Performance of a targeted next generation sequencing assay for mycobacterial identification and drug resistance from...

Accurate identification and genotypic drug resistance prediction in mycobacteria are essential for early diagnosis and appropriate clinical management. Targeted next-generation sequencing (tNGS) detects multiple genetic targets and enables direct testing from primary specimens. We evaluated identification using the Deeplex Myc-TB assay for Mycobacterium tuberculosis complex (MTBC) and nontuberculous mycobacteria (NTM) as well as MTBC drug-resistance prediction from 78 isolates and 102 sputum samples. DNA libraries were sequenced (Illumina MiSeq), and results were analyzed using defined acceptability criteria. Performance was compared against reference phenotypic and genotypic methods. Testing for MTBC was invalid in <2.5% of cases, but NTM testing was invalid more frequently, especially in sputum (>50%). Of the evaluable isolates/sputum, specificity was 100% for negative/nontarget organisms. Among isolates, sensitivity for organism identification was 92.7% (38/41) for MTBC, 100% (5/5) for rapidly growing mycobacteria (RGM), and 87.5% (14/16) for slowly growing mycobacteria (SGM). For sputum, sensitivity was 97.6% (41/42) for MTBC, 70% (7/10) for SGM, and 66% (2/3) for RGM. Categorical agreement of MTBC resistotypes against reference susceptibility methods was 91.9% (489/532) for isolates and 90.8% (472/520) for sputum, increasing to 96.8% (489/505) and 97.3% (472/485), respectively, when variants of unknown significance were excluded. Limit-of-detection studies confirmed published thresholds of 100-1,000 genomes/PCR in both isolates and sputum. Deeplex Myc-TB provides highly specific identification and reliable genotypic resistance profiling from isolates and sputum, especially for MTBC. While identification of rare NTM species and mixed infections remain challenging, tNGS demonstrates substantial promise for rapid mycobacterial diagnostics.IMPORTANCERapid and accurate diagnosis of mycobacterial infections is essential for initiating effective treatment, interrupting transmission, and improving patient outcomes. Traditional phenotypic methods for organism identification and antibiotic susceptibility testing are slow and may be unreliable for certain drugs. This study evaluates the Deeplex Myc-TB targeted next-generation sequencing assay to identify mycobacteria and resistance-associated mutations from the Mycobacterium tuberculosis complex (MTBC) for both isolates and clinical sputum. Our findings show strong correlation of MTBC identification and resistance markers compared with reference methods.