Evaluation of an integrated in vitro-in silico PBPK (physiologically based pharmacokinetic) model to provide estimates of human bioavailability.

PK express module is a physiologically based model of first pass metabolism, which integrates in vitro data with an in silico physiologically based pharmacokinetic (PBPK) model to predict human bioavailability (F(H)). There are three required inputs: FDp (Fraction dose absorbed, final parameter from iDEA absorption module), protein binding (fu) and disappearance kinetics in human hepatocytes. Caco-2 permeability, aqueous solubility (at multiple pH's), estimated dose and chemical structure are inputs required for the estimation of FDp (Norris et al., 2000; Stoner et al., 2004) and were determined for all compounds in our laboratory or obtained from literature. Protein binding data was collected from literature references and/or Pfizer database. Human hepatocyte data was generated in-house using an automated human hepatocyte method (using Tecan Genesis Workstation) as described previously (). Sixteen compounds (commercial and Pfizer compounds) were chosen to evaluate the PK express model and the bioavailability predicted from the module was compared with known clinical endpoints. For majority of the 16 compounds (approximately 80%), the PK express model F(H) values were comparable to the known human bioavailability (F(H)) (within 23.7 units of the known human (true) F, except for PF 3, PF 4, PF 6). In conclusion, the PK express model integrates a number of key readily available discovery parameters and provides estimates of human performance by integrating in silico and experimental variables built on a physiological based pharmacokinetic model. Information from this model in conjunction with other ADME data (e.g., P450 inhibition) will enable progression of most promising compounds for further in vivo PK and/or efficacy studies.

Validation of a semi-automated human hepatocyte assay for the determination and prediction of intrinsic clearance in discovery.

An automated high throughput human hepatocyte assay has been established with a 96-well format using a Tecan Genesistrade mark Workstation. Validation of this assay was performed with nine commercially available compounds and an additional 10 Pfizer compounds with varying hepatic extraction ratios (E(H)) ranging from 0.02 to approximately 1. The incubation conditions in the automated assay are readily and precisely controlled and cell viability of over 80% was achieved in the automated assay further confirming its utility for absorption, distribution, metabolism, and excretion (toxicity) (ADME (T)) screening. The results of the nine commercial compounds correlate with both manually executed (R(2)=0.97) and literature reported experimental results (R(2)=0.93). Overall, measured E(H)s were within two-fold of the literature values for approximately 90% of the 19 compounds tested. Additionally, good inter- and intra-day reproducibility was observed for all the 19 compounds. In conclusion, an automated and robust assay suitable for simultaneously testing up to 48 compounds with multiple time points has been validated. Throughput of 192 compounds per run can be achieved using 384-well plates to meet increasing needs in drug discovery. Currently, this automated assay is used to support early discovery profiling towards lead optimization of various discovery targets/programs.