Rapid prediction of multi-drug resistance in clinical specimens of Mycobacterium tuberculosis.

Abstract

Conventional drug susceptibility testing techniques, the ‘gold standard’ for M. tuberculosis are slow, requiring about 3-6 weeks from a positive culture. This diagnostic delay, before initiation of appropriate treatment, contributes to increased transmission rates. Molecular techniques provide rapid results and therefore present an alternative to conventional tests. The aim of this project was to develop an inhouse reverse line blot hybridization assay (RIFO assay) that could detect mutations associated with Rifampicin resistance directly in clinical specimens of patients in KwaZulu Natal. A 437 bp region of the rpoB gene was sequenced to ascertain the most frequently occurring mutations conferring resistance to rifampicin in isolates in KwaZulu-Natal. Wildtype and mutant probes designed to target these mutations, were immobilized on a Biodyne C membrane. Hybridization conditions were optimized using biotin labeled PCR products from culture. Detection was performed with peroxidase labeled streptavidin using enhanced chemiluminescence. Four DNA extraction methods were evaluated on sputum specimens to determine the one with the least inhibitory effect on amplification. A total of 11 mutations were found in 236 clinical isolates: 531TTG (109, 58.3%), 516GTC (26, 13%), 533CCG/516GGC (20, 10%), 533CCG (18, 9.6%), other mutations < 5% each. The chelex extraction method was found to be optimal for removing inhibitors in sputum specimens. Sputum specimens of 404 patients hospitalized at King George V Hospital between 2005 and 2006 were rifoligotyped. The RIFO assay was optimised on clinical isolates and then applied to sputum specimens. The RIFO assay on culture and sputum correlated well with the DST (sensitivity 92% and 94% respectively). However, the specificity was very low in both culture and sputum specimens compared to DST (38% and 35% respectively). This could be attributed to the presence of silent mutations, mixed infections, mixed populations of bacteria or the small number of susceptible strains used in this study. The in-house RIFO assay can be used directly on sputum specimens to predict Rifampicin resistance and therefore MDR-TB in less than a week compared to the gold standards. A total of 43 samples can be tested simultaneously at low cost and the membrane is reusable compared to commercial kits such as the Hains test that is expensive and strips are not reusable. A similar assay can be designed to target mutations for the detection of XDR-TB. Future studies should be conducted in a clinical setting on patients with sensitive strains to increase the specificity.