Prediction of human hepatic clearance from in vivo animal experiments and in vitro metabolic studies with liver microsomes from animals and humans

We investigated the quantitative prediction of human hepatic metabolic clea rance from in vitro experiments focusing on cytochrome P450 metabolism with eight model compounds, FK1052, FK480, zolpidem, omeprazole, nicardipine, n ilvadipine, diazepam, and diltiazem. For the compounds, in vivo human hepat ic extraction ratios ranged widely from 0.03 to 0.87. In vitro and in vivo hepatic intrinsic clearance (CLint) values for each compound were measured and calculated in rats and/or dogs and humans. CLint,in vitro was determine d from a substrate disappearance rate at 1 muM in hepatic microsomes, which was a useful method. CLint,in vivo was calculated from in vivo pharmacokin etic data using three frequent mathematical models (the well stirred, paral lel-tube, and dispersion models). The human scaling factor values (CLint,in vivo/CLint,in vitro) showed marked difference among the model compounds (0 .3-26.6-fold). On the other hand, most of the animal scaling factors were w ithin 2-fold of the values in humans, suggesting that scaling factor values were similar in the different animal species. When human CLint,in vitro va lues were compared with the actual CLint,in vivo, correlation was not neces sarily good. By contrast, using human CLint,in vitro corrected with the rat and/or dog scaling factors yielded better predictions of CLint,in vivo tha t were mostly within 2-fold of the actual values. Furthermore, successful p redictions of human CLoral, and hepatic extraction ratio (E-H) were obtaine d by use of the human CLint,in vitro corrected with animal scaling factors. The new variant method is a simple one, incorporating additional informati on from animal studies and providing a more reliable prediction of human he patic clearance.