Sylvain Fouliard and Marylore Chenel
Department of Clinical Pharmacokinetics, Servier, Suresnes, France
Objectives: To develop through a sequential approach a single PKPD model of a surrogate biomarker SB as a function of two independent variables (time=IDV1 and IDV2) after the administration of drug S, describing the effect of drug S and its active metabolite M, consistently with in vitro activity assays and drug mechanism knowledge.
Methods: A combined PK model of drug S and its active metabolite M was first developed using PK data from 25 healthy volunteers being administered drug S as a slow-release formulation. Individual PK parameters for S and M were obtained from the final PK model through Bayesian approach, and used as an input in a PKPD dataset along with PD measurements in order to fit SB as a function of IDV1 (time) and IDV2. A drug/receptor binding kon/koff model was implemented including the influence of S and M on SB. An identifiability analysis was performed using design evaluation software POPDES in order to assess the data's and model's ability to estimate both S and M activities. Comparison was also performed between sequential approach (fit of SB(IDV1), then fit of SB(IDV1, IDV2)) and simultaneous approach (direct fit of SB(IDV1, IDV2)) [1].
Results: The PK model of S and M concentration-time profile after administration of S as an extended-release formulation consisted in a mono-compartmental model for S and M, with a complex absorption (2 depot compartments and a time-dependent absorption rate) and first order elimination for both drugs. Inter-individual and inter-occasion variabilities were estimated on several parameters. A drug/receptor binding modelled the dependence of SB on IDV1 and a direct model described SB variation with IDV2. The identifiability analysis showed that it was not possible to estimate the effect of both S and M on SB thus their relative activity was fixed to a value from an in vitro assay. The sequential approach and the simultaneous approach showed similar results and allowed a satisfactory description of SB(IDV1, IDV2).
Conclusion: A PKPD model successfully described the effect of drug S and its metabolite M on SB with respect to two independent variables and will be used for further PKPD simulations providing a powerful tool in the context of model-based drug development.
References:
[1] Zhang, L., Beal, S. L., & Sheiner, L. B. (2003). Simultaneous vs. sequential analysis for population PK/PD data I: best-case performance. Journal of pharmacokinetics and pharmacodynamics, 30(6), 387-404.
Reference: PAGE 22 (2013) Abstr 2846 [www.page-meeting.org/?abstract=2846]
Poster: Other Modelling Applications