Extrapolation of the Hepatic Clearance of Drugs in the Absence of Albumin In Vitro to That in the Presence of Albumin In Vivo: Comparative Assessement of 2 Extrapolation Models Based on the Albumin-Mediated Hepatic Uptake Theory and Limitations and Mechanistic Insights.

The extrapolation of hepatic clearance (CL) from data determined in an in vitro assay in the absence of albumin (ALB) to that in the presence of ALB in liver in vivo was often inaccurate using traditional in vitro-to-in vivo extrapolation (IVIVE) methods for drugs binding to the ALB. It is recognized that considering an ALB-facilitated hepatic uptake phenomenon in the IVIVE can improve the extrapolation. Therefore, the present study provides a comparison of 2 existing models that account for the ALB-facilitated hepatic uptake phenomenon in the IVIVE of CL. These models assume an interaction of the ALB-bound drug complex with the hepatocyte membrane that enhanced the dissociation of the drug from ALB to result in increased unbound intracellular drug levels available for metabolism or transporter-mediated elimination. One model refers to the old facilitated-dissociation model (FDM), which is based on a binding isotherm and necessitates knowing the specific input parameters of the interaction (i.e., relative capacity of the interaction, dissociation constant, number of binding sites, and ALB concentration). The other model is based on the same theory but is recent and more speculative although it presumes that each interaction between the ALB-drug complex and the hepatocyte surface would at all times enhance and deliver the dissociated bound drug moiety into the hepatocytes and therefore, has the advantage to use less binding information. Consequently, this second model simply consists of adjusting the unbound fraction determined in plasma in vitro of each drug (fu) with the real differential of ALB concentration between the plasma and liver in vivo to estimate the corresponding differential of ALB-drug complex also assumed available to deliver the unbound drug moiety for hepatic uptake in vivo versus in vitro. Application of these 2 models (FDM and fu) significantly improved the IVIVEs of CL of drugs, and hence, the next step was to compare these 2 models with the same data set. Recently published data on the hepatic uptake of 2 organic anions, namely 1-anilino-8-naphthalene sulfonate and pitavastatin, provide all binding information. As expected, the results indicate that these 2 models are conceptually and mathematically equivalent as well as they successfully predicted the experimentally determined ratios of the unbound intrinsic CL (CL) in the presence of ALB in vivo to that in the absence of ALB in vitro. However, the 2 models were equivalent particularly for pitavastatin because its ALB-drug complex showed a relevant capacity of interaction and dissociation with the hepatocyte membrane. Conversely, for 1-anilino-8-naphthalene sulfonate, the model of fu overestimated the ratio of unbound CL by contrast to the FDM model because its ALB-drug complex demonstrated a significantly lower capacity of interaction with the membrane. The rational is simply because the model of fu presumably assumed an important facilitated-uptake phenomenon for each drug, whereas the FDM model was derived from binding data specific to each drug. Overall, these 2 models are complementary, and all contribute toward achieving the same objective of quantifying the ALB-facilitated uptake phenomenon; however, the FDM model is more specific, but its application necessitates collecting more binding data compared with the model of fu that can be used prospectively to predict the maximal effect of the facilitated-hepatic uptake in IVIVE.