Bacteriological methods as add on tests to fine-needle aspiration cytology in diagnosis of tuberculous lymphadenitis: can they reduce the diagnostic dilemma?

The diagnostic accuracy of fine-needle aspiration (FNA) cytology for the diagnosis of tuberculous lymphadenitis (TBLN) is confounded by mimicking cytomorphologic disorders. The objective of this study was to determine whether supplementing FNA cytology with bacteriological methods improves the overall accuracy of TBLN diagnosis.

Cytomorphology detected TBLN among 80% (160/200) of the presumptive cases. Culture results were available for 188 cases. Twelve samples were excluded due to contamination on both culture methods. Culture confirmed cases accounted for 78% (147/188) of which MTBc constituted 97.3% (143/147). Among presumptive cases, classified by FNA cytology as ‘abscess’, 11 were culture positive. Microscopy detected 31.3% (46/147) of culture confirmed mycobacterial lymphadenitis of which 11% (4/37) were diagnosed non-suggestive for tuberculosis (TB) by FNA cytology. Compared to culture (LJ & BACTEC MGIT 960) and AFB microscopy as composite gold standard, FNA cytology had a sensitivity of 88.4% and a specificity of 48.8%. The positive predictive value was 86.1% while the negative predictive value was 54.1%. The confirming power and the ROC curve area was 1.73 and 0.69, respectively.

FNA cytology showed a relatively high sensitivity but a low specificity. Combining bacteriological methods with FNA cytology in an endemic region like Ethiopia improves the overall accuracy of the diagnosis of mycobacterial lymphadenitis, which in turn may lead to better patient management.