Usefulness of the BACTEC MGIT 960 System for Isolation of Mycobacterium tuberculosis from Sputa Subjected to Long-Term Storage

The transportation of sputa for isolation of Mycobacterium tuberculosis from remote settings to laboratories located abroad usually takes more than 1 week and results in an increased contamination rate and a loss of positive cultures (3, 6). Detection of growth of M. tuberculosis from samples inoculated in solid media such Lowenstein-Jensen medium (LJ) may take several weeks, with a delay in patient treatment. The BACTEC MGIT 960 system (MGIT) is a fully automated, nonradiometric, faster technique with liquid medium that is largely used for M. tuberculosis recovery, but few investigations have been reported on its usefulness for the culture of sputa subjected to long-term storage (1, 3). Sample preservatives such as cetyl-pyridium chloride (CPC) have been shown to be useful for M. tuberculosis isolation on LJ from sputa subjected to long-term storage (1, 6). Using MGIT, we found that the addition of CPC to the sputa did not significantly increase the M. tuberculosis yield (data not shown).

In the present study, the comparison of M. tuberculosis recoveries and contamination rates from long-term-stored sputa without CPC that were inoculated in LJ and MGIT was performed. Sputa were collected by the humanitarian medical aid agency Médecins Sans Frontières at Guliripchi Tuberculosis Hospital, Sukhumi, Abkhazia, a region of the former Soviet Union, from December 2003 to December 2005. Smear microscopy by the Ziehl-Neelsen method was performed on all unconcentrated specimens, and results were reported as smear negative (S−) or smear positive (S+) (range, 1+ to 4+) (6). Sputa were kept refrigerated at +4°C at the hospital of Sukhumi, sent by car in groups of 20 to 30 per parcel to the office of Médecins Sans Frontières at Tbilisi, and finally mailed by an express courier to the Istituto Superiore di Sanità, Rome, for culture; overall, the mean time ± standard deviation between collection and sample processing was 17 ± 7 days (range, 7 to 39 days). Sputa were processed blindly by the N-acetyl-l-cysteine-NaOH method, using a commercial kit (MycoPrep; Becton Dickinson, Cockeysville, MD); the sediment was suspended in phosphate-buffered saline and inoculated in LJ (100 μl) (Biomérieux, Marcy l'Etoile, France) and MGIT tubes (Becton-Dickinson) (500 μl) according to the manufacturer's instructions. The LJ slants were incubated at 37°C in 5% CO₂ and examined weekly for 8 weeks. M. tuberculosis was identified in positive cultures by a DNA probe (Gene Probe, San Diego, CA) (5, 6).

A total of 1,101 sputa, including 344 baseline samples from 261 patients and 757 follow-up samples from 555 patients on antituberculosis treatment, were examined (Table 1). Inoculation of S+ sputa in MGIT significantly increased M. tuberculosis recovery from 50 to 66% and decreased the contamination rate from 33 to 13%, in comparison with inoculation in LJ. In particular, positivity increased from 62 to 82% for baseline samples and from 39 to 50% for follow-up samples; the contamination rate decreased from 35 to 14% for baseline samples and from 30 to 13% for follow-up samples, in comparison with inoculation in LJ. When the combination of LJ and MGIT media was considered, M. tuberculosis recovery increased from 2 to 3% in comparison with MGIT alone.

A similar trend was observed for S− sputa, with M. tuberculosis recovery increasing from 19 to 47% and the contamination rate decreasing from 41 to 13%, in baseline samples, in comparison with LJ. In S− samples collected during the follow-up, culture positivity was higher in MGIT than in LJ, but the difference did not reach statistical significance. The contamination rate for S− sputa was significantly lower with MGIT than with LJ and decreased by 2% or more when the combination of LJ and MGIT media was considered.

Correct collection and transportation of sputa are crucial to ensure accurate M. tuberculosis recovery from remote survey settings (3). Overall, our results indicated that the rate of M. tuberculosis recovery by MGIT, or by LJ plus MGIT, from S+ or S− long-term-stored sputa was clearly higher than that by LJ. The positivity rate by MGIT of our S+ samples was lower than that reported for routinely processed S+ samples cultured in the same medium (7). This can be explained by the knowledge that increasing storage time at room temperature reduces M. tuberculosis viability and increases contamination (4). After-collection samples were kept refrigerated, but ground transportation and mailing from Sukhumi to Rome occurred at room temperature and took several days; this period may have affected M. tuberculosis viability and contamination. Contamination rates similar to those found by us were also reported by other investigators when late arrival of the specimens delayed sample processing (7).

The use of one liquid and one solid medium is recommended by the Centers for Disease Control and Prevention (2). Indeed, when the combination of LJ and MGIT media was considered, overall positivity increased (68%) and the contamination rate decreased (11%). We are aware that comparative studies between long-term-stored and routinely examined samples are difficult because they are affected by several variables. When the data were stratified for baseline and follow-up samples, the positivity rate of S+ baseline samples processed by LJ plus MGIT was 85%, indicating that this method is helpful to increase M. tuberculosis isolation from long-term-stored sputa. MGIT was also very useful for M. tuberculosis recovery for S− sputa, either at baseline or at follow-up. In the baseline samples, M. tuberculosis isolation by MGIT (or LJ plus MGIT) increased from 19 to 47% compared to that by LJ, and in the follow-up the results were similar to those reported by other investigators (8). From the clinical point of view this observation is very important, since S− patients, either at baseline or during follow-up, have a higher chance of being treated with antituberculosis therapy if culture is performed in MGIT plus LJ, in comparison with LJ alone.

Overall, our results indicate that MGIT is a suitable liquid method to culture M. tuberculosis from samples collected in remote areas of the world.

• Copyright © 2007 American Society for Microbiology