Reply to “Pharmacokinetic Mismatch of Tuberculosis Drugs”

REPLY

We thank Dr. Peloquin (2) for his words of caution concerning our pharmacokinetic mismatch paper (4). We also thank him for the opportunity he raised to further discuss the subject of pharmacokinetic mismatch in the context of tuberculosis treatment. First, we agree with him that our results specifically pertain to the rifampin and isoniazid combination; our intent was to study the role of pharmacokinetic mismatch in the most commonly used regimen for short-course chemotherapy. This is because the pharmacokinetic mismatch hypothesis has been used by others to try and explain the reasons for resistance emergence in the standard short-course regimen (1). Our comments on the futility of pharmacokinetic mismatch were meant to pertain to this particular combination. We acknowledge in the paper that pharmacokinetic mismatching is often said to occur due to differences in half-life; namely, one drug with a short half-life and another with a long half-life. However, at the core of this hypothesis is that the drug with long half-life continues to exert pressure as monotherapy, leading to emergence of resistance to that drug (1, 3). We approached this by deliberately mismatching the schedule of administration to create different levels of overlap (or nonoverlap) of drug concentrations in a manner that could be better quantified, creating progressively different durations of monotherapy while controlling for cumulative drug exposure. Thus, despite the similarity of the half-life of isoniazid and rifampin, the design allowed for periods of monotherapy. We agree with Dr. Peloquin that different results may be obtained with rifabutin and rifapentine, and those studies are ongoing in our experimental systems. Nevertheless, it is interesting that in the clinical studies in which rifamycin monoresistance arose there was intermittent administration of a rifamycin and isoniazid. Intermittent regimens as used in clinical practice unfortunately also lead to reductions in cumulative area under the concentration-time curve (AUC) without advantageously increasing peak concentrations, since there is not enough of a compensatory dose increase of intermittently administered drug in true dose-fractionation fashion. Thus, even for rifamycins that have a long half-life when combined with isoniazid, the explanation for acquired drug resistance, while often attributed to pharmacokinetic mismatch, could equally be due to suboptimal cumulative drug exposures, as predicted by pharmacokinetic-pharmacodynamic studies. The pharmacokinetic mismatch hypothesis in that case would be true only if acquired resistance was greater in intermittent dosing schedules than with a more frequent dosing schedule in a design in which the colinear decrease in drug exposure and dosing schedule were controlled for. This hypothesis still needs to be affirmed or rejected for rifapentine and rifabutin under these controlled experimental conditions.