Drug properties; other

A computational model for predicting the PK of ATRA could aid the design of ATRA-dosing strategies and identify drug-drug interactions that could help overcome resistance caused by autoinduction of ATRA clearance. The ATRA model incorporated measurements of absorption of ATRA in a human colorectal epithelial cell line and plasma samples from healthy volunteers dosed with ATRA, biodistribution of ATRA in mice, metabolic clearance of ATRA, and ATRA-induced expression of cytochrome P450 26A1 (CYP26A1; CP26) in a human liver epithelial cell line to generate an algorithm that predicted the ATRA area under the curve (AUC) for plasma concentration-time in patients based on the size of ATRA dose. To predict ATRA-drug interactions that could reduce resistance by increasing the AUC, the model was combined with PK models for inhibitors of ATRA metabolism that incorporated the kinetics of multiple ATRA-catabolizing cytochrome P450 enzymes. In cancer patients dosed with ATRA, the model correctly predicted the AUC in 188 of 220 (85%) patients. In patients dosed with ATRA and the ATRA catabolism inhibitors ketoconazole or liarozole, the ATRA model combined with PK models for the inhibitors correctly predicted the high ATRA AUC resulting from ATRA-ketoconazole and ATRA-liarozole interactions in 31 of 31 (100%) and 11 of 19 (58%) patients, respectively. Next steps include using the model to predict optimal ATRA-dosing strategies in combination therapies.

The generic ketoconazole, a 14α-sterol demethylase inhibitor, is marketed to treat fungal infections...