RESUMO
BACKGROUND: Few studies in routine settings have confirmed the high accuracy of the Xpert MTB/RIF assay for detecting rifampicin resistance (RR) and the first-line probe assay (FL-LPA) for detecting both RR and isoniazid resistance (INHR). METHODS: The performance of Xpert MTB/RIF and MTBDRplus VER 2.0 LPA was evaluated in 180 Mycobacterium tuberculosis samples collected from January 2018 to December 2019 in Rio de Janeiro, Brazil. The results were compared with those from BACTEC MGIT 960 culture and drug susceptibility testing (DST). Whole-genome sequencing was performed on the samples with discordant results. RESULTS: The Xpert MTB/RIF assay showed a sensitivity (Se) of 93.3% and a specificity (Sp) of 97.6%, detecting RR. The performance of FL-LPA to identify RIF and INH resistance was, respectively, (Se) 100% and 83.3% and (Sp) 98.8% and 100%. Among 18 clinical isolates with INHR detected by FL-LPA, mutations in the katG gene were observed in 100% of samples, of which only two (11.1%) had mutations in both katG and inhA genes. Overall, the discordant results were identified in 9 (5%) samples. Among the four Xpert RIF-resistant and DST-sensitive, two harbored mutations in rpoB Leu430Pro. Among the four FL-LPA-sensitive and DST-resistant, one had a mutation in inhA 17G>T. FL-LPA showed high accuracy in detecting RR and INHR. CONCLUSIONS: The MTBDRplus test demonstrated excellent performance in detecting RR, and INHR in clinical isolates under routine conditions at a reference laboratory in Rio de Janeiro, Brazil. Incorporating both tests can improve drug-resistant tuberculosis treatment outcomes and monitor the INHR incidence.