Abstract

Background: Several methods have been proposed for rapid detection of drug resistance of Mycobacterium tuber-culosis (M. tuberculosis). High resolution melting (HRM) curve analysis has been developed for accurate and si-multaneous detection of resistance of M. tuberculosis isolates to rifampicin (RIF) and isoniazid (INH). Other tech-niques include the resazurin microtiter assay (REMA) which is one of the new colorimetric phenotypic methods. The present study evaluated the HRM curve analysis and REMA assay in comparison to the proportional method (PM) for rapid identification of multidrug resistant (MDR) M. tuberculosis isolates.
Methods: Thirty M. tuberculosis clinical isolates of known resistance phenotypes were used. An HRM curve was generated for each isolate to scan for mutations in the rpoB and katG genes to detect RIF and INH resistance, re-spectively. The REMA colorimetric assay was also evaluated using the same isolates. The results of both tech-niques were compared to the gold standard proportional method.
Results: The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the REMA assay for RIF and INH susceptibility testing were 100%. The HRM curve assay results for RIF sus-ceptibility testing were 92.3%, 100%, 100%, 94.4%, and 96.7%, respectively, and for INH they were 85%, 100%, 100%, 76.9%, and 90%, respectively.
Conclusions: REMA is a rapid non-conventional and inexpensive method and may serve as a replacement for the proportion method in resource limited settings, while the PPV and NPV of the HRM curve make this assay an ide-al screening method for the TB laboratory.

DOI: 10.7754/Clin.Lab.2012.121121