Pharmacokinetic/Pharmacodynamic (PK/PD) Modeling for Integrase Inhibitors with a Simple Viral Dynamic Model
Toshihiro Wajima, Ryuji Kubota
Clinical Pharmacology & Pharmacokinetics, Shionogi & Co. Ltd., Osaka, Japan
Objectives: S/GSK1349572, S/GSK1265744 and S/GSK364735 are potent, low nanomolar inhibitors of both recombinant HIV integrase and HIV replication in cell culture assays. Currently in clinical development, S/GSK1349572 and S/GSK1265744 are unboosted, once daily integrase inhibitors with different resistance profiles than raltegravir or elvitegravir. A mathematical representation of viral dynamics for integrase inhibitors combined with a pharmacokinetic model are useful to assess dose-effect and concentration-effect relationships and thus aid in dose selection for clinical studies. The objective was to develop a simple and practical PK/PD model for describing plasma concentration profiles and viral dynamics of integrase inhibitors.
Methods: A simple viral dynamic model was developed. The PK part of the model is a conventional 1 compartment model with first-order absorption, and the PD part consists of 1 compartment for describing viral dynamics with first-order viral depletion and viral count-related viral replication, which is inhibited by integrase inhibitors with a Emax model. The model was applied to the profiles of plasma concentrations and HIV-1 RNA counts from 3-Phase IIa 10-day monotherapy studies. The model was fitted to 3 integrase inhibitors with adjustment by in vitro protein-adjusted IC50 (PA-IC50). Model evaluation was performed using classical diagnostic plots and the visual predictive check.
Results: The simple viral dynamic model described well the profiles of plasma concentrations and HIV-1 RNA counts in short-term monotherapy studies for these integrase inhibitors. The PD profiles for these 3 integrase inhibitors were described with a common virus-related PD parameter (first-order viral depletion constant) using in vitro PA-IC50 for each compound with adjustment by in vitro-in vivo scaling parameter. The first-order viral depletion constant was estimated to be 0.00305 hr-1. The in vivo-in vitro scaling factor was estimated to be 2.26, suggesting in vivo IC50 was higher than in vitro PA-IC50. Simulations suggest that S/GSK1349572 and S/GSK1265744 will have robust efficacy with once daily dosing.
Conclusions: A simple PK/PD model was developed for describing the relationships between PK and PD for integrase inhibitors. This model can be used to predict future clinical trial results for the drugs of interest but can also be used for predicting the PK/PD relationships for other drugs in the same class.