Conventional and radiometric drug susceptibility testing of Mycobacterium tuberculosis complex

Abstract
A recently developed method of drug susceptibility testing of M. tuberculosis which measures the evolution of labeled CO2 from [1-14C]palmitic acid (BACTEC 460 system) was compared to 3 conventional methods. The proportion method of drug susceptibility testing was the standard against which all test results were compared. Indirect drug susceptibility to isoniazid, streptomycin, rifampin and ethambutol of 245 isolates belonging to the M. tuberculosis complex was determined. In 95% of the cases, results obtained by the radiometric method were available within 1 wk, as opposed to 3-6 wk needed in conventional methodology. Overall agreement was 96.4%. Specificity values were 0.98-1.0; sensitivity values of 1.0 for rifampin, 0.96 for streptomycin, 0.91 for isoniazid and 0.18 for ethambutol were obtained. The specificity of the absolute concentration and resistance ratio drug susceptibility testing methods were 0.99 and 1.0, respectively. The sensitivity of the former was higher than that of the radiometric method (0.99 vs. 0.92); that of the latter was lower (0.88 vs. 0.96). Further testing indicated that the low sensitivity determined for ethambutol may be due to the choice of the critical concentration used, rather than to a shortcoming of the procedure. The radiometric method thus does not significantly differ in reliability from conventional methods of drug susceptibility testing of M. tuberculosis.