Tuberculosis (TB) is a preventable and curable disease, yet it continues to pose a substantial global health challenge. According to the World Health Organization's (WHO) 2024 Global Tuberculosis Report, 8.2 million new TB cases were reported in 2023 [1], an increase from 7.5 million in 2022 and 7.1 million in 2019. Additionally, an estimated 1.25 million TB-related deaths occurred in 2023, potentially reclaiming TB as the leading cause of death from a single infectious agent, surpassing COVID-19 [1]. Despite its global burden, only 48% of newly diagnosed TB cases in 2023 underwent WHO-recommended testing. This diagnostic gap is largely attributed to the lack of highly sensitive, rapid and accessible diagnostic tools [2].

Drug-resistant TB remains a significant challenge in TB management, with isoniazid (INH)-monoresistance being the most common form of drug resistance [3]. Multidrug-resistant TB (MDR-TB), defined as resistance to at least INH and rifampicin (RIF), poses an even greater threat. In Taiwan, INH-monoresistant, RIF-monoresistant and MDR-TB accounted for 8.6%, 0.7% and 0.9%, respectively, of newly diagnosed TB cases in 2023 [4]. Treating INH-monoresistant TB with standard first-line therapy is associated with higher risks of treatment relapse and the development of MDR-TB [5], while MDR-TB is linked to high rates of treatment failure [6,7] and increased mortality [8].

The End TB Strategy emphasizes the importance of universal drug susceptibility testing (DST) for all confirmed TB cases [9]. However, traditional methods – mycobacterial culture and phenotypic DST (pDST) – are labour-intensive and time-consuming. Nucleic acid amplification tests (NAATs) provide a promising alternative by enabling direct detection of MTBC DNA along with drug resistance-associated mutations, through molecular DST (mDST) [10]. In particular, mutations in the rpoB gene (conferring RIF resistance) and in the katG gene or inhA promoter region (conferring INH resistance) can be detected more rapidly by mDST compared to pDST [11].

The Roche cobas MTB and MTB-RIF/INH assays, designed for use on the cobas 6800/8800 systems, have been endorsed by the WHO as part of the moderate-complexity NAAT class [10]. The cobas MTB assay is an automated real-time polymerase chain reaction (PCR) test for detecting MTBC DNA. It employs a dual-target approach using two sets of primers and probes that targets distinct DNA regions, namely, the 16S ribosomal RNA gene and esx gene family (esxJ, esxK, esxM, esxP and esxW) [10]. The MTB-RIF/INH assay is a reflex test for MTBC-positive samples that detects mutations associated with RIF and INH resistance in the rpoB, katG and inhA promoter regions [12]. However, limited comparative performance data are available for the cobas MTB and MTB-RIF/INH assays relative to culture and other commercial NAATs, such as the Xpert MTB/RIF Ultra assay [[12], [13], [14]]. Moreover, geographic variability in resistance-associated mutations may affect assay performance [12,13], underscoring the need for additional data from diverse settings to inform global and national policy decisions.

This study evaluated the diagnostic performance of the cobas MTB and MTB-RIF/INH assays for detecting MTBC and resistance to INH and RIF and compared it with that of the Xpert MTB/RIF Ultra assay using conventional mycobacterial culture with pDST as reference standards.