The indirect microscopic observation drug susceptibility assay demonstrated high concordance with the indirect MGIT...

The microscopic observation drug susceptibility (MODS) assay has been used for Mycobacterium tuberculosis detection and anti-TB drug susceptibility tests for several years, and our study aimed to evaluate the accuracy of MODS in detecting pyrazinamide resistance in MDR TB suspects.

One hundred and forty-eight clinical isolates were collected from 148 MDR TB suspects in the Nanjing Chest Hospital, and the MODS and the mycobacteria growth indicator tube (MGIT) 960 methods for pyrazinamide susceptibility testing were conducted for each isolate independently. pncA gene sequencing was applied to confirm results showing discrepancy between the MODS and MGIT 960 methods. The McNemar χ(2) test was employed to evaluate the paired 2 × 2 table.

Compared with the MGIT 960 method, the sensitivity and specificity of the MODS assay were 95.5% and 93.3%, respectively, with a high accuracy of 94.5%. The κ value of 0.89 showed near-perfect agreement for the two methods in determining pyrazinamide susceptibility. Eight clinical isolates showed inconsistent results by MODS and MGIT 960. After sequencing the pncA gene of the eight isolates, MODS and MGIT 960 showed no significant difference in detecting pyrazinamide resistance when compared with pncA gene sequencing (P = 0.655 for MODS and P = 0.564 for MGIT 960).

The high accuracy of MODS in detecting pyrazinamide resistance showed that the MODS method would be an alternative for pyrazinamide susceptibility testing; other mutations affecting resistance to pyrazinamide need to be identified in order to elucidate the discrepant results between phenotype- and genotype-based methods.