Does the MGIT 960 System Improve the Turnaround Times for Growth Detection and Susceptibility Testing of the Mycobacterium tuberculosis Complex?

The nonradiometric and fully automated Mycobacteria Growth Indicator Tube 960 system (MGIT 960; Becton Dickinson Microbiology Systems, Sparks, MD) has been proposed as a more efficient alternative to the semiautomated and radiometric BACTEC 460TB system (4, 5). However, it is noteworthy that to date, only two prospective studies (1, 2) have examined the turnaround time (TAT) of the first-generation protocol (the manual 4-ml MGIT) for this assay with respect to both growth detection and susceptibility testing of the Mycobacterium tuberculosis complex (MTBC). Importantly, those studies assessed only a limited number of samples drawn from repositories, rather than actual patient specimens, and only one of the two compared the MGIT with the BACTEC system (1). Similar studies to assess the performance of the fully automated 7-ml MGIT 960 have not yet been reported. Since the health care provider has to make decisions concerning discontinuation of certain drugs or significantly changing the drug regimen upon receipt of susceptibility testing results, it is important that the entire TAT be examined, from specimen collection (or receipt in the laboratory) through reporting of susceptibility results.

A prospective study was performed using 260 consecutive MTBC Fast Track clinical specimens (253 M. tuberculosis, 4 Mycobacterium africanum, and 3 Mycobacterium bovis) in a routine diagnostic setting to determine the TATs of the MGIT 960 system relative to the reference BACTEC 460TB system for growth detection, susceptibility testing, and specimen processing (from specimen receipt to reporting of susceptibility testing results) of MTBC.

The model Fast Track program for tuberculosis was initiated by the New York State Department of Health in 1993 to expedite testing of newly diagnosed smear positive, and therefore highly infectious, patients (3). After decontamination, the following media were inoculated for each specimen: one MGIT 960 tube, one BACTEC 12B vial, and one Lowenstein-Jensen Gruft slant. Growth detection and drug susceptibility testing assays using the automated MGIT 960 system and the reference BACTEC method were performed in line with the recommendations of the manufacturer (4, 6).

Of the 260 specimens, 29 were excluded from TAT calculations either due to nonmycobacterial contamination (20 [7.7%] specimens) or due to being a mixed culture with Mycobacterium avium (9 specimens [3.5%]). Of the remaining 231 specimens, 193 (83.5%) yielded pansusceptible MTBC, and 38 (16.6%) yielded drug-resistant MTBC. There was a statistically significant difference (P < 0.0001) in the mean TAT from specimen receipt to the detection of growth of MTBC between the MGIT 960 (8.6 days; range, 2 to 33 days) and the BACTEC system (6.4 days; range, 1 to 23 days).

An additional 29 specimens were omitted from the comparative analysis, in line with the instructions of the manufacturer, on the basis of delayed (>5 days after growth detection) MGIT 960 susceptibility setup (10 [4.3%] specimens), invalid susceptibility results with the MGIT 960 system (13 [5.6%] specimens), or invalid susceptibility results with the BACTEC 460TB system (3 [1.3%] specimens). In addition, in three (1.3%) specimens, growth was detected more than 4 days earlier in the MGIT 960 system than in the BACTEC system. For ethical reasons, MGIT 960 aliquots were transferred into the BACTEC susceptibility testing system to avoid further delays. MGIT 960 susceptibility testing produced an invalid result for one isolate of M. africanum (25%) and for one isolate of M. bovis (33%). These numbers are not significant, and additional research studies are clearly warranted to determine the suitability of the MGIT for detection of all members of the MTBC.

For the remaining 202 specimens, the mean TATs for susceptibility testing (from setup through reporting of susceptibility results) were 7.3 days (range, 5 to 12 days) with the MGIT 960 system and 4.3 days (range, 3 to 10 days) with the BACTEC 460TB system (P = 0.015, a significant difference). The mean TATs from specimen receipt to reporting of susceptibility results were 17.9 days (range, 10 to 47 days) with the MGIT 960 system and 14.6 days (range, 5 to 34 days) with the BACTEC system (P < 0.0001, a significant difference) (Fig. 1).

FIG. 1.

The number and distribution of mycobacterial isolates having total turnaround times (from specimen receipt through reporting of susceptibility results) either longer and shorter than the mean turnaround time ± standard error for the MGIT 960 and BACTEC 460TB systems.

The guidelines of the Centers for Disease Control and Prevention recommend that the TAT for susceptibility testing of MTBC be 2 to 4 weeks following receipt of the specimen (7). In the present study, this recommended TAT could be achieved for 75.0% of the 260 Fast Track specimens with the MGIT 960 system and for 76.5% of the specimens with the BACTEC 460TB system. Thus, although the MGIT 960 system brings increased safety, new technology, and more economical use of labor, it does not introduce any significant improvement that would benefit patient treatment compared to the radiometric BACTEC method. Therefore, we would encourage the manufacturer to modify the MGIT 960 system to produce shorter TATs, while retaining the advantages of its more efficient and safety-minded design. This, however, is likely to be a short-term gain, given that new goals outlined in Healthy People 2010 have proposed a detection time of 2 days for 75% of all tuberculosis cases (9, 10). It is imperative, therefore, that the clinical mycobacteriology laboratories pursue more-sensitive nucleic acid amplification assays in conjunction with molecular drug susceptibility assays that are performed directly on the clinical specimens. Such approaches will inevitably replace the growth detection-based systems (6, 8).