Abstract

The general workflow of PBPK modeling and simulation of test compounds in normal patients and patients with genetic variations consisted of the following steps (see Fig. 1). First, the PBPK model was initially developed using the physicochemical, biopharmaceutical, and pharmacokinetic parameters obtained from literature or estimated by ADMET Predictor(Simulation Plus Inc.); simulated mean plasma concentration-time profiles were compared with those obtained from clinical studies including single and multiple oral dosing. Population-dependent physiological parameters in human PBPK models were obtained using the Population Estimates for Age-Related Physiology™ module in GastroPlus, where the generic demographic and physiological parameters (e.g., age, tissue volume, and perfusion rate) for a normal American population were based on the National Health and Nutrition Examination Survey as described previously [1]. Each simulation trail included 50 virtual subjects with demographics in terms of age, ethnicity, and percentage of women similar to those reported in the corresponding clinical trails [2-12].Human pharmacokinetics profiles of test compounds were taken from literature by scanning with Get Data Graph Digitizer version 2.26 and were then used in the modeling. Second, the drug-specific parameters underwent refinement if predicted PK profile and parameters deviate significantly (<0.8-fold or >1.2-fold) from the observed data. Third, the PBPK model was used to predict the plasma concentration-time profiles and PK parameters in individuals with genetic variations by incorporating these alterations and compared to observed genetic polymorphisms data. Fourth, the PBPK model was linked to a PD model to evaluate the relationship between brain drug concentration and drug effect (Figure 1).

By Saeed Alqahtani, Amal Kaddoumi