Objective: Pharmacokinetic interactions are expected when human immunodeficiency virus (HIV) protease inhibitors are coadministered because many are both substrates for and inhibitors of CYP3A4. The goal of this model-based pharmacokinetic analysis was to describe the differences observed in amprenavir pharmacokinetics among treatment arms in the Adult AIDS Clinical Trial Group (AACTG) study protocol 398 and to propose mechanisms to account for them.
Methods: One hundred seventy-six HIV-positive subjects receiving 1200 mg amprenavir twice daily as part of AACTG protocol 398 were included in the pharmacokinetic study. All patients also received background medications efavirenz, adefovir dipivoxil, and abacavir and, depending on the study arm, placebo or one of the following protease inhibitors: nelfinavir, indinavir, or saquinavir. A population pharmacokinetic model was fitted to a total of 565 amprenavir concentration measurements. The blood samples for concentration measurements were drawn at week 2 (12-hour pharmacokinetic study, approximately 7 samples per study; 46 patients) and at week 24 (6-hour pharmacokinetic study, approximately 5 samples per study; 10 patients). In addition, samples were collected at 1 or more follow-up visits (population pharmacokinetic study, 1 to 3 occasions per patient; 150 patients).
Results and conclusion: Amprenavir intrinsic clearance was significantly reduced relative to placebo by nelfinavir (-41%) and indinavir (-54%) but not by saquinavir. The absolute magnitude of amprenavir intrinsic clearance suggests that CYP3A4 inhibition by nelfinavir and indinavir is balanced by enzymatic induction in the presence of the background drug(s), most likely efavirenz. Amprenavir intrinsic clearance apparently increases by more than 30% between weeks 2 and 24, possibly because of the time course of CYP3A4 induction.