Abstracts

Rationale: The objective of this study was to assess, for drugs that auto-induce their own metabolism, whether the conclusions obtained from single dose bioequivalence (BE) studies in healthy subjects are valid for patients whose metabolizing enzymes are induced after chronic administration using population PBPK simulation.Methods: Carbamazepine (CBZ), a strong metabolizing enzyme inducer, was used in the current population PBPK simulation case study. A mechanistic model incorporating in vivo enzyme induction profile based on in vitro data was built using the mechanistic dynamic model within Simcyp population-based simulator. The model was validated by comparing the model-predicted with the observed plasma concentrations of CBZ and its metabolite. Thereafter, single dose (2-way crossover and 4-way crossover studies) and multiple dose (2-way crossover studies) BE studies in healthy and patient populations were simulated. Test products had the same formulation design as the reference product, i.e. controlled released formulation, but had differences in the rate of release or fraction absorbed compared with reference products. The log-normal statistical distributions of intrasubject variabilities were incorporated in the pharmacokinetic parameters. As CBZ has already been classified as a narrow therapeutic index drug (NTI), the simulated 4-way crossover studies were assessed by the approach of reference scaled average BE (SABE), and the unscaled average BE approach (ABE) (90% CI of 80-125%) was applied for simulated 2-way crossover studies. The sensitivity to detect formulation differences was assessed by comparing the passing rate of Cmax and AUC in each study using SABE or ABE criteria.Results: It was shown that the probability of BE failure for the Cmax (sensitive to the changes in the release rate and the fraction absorbed) and AUC (sensitive to the changes in the fraction absorbed) in 2-way crossover single dose studies was always greater than that in 2-way crossover multiple dose studies in healthy population. Moreover, the passing rate of Cmax in a 4-way crossover single dose study with SABE criteria in healthy population is more sensitive to the changes in the release rate compared with a 2-way crossover multiple dose study with ABE criteria in patient population.Conclusions: The population PBPK simulation results suggest that single dose BE study in healthy subjects is more sensitive assessment of bioequivalence than multiple dose BE study in patients.