Characterization of drug disposition in humans using novel in vitro methodologies based on the Extended Clearance Model

Safety and efficacy of drugs depend on their exposure in the body, which is determined by dose and bioavailability, but also by drug disposition as a result of tissue distribution and elimination processes. Knowledge about drug disposition in humans is therefore critical for the successful development of new drugs, with clinical information being unavailable at early development stages. To overcome this limitation, the pharmacokinetic properties of new drug candidates are routinely characterized using cell-based in vitro methods and in vitro-in vivo extrapolation (IVIVE) models. However, the assessment of drug distribution and elimination remains challenging. It was therefore the aim of this thesis to 1) establish a mechanistic in vitro model to study the hepatic distribution of unbound drug and to validate the model by predicting the clinical risk of drug-induced cholestasis, 2) investigate the applicability of additional in vitro methods for the determination of hepatic distribution of unbound drug, and 3) develop an in vitro model for the prediction of total (hepatic and renal) drug clearance and elimination pathway contributions in humans.
Knowledge about the drug distribution into tissues and the corresponding unbound intracellular drug concentrations is of particular interest in the context of intracellular drug effects related to toxicity, pharmacokinetics, and pharmacodynamics. For instance, prediction of drug-induced cholestasis due to inhibition of the intrahepatic bile salt export pump (BSEP) is commonly conducted using the unbound systemic drug exposure as a surrogate for the unbound intrahepatic concentration following the “free-drug hypothesis”. However, this assessment offers limited translatability to the clinical cholestasis risk since the effective unbound intrahepatic drug concentration is affected by active transport and/or metabolic processes. To improve such evaluations of intrahepatic drug interactions, the determination of the liver-to-blood partition coefficient for unbound drug at steady-state (Kpuu) was established based on in vitro measurements of active and passive sinusoidal uptake permeability, sinusoidal efflux permeability, hepatic metabolism, and biliary secretion according to the Extended Clearance Model (ECM). Following successful validation of the ECM-based Kpuu approach by in vitro-in vivo correlation in rats, human Kpuu data of 18 drug compounds were used to calculate unbound intrahepatic drug concentrations based on clinical drug exposure. This assessment significantly improved the translation of BSEP inhibition in vitro data to human and allowed the prediction of the clinical cholestasis frequency. Moreover, usefulness of the ECM as a drug classification system and for the quantitative evaluation of genetic and physiological risk factors for the development of cholestasis was demonstrated. The determination of unbound intrahepatic drug concentrations using the ECM-based hepatic Kpuu is therefore expected to improve early risk assessment of drug-induced cholestasis as well as of other intrahepatic drug interactions.
The ECM-based determination of Kpuu was successfully established and validated. However, this approach is labor and cost-intensive. A second project therefore aimed at comparing alternative in vitro Kpuu determination methods for the previously investigated compound set. For this purpose, three straightforward approaches were selected that rely on separate in vitro measurements of the liver-to-blood partition coefficient for total drug at steady-state (Kp) and the unbound fraction in hepatocytes (fuhep). Kp was generally determined in hepatocellular drug accumulation experiments in the absence of intrinsic metabolic and biliary clearance processes, whereas fuhep was either measured in hepatocellular drug accumulation experiments on ice (temperature method), using homogenized hepatocytes in equilibrium dialysis experiments (homogenization method), or calculated from the distribution coefficient logD7.4 using an empirical model (logD7.4 method). All investigated methods indicated deviations to ECM-derived Kpuu data, which were closely linked to the pharmacokinetic and physicochemical compound properties, namely the extent of intrinsic hepatic clearance, logD7.4, and molecular weight. The usefulness of the alternative Kpuu determination methods is therefore limited, with the ECM remaining the preferred approach for an integrated assessment of hepatic Kpuu. Nevertheless, the alternative methods can provide valid fuhep data if the physicochemical compound properties are considered for the selection of the appropriate method.
During drug development, hepatic drug clearance is routinely predicted using in vitro approaches such as the ECM. In contrast, appropriate in vitro models for the prediction of renal drug clearance are lacking. Thus, the assessment of total clearance for new drug candidates is strongly limited. To overcome this drawback, an empirical in vitro model was established that provides estimates of the relative hepatic metabolic, biliary, and renal elimination pathway contributions in humans, based on in vitro sinusoidal uptake permeability data. This assessment subsequently allows the extrapolation of hepatic into total drug clearance. Under consideration of ECM-based hepatic clearances, the model provided accurate predictions of total human clearance for 10 developmental compounds. Moreover, it was demonstrated that the Extended Clearance Concept Classification System (ECCCS) is applicable to evaluate the relevance of metabolic, biliary, and renal drug elimination, which provides useful guidance for the design of follow-up enzyme and transporter phenotyping studies. Thus, the established model allows a simple and highly reliable assessment of total drug clearance and relative elimination pathway contributions in humans based solely on hepatic in vitro data, facilitating a tailor-made pharmacokinetic assessment during early drug development.