Model-Based Approach for a New Prolonged-Release (PR) Formulation of Ivabradine: PK and PKPD Assessment to Support Bioequivalence
Myra Skerlep1, Enrica Mezzalana3, Marylore Chenel3, Nathalie Gendrot2, Sylvain Fouliard1
1Quantitative Pharmacology, Translational Medicine; Servier, 2Cardiovasc & Metabolic Disease Therapeutic Area; Servier, 3Pharmetheus
Introduction: Ivabradine is a specific HR-reducing agent, named If inhibitors, that was developed for the treatment of chronic heart failure and stable angina. Ivabradine is currently available as immediate release (IR) tablets designed for twice daily (bid) dosing. Therefore, the introduction of a prolonged release (PR) formulation, which requires once daily (qd) dosing may have the potential to reduce plasma drug concentration fluctuations and to enhance patient adherence. Pharmacometric modeling has been proposed as a tool to compare the pharmacokinetics (PK) and pharmacodynamics (PD) of ivabradine and its active metabolite after administration of IR or PR formulations in the pivotal phase I study. The objectives of this work were: -to characterize the population pharmacokinetics (popPK) of ivabradine and its metabolite (S18982) after single and repeated administrations of both PR and immediate release (IR) tablet formulations, -to identify the relationships between the pharmacokinetics of ivabradine and its active metabolite (S18982) and heart rate (HR) following repeated administrations of these formulations, - to assess the bioequivalence (BE) between the PR and IR tablet formulations, both in terms of PK exposure at steady state and in terms of PKPD, as mean HR change over 24h from baseline. Methods: The data for this analysis originated from a completed 2-way cross-over phase I study, with PK and PD data available from 37 and 36 healthy volunteers, respectively. A sequential modeling approach was first performed to estimate the parent and then the metabolite parameters. In the final model, ivabradine parameters were estimated simultaneously with those of its metabolite. A joint parent-metabolite population PK (popPK) model was developed to characterize the PK of ivabradine and its active metabolite (S18982) in adult healthy subjects following the administration of o.d. PR formulation and b.i.d. IR formulation. Additionally, a popPK/PD model was developed to characterize the relationships between the PK of ivabradine and its metabolite (S18982) and the change in HR, after repeated administrations of both PR and IR tablet formulations. The body weight was included a priori on all the clearance and volume parameters using allometric scaling for the popPK analysis. The evaluated covariates included formulation and subject characteristics (age, sex) for the PK and the PKPD. Potential exploratory covariate-parameter relationships were evaluated using the stepwise covariate model (SCM) building procedure with adaptive scope reduction (ASR). The population modeling analyses were performed using the first-order conditional estimation method with interaction (FOCEI) method in NONMEM. Results: For the PK, a two-compartment model with first-order elimination best described the disposition of ivabradine and its metabolite. For IR formulation, absorption was characterized by a first-order absorption with an effect of morning versus evening dose on mean absorption time. A mixture model was estimated to split the subjects into two different populations: one with typical lag-time (tlag) and one showing longer tlag after the IR evening dose. For the PR formulation, a parallel zero- and first-order absorption model was used. A full transformation of ivabradine into its metabolite was considered for both formulations. Concerning the popPKPD, the baseline model was characterized by a constant baseline and a 24 hour circadian rhythm model, and the drug effect was modeled as a competitive inhibitory model due to the effect of parent and metabolite concentrations with a delayed effect. An effect of sex on constant baseline and of formulation on the EC50 parameter of ivabradine were found with 12.3 % higher HR baseline in female subject and the PR formulation having a 37 % lower EC50 compared to the IR formulation. The modeling supported the analysis of the potential deviation between formulations, through the computation of AUC0–24ss and the mean HR change from baseline over 24 hours. Simulations were performed based on the final parent-metabolite PK model and the final PK/PD model and allowed to support the results from the bioequivalence study between IR and PR formulation in term of PK and PD. Conclusion: This work underlines the importance of modelling and simulations for the analysis of clinical studies for development of new formulations especially in a complex setting where PD readout is impacted by an active metabolite.