Article Text
PDF
Abstract
Background Five percent of disease due to Mycobacterium tuberculosis is reported to be associated with a pleural effusion. We conducted a retrospective review of all patients at our institution treated for pleural tuberculosis (TB) between 2010 and 2015, reviewing demographics, fluid microbiology, histology, and radiology to characterise our cohort and review diagnostic certainty.
Results 71% (36/51) of our cohort were male; with a mean age of 41 years. 59% of cases had clear evidence of concurrent parenchymal involvement (five did not have cross sectional imaging). All effusions were exudates. 46 patients had imaging one year after treatment started. 16/46 had residual fluid, 16/46 demonstrated complete resolution and 14/46 had pleural thickening. Pleural fluid TB culture was performed in 47/51 patients and was positive in 43% of cases. Pleural fluid TB PCR was performed in 18/51 patients. 3/18 were PCR positive; in 1 patient PCR provided a diagnosis 10 days before culture. Two PCR Results provided diagnoses despite negative cultures and demonstrated a rifampicin resistance mutation in one case. 21 pleural biopsies were done, 13 were video assisted thoracoscopic procedures and 8 were taken with closed cutting needles. Pleural biopsy TB culture was performed on 16/21 samples. 63% (n=10) were culture positive. 4/6 culture negative samples that were sent for histopathology also demonstrated granulomas. The combined yield of fluid and biopsy culture in addition to fluid PCR was 57% in our cohort increasing to 65% when histology is included. 4 patients had a positive culture or PCR from another site. 15/51 (29%) of patients were treated empirically without positive TB cultures, positive TB PCR results, or granulomas on biopsy.
Conclusions This analysis demonstrates that the characteristics of our cohort are similar to previously reported cohorts; that TB PCR of pleural fluid has provided some diagnostic benefits, but that there remains diagnostic uncertainty in a significant proportion of patients. New tools are required to improve diagnostic accuracy of this difficult disease and we propose that NICE-approved analysis of fluid for adenosine deaminase may be a useful additional investigation for this patient group.
