Characterization of the population pharmacokinetics of UK-369,003 using a semi-mechanistic model
N.H. Prins (1), A. Cleton (2), M.O. Magnusson (2) , P. Johnson (2), A. Heatherington (2)
(1) Pharsight, a Certara company, St. Louis, MO, USA; (2) Pfizer Ltd, Clinical Pharmacology And Pharmacometrics, Sandwich, UK
Objectives: Repeated dosing of a PDE5-inhibitor, UK-369,003, indicated that the plasma exposure dose-dependently decreased to equilibrate between 1 and 2 weeks after start of treatment. This suggested auto-induction of clearance, although the mechanism by which (hepatic or gastrointestinal clearance) was unclear. A semi-mechanistic model was developed to characterize UK-369,003 pharmacokinetics (PK) following single dose intravenous (IV) administration and single and multiple doses of a modified release (MR) formulation by determining if and to what extent the dose-disproportional PK was due to hepatic and/or gastrointestinal enzyme induction.
Methods: The PK data set consisted of 6602 plasma concentrations from 610 subjects from 7 dense PK sample phase I studies (3463 concentrations) and 2 sparse PK sample phase II studies (3139 concentrations). A semi-mechanistic model allowing partitioning of induction of clearance between the gut wall and liver clearance [1] was fitted to the data in NONMEM VI. The PK model featured a tablet release into the depot compartment followed by a first-order absorption process into the liver compartment, where the fraction absorbed was estimated. The liver compartment was linked to the central (sampling) compartment with a bi-directional blood flow and a peripheral compartment completed the 2 compartment disposition model. The intrinsic hepatic clearance was assumed to be proportional to the metabolic enzyme amount in a turnover model, where a power model described the relationship between dose and enzyme production rate. Interoccasion variability was allowed for the absorption component of the PK model.
Results: The model fit suggested that the induction of intrinsic clearance was entirely hepatic and neglible contribution of gut wall induction was observed. The half-life of the enzyme induction was estimated at 64 hr, which is consistent with half lives of ~70 hr reported in the literature [1,2]. UK-369,003 metabolism in the liver was estimated to approximately proportionally increase with dose, and resulted in a near doubling of hepatic clearance at 200mg as compared to 10mg at steady state. Age was found to decrease intrinsic clearance (~10% reduction for a 80-yr old compared to a 20-yr old) albeit this relationship came with substantial uncertainty.
Conclusion: UK-369,003 plasma concentrations across time and dose were adequately described by a semi-mechanistic model, which includes a dose-dependent auto-induction of hepatic clearance.
References:
[1] Magnusson, M.O., Dahl M-L, et al. Clin. Pharmacol. Ther. 84, 52-62 (2008)
[2] Ghanbari, F. et al. Curr. Drug Metab. 7, 315-334 (2006).