Which Unbound Fraction Should We Use in the Well-Stirred Model for More Accurately Predicting Hepatic Clearance of Drugs for Humans?

Abstract

As the hepatic clearance (CL_H) of drugs becomes independent of hepatic blood flow rate, CL_H depends primarily on intrinsic clearance (CL_int). Incubations of microsomes or hepatocytes are commonly used to generate CL_int. Therefore, CL_int estimates corrected for binding to the in vitro systems and scaled to whole liver, are applied to a well-stirred liver model to obtain CL_H predictions for drugs. In other words, CL_int is extrapolated with the ratio of unbound fraction between the plasma (fu_p) and incubation medium (fu_inc). However, this binding correction resulted to an important underprediction bias of CL_H. Therefore, the approach considering fu_p and fu_inc needs to be better understood for more precisely scaling CL_int. The objective of this study was to explain the underprediction bias of CL_H based on physicochemical properties of drugs. Analysis-ready datasets have been collected so that evaluation of the data generates a mechanistic understanding of the impact of unbound fraction on the prediction of CL_H of human for 128 drugs. Experimental values of fu_inc for liver are quantifying solely the binding to lipids in microsomes or hepatocytes in the absence of plasma proteins in the incubations. However, the experimental values of fu_p for plasma can estimate the binding to lipids and plasma proteins. Therefore, drugs binding primarily to lipids in the liver and plasma showed a less pronounced underprediction bias of CL_H by using the ratios of fu_p/fu_inc in the well-stirred model. In contrast, drugs binding primarily to the plasma proteins in the liver and plasma showed a larger underprediction bias of CL_H. Furthermore, for the ionized drugs, values of fu_p and fu_inc are not covering the pH gradient effect between plasma and hepatocytes, which also impacted the CL_H predictions. For these reasons, a mechanistic approach was proposed to replace the conventional fu_p value with an adjusted fu_p (fu_-adjusted) that accounts for differences in proteins/lipids binding and pH gradient effect between the liver and plasma. Hence, replacing fu_p with fu_-adjusted did provide a universal and mechanisms-based approach removing the underprediction bias for all datasets of drugs. Overall, this study indicates which drug properties generated the largest underprediction bias of CL_H and suggests that the Poulin et al. method referring to fu_-adjusted could be the most proper unbound fraction to reduce that bias with the well-stirred model.