Protocol Synopsis

While developmental pharmacokinetic (PK) and pharmacodynamic (PD) studies are routinely performed for new drugs, there are few clinical PK and PD studies of currently licensed tuberculosis (TB) drugs.  Notably, detailed knowledge of PK/PD relationships can markedly accelerate drug development for antimicrobial agents.  For example, because of information on the PD of antiretroviral drugs and the availability of a convenient surrogate endpoint, Phase 2 trials of new protease inhibitors can be performed using as few as 20 patients per arm.  Thus a better understanding of PD relationships of antimycobacterial drugs could accelerate the difficult task of going from many possible combinations of drugs, doses, and drug combinations to the few that can be evaluated in large Phase 3 studies.

Standard PK studies often include multiple sampling time points over 24 (or more) hours to determine AUC and peak concentration.  However, such dose-ranging studies of PK are intensive and unsuitable for critically ill patients who are much more likely to have highly abnormal PK that may be associated with differences in treatment outcome.  The analytic technique of population PK modeling is well-suited to the study of the PK and PD of TB drugs.  In population PK, models are fitted to all available data simultaneously instead of fitting models to each individual’s data.  This makes the population approach useful for sparse data sets, for which it is impossible to use the more traditional analytical approaches.  Also, population PK modeling is a powerful tool for studying the effects of key covariates on PK like HIV infection, degree of immunodeficiency, or the presence of liver disease.  The cohort of patients in TBTC Study 29X will be studied by either single specimen sampling per the parent Study 29X protocol or sparse sampling (3 samples) or intensive sampling (7 samples) as described in this sub-study protocol to support the construction of a rifapentine population PK model.  This 29X PK sub-study with intensive sampling of up to 80 patients (48 to 60 patients with rifapentine treatment and 20 controls with rifampin) will provide initial estimates for the iterative 2-stage Bayesian (IT2B) parametric analysis of the population PK model. We anticipate that approximately 70 patients will enroll for sparse sampling and  up to 110 additional patients will provide just a single PK sample in the parent Study 29X protocol.  Thus, the population PK model will include data regarding rifapentine in up to 240 patients. 

The primary objective of this study is to characterize rifapentine and 25-desacetyl rifapentine pharmacokinetic parameters (AUC_0-24 and peak concentration) in patients with TB treated with 10, 15 or 20 mg/kg rifapentine daily.