P155 A New Mechanism-Based Human In Vitro Screen (C-DILI™ Assay) for Prediction of Cholestatic Liver Toxicity

Jonathan P. Jackson , Qualyst Transporter Solutions, Durham, NC
Kimberly M Freeman , Qualyst Transporter Solutions, Durham, NC
Robert L. St. Claire III , Qualyst Transporter Solutions, Durham, NC
Matthew K Palmer , Qualyst Transporter Solutions, Durham, NC
Christopher B Black , Qualyst Transporter Solutions, Durham, NC
Kenneth R Brouwer , Qualyst Transporter Solutions, Durham, NC
Cholestatic Drug Induced Liver Injury (DILI) in humans has been associated with bile salt export pump (BSEP) inhibition, however, in vitro hBSEP IC50 concentrations do not correlate with in vivo cholestatic DILI severity. Sandwich-cultured human hepatocytes (SCHH) when treated with BSEP inhibitors respond to the resulting increased intracellular concentrations of bile acids (BA) via activation of FXR (adaptive response). This results in decreased synthesis of BA and increased basolateral efflux of BA via OSTα/β, which reduces the intracellular concentration of BA and decreases the potential for cholestatic hepatotoxicity. Inhibition of BSEP or basolateral efflux and/or interference with the adaptive response (FXR antagonism) may increase the potential for drug-induced cholestatic BA hepatotoxicity. Cryopreserved, Transporter Certified™ human hepatocytes in a sandwich configuration (96-well format) were cultured using QualGro™ Media for 5 days. On Day 5 of culture, hepatocyte cultures were exposed to test compounds at 20X (or limit of solubility) of their systemic Cmax in Qual-Gro™ Sensitization Media containing physiologically relevant concentrations of lipids and bile acids for 24 hours. Approximately 50 compounds with a range of hBSEP IC50s and clinical cholestasis were evaluated. Treatment with DMSO (0.1%) and standard QualGro™ Media were used as controls. The cultures were measured for ATP content (CellTiter-Glo® Promega) and LDH secretion (CytoTox-ONE™ Promega) following 24 hours of treatment. The LiverTox data base was used to identify compounds that were consistent with hepatocellular injury. LDH readout was used as a surrogate marker of cholestatic bile acid toxicity. The C-DILI™ assay correctly predicted hBSEP False Positive (Rosiglitazone) as having low potential for hepatocellular cholestasis, and hBSEP False Negative (Deferasirox) as having a high potential. Troglitazone and Ketoconazole, both True Positives in the hBSEP analysis were also correctly predicted in the C-DILI™ Assay. Overall, the C-DILI™ assay had a sensitivity and selectivity of 87.5% and 97.3%, respectively, and an accuracy of 96%. A whole cell assay integrating BSEP inhibition and effects on the adaptive response will more accurately predict cholestatic DILI.