Background: Accurate identification and drug susceptibility testing (DST) of nontuberculous mycobacteria (NTM) are critical for the management of NTM pulmonary disease. Conventional methods such as Sanger sequencing and line probe assays have limited ...
Background: Accurate identification and drug susceptibility testing (DST) of nontuberculous mycobacteria (NTM) are critical for the management of NTM pulmonary disease. Conventional methods such as Sanger sequencing and line probe assays have limited discriminatory power and scope. This study aimed to evaluate the feasibility of implementing next-generation sequencing (NGS) for routine NTM identification and genotypic resistance prediction in a clinical microbiology laboratory.
Methods: The study was conducted in four phases: (1) evaluation of DNA extraction methods, (2) a pilot phase assessing rare or inconclusive isolates, (3) a comparative evaluation with Sanger sequencing (n=106), and (4) routine diagnostic application (n=236). NGS was performed using the NovaSeq 6000Dx platform (Illumina, USA), and taxonomic classification was based on Kraken2 and Bracken, followed by de novo genome assembly and average nucleotide identity (ANI) analysis. Genotypic resistance to clarithromycin and amikacin was predicted using known mutations in rrl, rrs, and erm(41).
Results: In the comparative evaluation phase, genome-based identification showed 91.4% concordance with Sanger sequencing (96/106). NGS identified mixed Mycobacterium infections (3.8%), rare species (4.7%), and a non-mycobacterial organism (0.9%) that were not detected by Sanger sequencing. Among 78 isolates of M. avium, M. intracellulare, and M. abscessus complex, genotypic predictions for clarithromycin and amikacin showed 100% concordance with phenotypic DST. In the routine diagnostic phase, NGS was applied to 236 isolates. A single Mycobacterium species was identified in 226 cases (95.8%), mixed infections in 2 cases (0.8%), and non-mycobacterial organisms in 8 cases (3.4%). The most common species were M. intracellulare (37.7%) and M. avium (31.8%). Among 160 isolates with DST data, genotypic-phenotypic concordance for clarithromycin was 100% in M. avium and M. massiliense, 99.1% in M. intracellulare, and for amikacin, 100%, 100%, and 97.4%, respectively. Inducible macrolide resistance in M. abscessus was also accurately predicted by erm(41) alleles. NGS reduced time from specimen receipt to test initiation (median 6 vs. 9 days, P < 0.001), but required more time for post-sequencing analysis. Total turnaround time was comparable to Sanger sequencing (median 14 vs. 13 days, P = 0.065).
Conclusions: NGS provides comprehensive species-level identification and accurate genotypic resistance prediction for NTM and can be integrated into routine clinical diagnostics with turnaround times similar to conventional methods.