Clinical evaluation of multiplex PCR for the differential diagnosis of major pathogenic mycobacteria in East China.

Distinguishing between Mycobacterium tuberculosis (MTB) and nontuberculous mycobacteria infections remains clinically challenging due to their drastically different treatment regimens, underscoring the critical need for rapid and accurate mycobacterial identification to guide effective patient management. This study aimed to develop a multiplex polymerase chain reaction (PCR) assay for the detection and typing of major pathogenic mycobacteria - including MTB, Mycobacterium avium complex (MAC), Mycobacteroides chelonae - Mycobacteroides abscessus group (MCAG), and Mycobacterium kansasii - and to evaluate its diagnostic performance. The assay was designed with species-specific primers targeting 5 key genes: 16S rRNA, RD9, ITS (internal transcribed spacer) region-MAC, ITS region-MCAG, and mkan-rs12360. Its feasibility and lower limit of detection (LOD) were validated using 6 standard mycobacterial strains; diagnostic accuracy was further assessed using 68 clinical isolates and 130 clinical sputum samples, with whole-genome sequencing and targeted next-generation sequencing serving as the respective gold standards for isolates and sputum specimens. The multiplex PCR assay exhibited high specificity for the target genes and primers: for standard strains, the LOD ranged from 0.7 to 5.0 pg/μL; for artificially simulated sputum samples, the LOD was 5.6 to 26.7 pg/μL for single infections and 7.1 to 43.9 pg/μL for mixed infections. Against whole-genome sequencing, the assay achieved 100% detection rate and accuracy for clinical isolates. When compared with targeted next-generation sequencing for clinical sputum samples, the detection rates for single infections of MTB, MAC, MCAG, and M kansasii were 26%, 75%, 56.5%, and 50%, respectively, with an overall detection rate of 51.2% and 100% diagnostic accuracy. This multiplex PCR assay enables rapid and accurate identification of major pathogenic mycobacteria in clinical isolates; however, its sensitivity limitations - evidenced by relatively low detection rates in clinical sputum samples, especially mixed infections - highlight the need for further optimization with more sensitive probe-based technologies to enhance clinical applicability.