Selecting drug combinations for treatment of drug-resistant mycobacterial diseases.

Mixtures of antituberculosis drugs were evaluated for their in vitro effects on drug-resistant isolates of Mycobacterium tuberculosis and M. avium-intracellulare. The response of individual isolates to representative drug combinations was not always predictable from the results of single-drug sensitivity assays. For the case of M. tuberculosis, combinations of drugs were often bactericidal even under conditions where two or more drugs were without effect when tested singly. The more widely drug-resistant M. avium-intracellulare demonstrated increased growth inhibition when subcultured in the presence of single drugs, particularly rifampin and streptomycin. However, these conditions favored the selection of highly resistant strains. Alternatively, multiple drugs were often bacteriostatic; and under conditions where isolates demonstrated growth inhibition, the selection of highly drug-resistant strains was delayed. These results suggest a role for multiple-drug sensitivity assays in selecting drug combinations to be used in the treatment of drug-resistant mycobacterioses.