Variation in hepatic extraction ratio with unbound drug fraction: discrimination between models of hepatic drug elimination.

Systematic examination of model-dependent predictions of changes in the hepatic extraction ratio (E), following alteration in the unbound fraction of drug in plasma (fub), should allow sensitive discrimination between the venous equilibrium model (model I) and the sinusoidal perfusion model (model II) of hepatic sinusoidal function if drugs which show high clearance of free drug are used. Analysis of experimental data from the literature confirmed the utility of this approach. Specifically, data related to diazepam (E = 0.95 at fub = 1) clearly conformed to the predictions of the sinusoidal perfusion model and differed markedly from those of the venous equilibrium model. Conversely, data for phenytoin (E = 0.69 at fub = 1) failed to discriminate between models, as predicted by our analysis. We identify a sensitive, convenient method for discrimination between current models of hepatic sinusoidal function and establish for the first time that a drug substrate (diazepam) conforms closely to the predictions of the sinusoidal perfusion model.