Pharmacokinetic mismatch does not lead to emergence of isoniazid- or rifampin-resistant Mycobacterium tuberculosis but to better antimicrobial effect: a new paradigm for antituberculosis drug scheduling

Abstract

Multidrug resistant-tuberculosis is a pressing problem. One of the major mechanisms proposed to lead to the emergence of drug resistance is pharmacokinetic mismatch. Stated as a falsifiable hypothesis, the greater the pharmacokinetic mismatch between rifampin and isoniazid, the higher the isoniazid- and rifampin-resistant subpopulation sizes become with time. To test this, we performed hollow-fiber-system studies for both bactericidal and sterilizing effects in experiments of up to 42 days. We mimicked pharmacokinetics of 600-mg/day rifampin and 300-mg/day isoniazid administered to patients. Rifampin was administered first, followed by isoniazid 0, 6, 12, and 24 h later. The treatment was for drug-susceptible Mycobacterium tuberculosis in some experiments and hollow fiber systems with inoculum preseeded with isoniazid- and rifampin-resistant isogenic Mycobacterium tuberculosis strains in others. Analysis of variance revealed that the 12-h and 24-h-mismatched regimens always killed better than the matched regimens during both bactericidal and sterilizing effects (P < 0.05). This means that either the order of scheduling or the sequential administration of drugs in combination therapy may lead to significant improvement in microbial killing. Rifampin-resistant and isoniazid-resistant subpopulations were not significantly higher with increased mismatching in numerous analysis-of-variance comparisons. Thus, the pharmacokinetic mismatch hypothesis was rejected. Instead, sequential administration of anti-tuberculosis (TB) drugs (i.e., deliberate mismatch) following particular schedules suggests a new paradigm for accelerating M. tuberculosis killing. We conclude that current efforts aimed at better pharmacokinetic matching to decrease M. tuberculosis resistance emergence are likely futile and counterproductive.

Fig. 1.

Effect of pharmacokinetic mismatch on total bacillary population during bactericidal activity. There was no significant difference between the killing curves of the perfectly matched and 6-h-mismatched regimens; however, the regimens in which rifampin and isoniazid were given 12 h and 24 h apart killed better.

Fig. 2.

Effect of pharmacokinetic mismatch on the isoniazid-resistant subpopulation. The isoniazid-resistant subpopulation emerged by day 7 and was highest with closer matching. However, this drug-resistant subpopulation was transient.

Fig. 3.

Effect of pharmacokinetic mismatch during sterilizing activity. Bacilli in the untreated control systems grew very slowly, validating the semidormant bacillary population. Killing slopes for the perfectly matched regimen were significantly poorer than those for all mismatched regimens, with the greatest bacillary decrease encountered with 24-h mismatching.

Fig. 4.

Effect of pharmacokinetic mismatch on rifampin-resistant subpopulation. Rifampin resistant subpopulations on day 7 were identified using either the standard critical concentrations or recently proposed breakpoints.