In vitro hepatocyte uptake-based estimation of tissue partition coefficients of basic compounds for PBPK modeling

Accurate prediction of tissue-to-plasma partition coefficients (Kp,_tissue) is critical for physiologically based pharmacokinetic (PBPK) modeling, particularly for basic compounds (pKa > 7) that display complex distribution driven by lysosomal sequestration and acidic phospholipid binding. We developed a novel in vitro method to estimate the hepatocyte partitioning coefficient (Kp,_hep) using isolated, metabolically inactivated rat hepatocytes. For selected basic compounds, in vitro Kp,_hep correlated strongly with in vivo liver partitioning (R² = 0.97), supporting its utility as a surrogate for hepatic distribution. Two strategies were then evaluated to extrapolate this information to other tissues. A scaling-factor approach based on the Poulin and Theil tissue composition-based model applied a liver-derived scalar across 11 tissues, improving prediction accuracy, with 60% and 73% of Kp values within 2-fold and 3-fold of observed data, respectively (AFE 0.70; AAFE 2.2). A second, regression-based approach adapted a prior muscle-derived model by estimating muscle Kp from hepatocyte Kp and extending predictions to additional tissues, yielding 63% and 83% of values within 2-fold and 3-fold, respectively (AFE 1.5; AAFE 1.8). These findings demonstrate that hepatocyte-based partitioning provides a practical, mechanistically relevant, and scalable method for predicting tissue distribution of basic compounds in PBPK modeling.