Evaluation of common in vitro assays for the prediction of oral bioavailability and hepatic metabolic clearance in humans

Abstract

IntroductionIntrinsic hepatic metabolic clearance (CLint) measured with human hepatocytes, apparent intestinal permeability (Papp) obtained using the Caco-2 model, unbound fraction in plasma (fu) and blood-to-plasma concentration ratio (Cbl/Cpl) are commonly used for predicting the hepatic clearance (CLH) and oral bioavailability (F) of drug candidates in humans. The primary objective was to select drugs whose in vitro hepatocyte CLint, Caco-2 Papp, fu and Cbl/Cpl have been measured in various laboratories and studies, and estimate correlation coefficients (R2) for predicted and observed F and log plasma CLH. Secondary aims were to estimate the laboratory/study variability and its impact on predictions and to compare results to in silico and animal model-based predictions. Materials and MethodsA literature search was done in order to find unbound hepatocyte CLint, (and corresponding predicted in vivo CLint), Caco-2 Papp, fu and Cbl/Cpl data. Compounds with multiple measurements for the four assays, without significant in vivo solubility/dissolution limitations and with known in vivo CLH and F, were selected. Min, max and mean estimates were used in the analysis. Results and DiscussionThirty-two compounds with data (in total 561 estimates) produced by 21 major pharmaceutical companies and universities met the inclusion criteria. The predicted vs observed R2 for log mean CLint, log mean CLH and mean F were 0.32, 0.08 and 0.20, respectively. Exclusion of atenolol increased the R2 for CLH to 0.20. R2-values were considerably lower than those presented in many studies, which seems to be explained by selection bias (choosing favorable reference values). There was considerable interstudy variability for measured and predicted CLint (80- and 1,476-fold mean and max differences, respectively) and measured fu (6.6- and 50-fold mean and max differences, respectively). For F, higher predictive performance was found for in silico (Q2=0.58; head-to-head) and animal in vivo models (R2=0.30). ConclusionThe combination of data from many laboratories and the use of mean values resulted in reduced selection bias and predictive accuracy. Overall, the predictive accuracy (here R2) for log CLint, log CLH and F was low to moderately low (0.08-0.32). The halved R2 compared to individual studies where high performance was demonstrated seems to be explained be selection bias (enabled by large data variability). Animal in vivo models, and in particular, in silico methodology, outperformed in vitro methodology for the prediction of F in man.