot done. The rest 26 (51%) patients could not undergo Elispot because of inadequate pleural fluid mononuclear cell yield. Patients with TB and those without had similar characteristics except for a higher axillary temperature among those with TB. The prevalence of TB pleural effusion was 64%. The sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios were 81.3%, 55.6%, 76.5%, 62.5%, 1.83 and 0.34 respectively. Among the 51 pleural biopsies done, 35 (68.6%) had TB pleural effusion by combined gold standards. Among the 25 Elispot assays done, 16 (64%) had TB by combined gold standards and 13 (52%) had TB by both Elispot and gold standards. Median spot forming cell counts were higher in patients with tuberculous pleural effusion compared to those without for both ESAT-6 (790 vs. 142, p=0.01) and CFP-10 (731 vs. 27, p=0.11). Conclusions: The diagnostic accuracy of commercial (T-SPOT.TB) Elispot assay may be low when used alone. It may be a clinically useful adjunct test for diagnosing TB pleural effusion. Recommendations: Larger studies are needed in our setting to determine accurately the diagnostic value of pleural fluid Elispot assay for TB pleurisy. Future studies should carefully consider procedures for collecting, processing and storing pleural fluid mononuclear cells for Elispot assay.