Exploring the boundaries for in vitro-in vivo extrapolation: Use of isolated hepatocytes in co-culture and impact of albumin binding properties in the prediction of clearance of various drug types.

Abstract

Prediction of hepatic clearance of drugs (via uptake or metabolism) from in vitro systems continues to be problematic. This is particularly the case when plasma protein binding is high. The following work explores simultaneous assessment of both clearance processes, focusing on a commercial hepatocyte-fibroblast co-culture system (HµREL), over a 24 hour period, using a series of six probe drugs (ranging in metabolic and transporter clearance and low-to-high plasma protein binding). A rat hepatocyte co-culture prototype assay was established using drug depletion (measuring both medium and total concentrations) and cell uptake kinetic analysis, both in the presence (1 and 4%) and absence of plasma protein (BSA). Secretion of endogenous albumin was monitored as a marker of viability and this reached 0.004% in incubations (at a rate similar to in vivo synthesis). Binding to stromal cells was significant and required appropriate correction factors. Drug concentration-time courses were analysed both by conventional methods and a mechanistic cell model, prior to in vivo extrapolation. CL assayed by drug depletion in conventional suspended rat hepatocytes provided a benchmark to evaluate co-culture value. Addition of albumin appeared to improve predictions for some compounds (where fu<0.1); however, for high binding drugs, albumin significantly limited quantification and thus predictions. Overall, these results indicate the prospect of reliable measurement of hepatically cleared compounds, including (of novelty) those with low metabolic turnover but high active uptake, through the longevity of co-culture systems combined with judicious sampling. Significance Statement Co-culture systems, combining hepatocytes with stromal cells, offer a more advanced tool than standard hepatocytes, with the ability to be cultured for longer periods of time, yet their potential has not been comprehensively assessed. We evaluate the strengths and limitations of the HµREL system via the use of six drugs representing various metabolic and transporter-mediated clearance pathways with various degrees of albumin binding. Studies in the presence and absence of albumin allow IVIVE and a framework to maximise their utility.