Clinical trial simulations to design a crossover study assessing the equivalence on the pharmacodynamic surrogate marker between an immediate and a modified release formulations
Anne Dubois (1), Marylène Gaignet (1), Jacques Thuet (1), Marylore Chenel (1)
(1) Department of Clinical Pharmacokinetics, Institut de Recherches Internationales Servier, Courbevoie, France
Objectives: After the marketing authorisation of a b.i.d. immediate release formulation (IR) of drug X, Servier developed three o.d. modified release formulations (MR) in order to improve patient compliance. A study comparing the pharmacodynamic surrogate marker (SM) between IR and at least one of the three MR has been planned. Modelling & Simulation techniques are playing a key role in designing this study in selecting the MR formulation(s), the equivalent doses of IR and MR, and the measurement times.
Methods: Plasma concentrations of the parent drug and its active metabolite, obtained from a crossover trial performed on 14 healthy volunteers (HV) in fasted conditions, were used to jointly model the IR and each MR formulation. The model assumes the same disposition for IR and MR. It takes account of the hepatic first pass effect and the transformation of the parent drug into metabolite. For the pharmacodynamics (PD), an agonist Emax PK/PD model was previously developed in HV to describe the time course of effect of the IR parent drug and its metabolite on SM. For each MR, we used the PK parameter estimates of the joint IR/MR PK model and the PD parameter estimates of the previously developed PK/PD model to simulate crossover trials on 24 HV at steady state. We simulated 100 trials with different doses. For each simulated trial, we performed an equivalence test on the daily mean SM assuming an equivalence interval of +/-2, 3 or 5 SM units (SMu). For each MR, each tested dose and each equivalence interval, we computed the percentage of trials where the equivalence is demonstrated (Peq). This analysis was first performed assuming no food effect on the IR and MR PK and using a 28-measurement design. As a food effect was found on the IR but not on the MR, we also simulated with a food effect for the IR. At last, the impact of different measurement designs on Peq was evaluated. We used NONMEM VI and SAS 9.1 for this analysis.
Results: Based on both simulations results and clinical relevance, the equivalence interval was fixed to +/-3 SMu. For this value, when an IR food effect was simulated with the 28-measurement design, Peq was above 80% for the 3 MR formulations. Results were similar for all tested measurement designs.
Conclusion: The present clinical trial simulations were determinant to design the PD equivalence crossover study as its results allowed to choose the equivalence interval, the MR formulation, the equivalence dose and the measurement design.