Comparison of different methods of assessing in vitro resistance of Mycobacterium tuberculosis to rifampicin.

ABSTRACT

BACKGROUND & OBJECTIVES: Definitions of in vitro resistance to rifampicin in strains of Mycobacterium tuberculosis by different methods have not been consistent, leading to variations in the interpretation and validity of results. This study compared three methods of defining in vitro resistance to rifampicin. METHODS: (i) A total of 598 clinical isolates of M. tuberculosis were concurrently compared by the minimal inhibitory concentration (MIC) and the proportion method on Lowenstein-Jensen medium; (ii) 54 strains tested by the MIC method were retested by the proportion method and the BACTEC radiometric method; and (iii) 72 strains which yielded an MIC of 64 mg/l by the MIC method were retested by the same method. RESULTS: Out of 598 cultures tested by the MIC and the proportion methods, identical classification as susceptible or resistant was observed in 99.7 per cent. A 100 per cent agreement was observed when 54 strains were tested by the MIC, proportion and BACTEC radiometric methods. When 72 strains with an MIC of 64 mg/l were retested by the same method, 61 (85%) yielded a lower MIC, 9 (12%) gave the same MIC while 2 (3%) yielded a higher MIC of 128 mg/l, reflecting perhaps the inherent limitations of the variations in the inoculum size. INTERPRETATION & CONCLUSION: All 3 definitions of resistance, viz., an MIC of 128 mg/l, a proportion of 1 per cent or more on 40 mg/l by the proportion method, both on L-J medium and a growth of 1 per cent or more on 2 mg/l by the radiometric method were found to be equally satisfactory.