Prediction of in vivo nonlinear first-pass hepatic metabolism of YM796 from in vitro metabolic data.

Recent cumulative evidence suggests the possibility of predicting the in vivo metabolic clearance and/or hepatic availability (Fh) from in vitro metabolism data under linear conditions. Under nonlinear conditions, however, it is essential to consider the rate constant for the absorption (ka) for predicting Fh after oral administration, because the time profiles for the portal vein concentration depends on ka. In our study, we numerically solved the dispersion model under nonlinear conditions to propose a method to predict Fh after oral administration by taking ka into consideration. As a model compound, (S)-(-)-2,8-dimethyl-3-methylene-1-oxa-8-azaspiro [4,5] decane-L-tartrate monohydrate (YM796) was used. After oral administration, we found that the dose-normalized AUC (AUCoral/dose) was markedly increased in rats from 5.0 x 10(-6) to 33 x 10(-6) hr/ml.kg as the dose increased from 1 to 10 mg/kg, whereas the same value was relatively constant in dogs (87.7 x 10(-6) to 105 x 10(-6) hr/ml.kg at 1 to 10 mg/kg) and in humans (1260 x 10(-6) to 1768 x 10(-6) hr/ml.kg at 5 to 60 mg/body). Kinetic analysis indicated that AUCoral could be accurately predicted at each dose if ka value was assumed as 0.07 min-1 for all animal species examined in our study. These results suggest that it is possible to predict Fh even if the metabolism is composed of non-linear process by considering the absorption rate into the portal vein.