Comparison of MAS-PCR and GenoType MTBDR assay for the detection of rifampicin-resistant Mycobacterium tuberculosis.

SETTING: Pham Ngoc Thach Hospital for Tuberculosis and Lung Diseases, the tertiary referral hospital for tuberculosis (TB) in Southern Vietnam. OBJECTIVE: To develop and evaluate a simple, rapid and accurate multiplex allele specific polymerase chain reaction (MAS-PCR) test to detect rifampicin (RMP) resistance point mutations at codons 516, 526 or 531 in the rpoB gene of Mycobacterium tuberculosis. DESIGN: The novel MAS-PCR was compared with the commercial M. tuberculosis Drug Resistance (MTBDR) test in 104 RMP-resistant and 50 RMP-susceptible routine isolates, defined by conventional 1% phenotypic susceptibility testing. RESULTS: The sensitivity of the MAS-PCR and MTBDR tests was respectively 83.7% (95%CI 75.1-90.2) and 93.3% (95%CI 86.6-97.3). Both tests were 100% specific. The negative predictive value was 74.6% (95%CI 65.3-83.1) for the MAS-PCR and 87.7% (95%CI 80.0-93.6) for the MTBDR test. CONCLUSION: The MTBDR test, although more sensitive, is currently prohibitively expensive in resource-poor, high-burden settings. The MAS-PCR described here presents a less laborious economic alternative. A susceptible result returned by either test cannot be used to exclude multidrug-resistant TB.

Fluoroquinolone resistance detection in Mycobacterium tuberculosis with locked nucleic acid probe real-time PCR.

SETTING: Pham Ngoc Thach Hospital for Tuberculosis and Lung Diseases, Ho Chi Minh City, Vietnam. OBJECTIVE: Fluoroquinolones (FQs) are increasingly used in the treatment of tuberculosis (TB) and are the second-line drugs of choice for treatment of multidrug-resistant TB. We aimed to set up a polymerase chain reaction (PCR) based assay to detect the most common FQ-resistance-associated mutations in gyrase A (gyrA) of Mycobacterium tuberculosis. DESIGN: A total of 42 FQ-resistant and 40 FQ-susceptible isolates were collected in 2005-2006 and sequenced in gyrA. Using sequencing results as gold standard, a real-time PCR using three locked nucleic acid probes (LNA-PCR) was designed to detect mutations at positions 90, 91 and 94 (97% of gyrA FQ-resistance-associated mutations) and evaluated. RESULTS: Sequencing of 42 FQ-resistant isolates revealed no gyrA mutations in 10 isolates, 20 isolates had a single mutation and 12 isolates showed double peaks at resistance-associated alleles, suggesting a heterogeneous population. With LNA-PCR, all wild-type and 19/20 mutant isolates were correctly identified. Eleven of 12 heterogeneous isolates were correctly identified as resistant mutants. Overall, 71% ([19 + 11]/42) of phenotypically FQ-resistant isolates were detected. Specificity was 100% on 40 FQ-susceptible isolates. CONCLUSION: This assay provides a simple and rapid means to reliably detect FQ-resistance-associated gyrA mutations in M. tuberculosis.

Molecular analysis of Mycobacterium tuberculosis causing multidrug-resistant tuberculosis meningitis.

SETTING: Tertiary referral hospitals in southern Vietnam. OBJECTIVE: Molecular characterisation of multidrug-resistant (MDR) tuberculous meningitis (TBM). DESIGN: Mycobacterium tuberculosis isolates from the cerebrospinal fluid (CSF) of 198 Vietnamese adults were compared with 237 isolates from patients with pulmonary tuberculosis (PTB) matched for age, sex and residential district. Isolates resistant to isoniazid or rifampicin (RMP) were sequenced in the rpoB and katG genes, inhA promoter and oxyR-ahpC intergenic regions. RESULTS: While drug resistance rates were lower in the CSF (2.5% MDR) than pulmonary isolates (5.9% MDR), the difference was not significant. The most commonly mutated codons were 531, 526 and 516 in rpoB and 315 in katG. Four novel triple mutants in rpoB were identified. CONCLUSION: RMP resistance is a good surrogate marker for MDR-TBM in this setting. However, probes directed against these three codons would have a maximum sensitivity of only 65%. A rapid phenotypic detection test may be more applicable for the diagnosis of MDR-TBM.