Line probe assay for detection of Mycobacterium tuberculosis complex: An experience from Central India

BACKGROUND & OBJECTIVES: Mycobacterium tuberculosis complex may sometimes not be detected in sputum samples of suspected multidrug-resistant tuberculosis (MDR-TB) patients by line probe assay (LPA) even though they are smear positive for acid-fast bacilli (AFB). This retrospective analysis was attempted to understand and document our experience with LPA for detection of M. tuberculosis complex and diagnosis of MDR-TB under programmatic conditions. METHODS: One thousand two hundred and ninety four sputum samples of MDR-TB suspects that were smear positive for AFB, and received from February to November 2013, were tested by LPA for the presence of M. tuberculosis complex and resistance to isoniazid (INH) and rifampicin as per the diagnostic mandate of an accredited reference laboratory. As per the mandate, those samples that were negative for M. tuberculosis complex were cultured, and the growth again tested by LPA. A retrospective analysis of the results was carried out. RESULTS: M. tuberculosis complex could be detected in 1217 (94.04%) but not in 77 (5.9%) of smear-positive sputum samples. Of the 1217 positive samples, 232 (19.1%) were MDR, 130 (10.6%) were rifampicin monoresistant and 101 (8.3%) were INH monoresistant. Seven hundred and fifty four (61.9%) strains were found to be pansensitive. Overall, 5.1 per cent of the sputum samples were negative for M. tuberculosis complex by LPA and culture. In at least 10 (0.77%) sputum samples smear positive for AFB, M. tuberculosis complex could not be identified by LPA though M. tuberculosis was present, as evidenced by culture positivity. INTERPRETATION & CONCLUSIONS: LPA is a robust technique for diagnosis of drug-resistant TB that has provided the basis for rapid and effective control of drug-resistant TB in India. While the reasons for concomitantly negative LPA and culture results of smear-positive sputum samples from MDR-TB suspects may be many, the possible presence of non-tubercular mycobacteria in these samples and the likelihood of inappropriate therapy in these patients cannot be ruled out. Addition of culture to the diagnostic algorithm may enhance the diagnostic yield.