The limit of detection (LOD) and clinical application value of MassARRAY were evaluated using reference strains and clinical isolates. MTB in bronchoalveolar lavage fluid (BALF) and sputum samples were detected using MassARRAY, quantitative real-time polymerase chain reaction (qPCR) and MGIT960 liquid culture (culture). Using culture as the standard, the efficacy of MassARRAY and qPCR for the detection of TB was analyzed. Mutation of drug resistance genes in MTB clinical isolates was tested using MassARRAY, high-resolution melting curve (HRM), and Sanger sequencing. Using sequencing as the standard, the efficacy of MassARRAY, and HRM for the detection of each drug resistance site of MTB was analyzed. Simultaneously, the mutation of drug resistance genes by the MassARRAY method was compared with the results of drug susceptibility testing (DST), and the genotype–phenotype relationship was analyzed. The ability of MassARRAY to discriminate mixed infections was detected using mixtures of standard strains (M. tuberculosis H37Rv) and drug-resistant clinical isolates and mixtures of wild-type and mutant plasmids.
In MassARRAY, 20 related gene mutations could be detected by two PCR systems. All genes could be accurately detected when the bacterial load was 104 CFU/mL. When the load of wild-type and drug-resistant MTB mixture was 105 CFU/mL (respectively reached 104 CFU/mL), variants and wild-type genes could be detected simultaneously. The sensitivity of MassARRAY (96.9%) for identification was higher than that of qPCR (87.5%) (
MassARRAY can obtain base mutation information and identify heteroresistance infections simultaneously when the mutant proportion was at least 5–25%. It has good application prospects in the diagnosis of DR-TB with high throughput, accurate and low-cost.