A practical framework to predict human hepatic clearance for Extended Clearance Classification System class 1B acidic...

The Extended Clearance Classification System (ECCS) predicts predominant systemic clearance mechanisms of small-molecule drugs based on molecular weight, ionization, and permeability. Highly permeable acids and zwitterions with a molecular weight exceeding 400 Da (ECCS class 1B) are expected to rely on transporter-mediated hepatic uptake. For this class, human hepatic clearance is frequently underpredicted when extrapolated from in vitro metabolic stability, complicating prospective pharmacokinetic predictions. Here, we systematically evaluated the performance of metabolism-based intrinsic hepatic clearance predictions using curated clinical intravenous pharmacokinetic data for 16 ECCS class 1B drugs. Although approximately 70% of compounds were predicted within 3-fold of observed values, clinically meaningful underprediction persisted for a subset (atorvastatin, bosentan, pitavastatin, lovastatin acid, and telmisartan). To prospectively identify compounds at risk of >3-fold underprediction, we compared 4 approaches: (1) rat in vitro-in vivo correlation (IVIVC), (2) nonhuman primate (NHP) IVIVC, (3) IVIVC in chimeric mice with humanized livers (PXB-mice), and (4) measurement of the fraction of active uptake in suspended human hepatocytes. Rat IVIVC reliably identified compounds scaling adequately in humans but showed lower discriminatory power to identify compounds with >3-fold underprediction. In contrast, a human hepatocyte active uptake fraction >50% robustly identified the compounds with human hepatic clearance underprediction. NHP and PXB-mouse IVIVC provided lower discriminatory power. For compounds flagged to be at risk of >3-fold underprediction by the hepatic uptake assay, incorporation of measured hepatic uptake clearance into scaling quantitatively improved human clearance predictions, achieving accuracy comparable to single-species NHP allometric scaling. SIGNIFICANCE STATEMENT: Accurate clearance prediction for Extended Clearance Classification System class 1B drugs is complicated by transporter-mediated hepatic uptake, which can cause substantial underprediction when scaling from metabolic stability. This work introduces a practical decision framework combining rat in vitro-in vivo correlation and human hepatocyte uptake screening to identify compounds at risk of >3-fold clearance underprediction. For the identified compounds, uptake-informed scaling improves human pharmacokinetic predictions, reduces uncertainty during lead optimization, and lowers reliance on nonhuman primates in drug discovery.