Improving antitubercular drug susceptibility testing with liquid media.

Liquid media such as the mycobacteria growth indicator tube (MGIT) are widely used for antitubercular drug susceptibility testing (DST) but may cause discordant or invalid results due to ordinary bacterial contamination or mixed mycobacterial colonies. To overcome the drawbacks of DST with MGIT (DSTMGIT), an experiment with a modified method that incorporated a subculture step on the Löwenstein-Jensen (LJ) media with positively identified liquid media of MGIT (modified DSTMGIT) was performed. For 307 samples identified as Mycobacterium tuberculosis complex from 2010 to 2012, DSTMGIT and modified DSTMGIT were performed for isoniazid, rifampin, streptomycin, and ethambutol using BACTEC MGIT 960. Those results were compared to the results of DST with the solid media, LJ (DSTLJ). We obtained drug susceptibility test results from 287 (93.5%) specimens; however, 20 (6.5%) specimens showed invalid results with error messages from the system using DSTMGIT. Thirteen (65.0%) of the 20 invalid DSTMGIT results were made valid using a modified form of DSTMGIT. DSTMGIT and the modified DSTMGIT methods had 41 (14.3%) discordant results to the reference method , DSTLJ. Also, five of these samples were falsely identified as multidrug-resistant. The percent of discordant results was similar between DSTMGIT and the modified DSTMGIT (P=0.624), but the bacterial contamination or mycobacterial mixture rate was significantly lower in the modified DSTMGIT (P=0.029). Modified DSTMGIT can reduce bacterial contamination or mixed mycobacterial cultures and can be useful for samples with discordant or invalid DST results.