A Novel Approach to the Prediction of Drug-Drug Interactions in Humans Based on the Serum Incubation Method

抄録

  A novel method for the prediction of drug-drug interaction has been established based on the in vitro metabolic stability in the “serum incubation method” using cryopreserved human hepatocytes suspended in 100% human serum. As a novel approach, the inhibitory effect of inhibitors on the metabolism of substrates during the first-pass elimination process in the liver (hepatic availability) and in the elimination process from the systemic circulation (hepatic clearance) were separately predicted with the anticipated inhibitor/substrate concentrations during absorption and in the systemic circulation, respectively. Ketoconazole strongly inhibited CYP3A4-mediated terfenadine metabolism in vitro, and the method predicted 6- to 37-fold increase of terfenadine AUC by the concomitant dosing of ketoconazole, which reasonably well agreed with the observed 13- to 59-fold increase of AUC in clinical studies. The CYP3A4-mediated metabolism of indinavir was also subject to the inhibition by ketoconazole in vitro at the lower indinavir concentration (2 μM), whereas no substantial inhibition was observed at 12 μM due to the saturation of indinavir metabolism. Predicted no interaction between ketoconazole and indinavir was consistent with the minimal increase (1.3-fold increase) of indinavir AUC by ketoconazole observed in clinical study. In addition, the method was applied to the CYP2D6-mediated desipramine-quinidine interaction: the predicted 6.4-fold increase of desipramine AUC by quinidine was consistent with the observed 6.7-fold increase of AUC in the clinical drug-drug interaction study. On the other hand, desipramine metabolism was little affected by ketoconazole in vitro, and consequently, it predicted no drug-drug interaction between desipramine and ketoconazole in humans, which agreed with the negligible interaction observed in clinical study. The accuracy of predictions for drug-drug interaction by the serum incubation method was evaluated by comparing the predicted increase of AUC after an oral administration by the inhibitor with the corresponding drug-drug interaction reported from clinical studies. These data demonstrated that the newly established method provides an in vitro tool for the prediction of drug-drug interaction with the accuracy ranging from 0.46 to 1.5.