Non-Linear Population Pharmacokinetic Model for Otamixaban in Healthy Male Subjects
Paccaly, A (1), Speth, H (2), Jensen, BK (1), Maas, J (2)
Sanofi-Aventis Science & Medical Affairs, MPK (1) Bridgewater NJ USA and (2) Franfurt Germany
Objectives: Otamixaban is a direct Factor Xa inhibitor for the intravenous treatment of arterial thrombosis. Healthy subjects (Phase I) showed a more than dose-proportional plasma exposure with increasing doses over a 100-fold dosing range. This phenomenon could be possibly related to a partial saturation of the biliary elimination and/or metabolism of otamixaban, while renal excretion is maintained. This work describes the elaboration of a population pharmacokinetic model with Michaelis-Menten elimination that accounts for the non-linear pharmacokinetics of otamixaban in healthy subjects
Methods: Otamixaban was administered intravenously to healthy male subjects (6 per dose group, 14 dose groups) as 6h- or 24h-infusions (1.7 to 183 mcg/kg/h) with- or without a bolus dose (60 or 140 mcg/kg). Otamixaban plasma concentrations at 13 to 23 pre-defined sampling times were measured by LC/MS/MS. Nonlinear mixed effects model analysis was performed using NONMEM/S-PLUS programs in a Linux Cluster environment. FO and FOCE-1 estimation methods were subsequently used throughout the model building process. Two- and three-compartment models with proportional error, using ADVAN3 and ADVAN11 subroutines, respectively, were built and converted into differential equations (ADVAN6). A $DES block was added that accounted for a dual elimination process with a saturable (Michaelis-Menten) (CLS) and a non-saturable (CLR) component. Final model selection was based on the minimal value of objective function (MVOF) and on the goodness of fit derived from the model parameter values and diagnostic plots.
Results: A three-compartmental model (FO) with proportional error and dual elimination was retained. This model proved to be superior to a two-compartment model (FO) as indicated by the difference in objective function DOF=1140. The dual elimination process with a Michaelis-Menten component improved the three-compartment model fit (DOF=233). The mean (CV%) values for the main parameters were V=13.3 L (6.1%), CL=18.6 L/h (24.7%), CLS=22.9 L/h (19.9%), CLR=8.99 (5.7%) L/h, C50=372 (29.0%) ng/mL, KM=1.72 h-1.
Conclusions: This approach illustrates the integration of a plausible physiologic mechanism into the population pharmacokinetic model of otamixaban, namely a partial saturation of the biliary elimination and/or metabolism, as an attempt to better understand and describe the time course of the plasma concentrations with increasing doses in healthy subjects. A three compartment non-linear PK model with dual elimination was retained based on the lowest VOF. However, a less complex model might also be suitable and is preferred for future drug development.
Reference: Paccaly A, Frick A, Rohatagi S, Liu J, Shukla U, Rosenburg R, Hinder M, Jensen BK. J Clin Pharmacol 2006; 46: 37.