Xavier Nicolas(1), Clemence Rauch(1), Eric Sultan(3), Celine Ollier(2) and David Fabre(1)
(1)Modeling and Simulation Entity, (2)Pharmacokinetic Entity, (3)Dedicated Project Expert, Drug Disposition Domain, Disposition, Safety and Animal Research, Sanofi-Aventis R&D Montpellier, France
Objectives: – Understand the differences observed in the pharmacokinetic analysis (steady state and accumulation assessments) of a repeated dose administration of a compound with bi-phasic disposition and long terminal t1/2 in a pool of healthy young and elderly subjects using NCA (Non Compartmental Analysis) and PopPK analysis approaches. – Show NCA limitation for estimation of steady state and accumulation.
Methods: Data were collected in two clinical studies (Phase I). The data Set was composed of 52 subjects (1591 samples) for who age ranged from 18 to 79 years. The Non Compartmental Analysis was performed with WinNonLin software. The PopPK model was developed and validated with the NONMEM computer program (version VII) running on LINUX. Steady state achievement was determined graphically, based on individual and mean Ctrough, for NCA and by simulation using individual parameters for PopPK.
Results: The structural PK model was a bi-compartment model with a lag time (LAG, h-1), an absorption constant (Ka, h-1), characterizing the first-order absorption process from the depot to the central compartment which was described by an apparent central distribution volume (V2/F, L). The peripheral compartment was related to the central one by an inter-compartmental clearance (Q/F, L/h) and described by an apparent distribution volume (V3/F, L). The elimination was characterized by a first-order process and described as an apparent clearance (CL/F, L/h). Terminal t1/2 estimated by NCA and PopPK were consistent (300-600 hrs). Results regarding steady-state achievement show inconsistency between the two approaches, 12 days for NCA vs. 33 (young) to 77 (elderly) days for PopPK. With Boxenbaum’s formula (based on RAC AUC), t1/2eff was 41 hours compatible with steady-state achievement predicted graphically. When the t1/2eff was computed with the t1/2 value and the contribution of each phase (Rowland), the t1/2eff of 450 hours obtained was compatible with results obtained by PopPK simulation.
Conclusions: NCA has a limited use to characterize steady-state achievement for compound combining both poly-exponential disposition and long terminal t1/2. The PopPK model and t1/2eff formula based on t1/2 and contribution of each phase works well. Anyway the appropriate solution is the simulation based on individual PopPK parameters, which takes into account all the PK data available, including those of the disposition phases after the last dose.
Reference: PAGE 22 () Abstr 2958 [www.page-meeting.org/?abstract=2958]
Poster: Model evaluation