Comparison of the Mycobacteria Growth Indicator Tube (MGIT) with radiometric and solid culture for recovery of acid-fast bacilli

In a multicenter study involving three reference centers for mycobacteria, the rate of recovery of acid-fast bacilli (AFB) and the mean time to their detection from clinical specimens was determined by using the Mycobacteria Growth Indicator Tube (MGIT). These parameters were compared to those assessed by the radiometric BACTEC 460 TB system and by cultivation on solid media. Clinical specimens (n = 1,500) were pretreated with N-acetyl-L-cysteine (NALC)-NaOH. The contamination rates for MGITs were 2.0% (center 1), 13.8% (center 2), and 6.1% (center 3). A total of 180 mycobacterial isolates were detected (M. tuberculosis complex, n = 113; nontuberculous mycobacteria [NTM], n = 67). When using a combination of liquid and solid media (the current "gold standard" for culture), MGIT plus solid media detected 156 (86.7%) of the isolates, whereas BACTEC plus solid media recovered 168 (93.3%) of all AFB. Between these two gold standards there was no statistically significant difference (P > 0.05). The combination of MGIT plus BACTEC detected 171 (95.0%) of all isolates (compared with MGIT plus solid media, P < 0.01; compared with BACTEC plus solid media, P > 0.05). Considering the efficacies of the different media separately, MGIT was superior to solid media (although not significantly; P > 0.05) in detecting AFB but was inferior to the BACTEC system (P < 0.01). The mean time to the detection of M. tuberculosis complex was 9.9 days with MGIT, 9.7 days with BACTEC, and 20.2 days with solid media. NTM needed, on average, 11.9, 13.0, and 22.2 days to appear by the three methods, respectively. In conclusion, MGIT proved to be a valuable alternative to the radiometric cultivation system.