Prediction of in vivo potential for metabolic activation of drugs into chemically reactive intermediate: Correlation of in vitro and in vivo generation of reactive intermediates and in vitro glutathione conjugate formation in rats and humans

The covalent binding of reactive intermediates to macromolecules might have potential involvement in severe adverse drug reactions. Thus, quantification of reactive metabolites is necessary during the early stage of drug discovery to avoid serious toxicity. In this study, the relationship between covalent binding and glutathione (GSH) conjugate formation in rat and human liver microsomes were investigated using 10 representative radioactive compounds that have been reported as hepatotoxic or having other toxicity derived from their reactive intermediates: acetaminophen, amodiaquine, carbamazepine, clozapine, diclofenac, furosemide, imipramine, indomethacin, isoniazid, and tienilic acid, all at a concentration of 10 mu M. The GSH conjugate formation rate correlates well with the covalent binding of radioactivity (both rat and human, r(2) = 0.93), which suggests that quantification of the GSH conjugate can be used to estimate covalent binding. To quantify the GSH-conjugation rate with non-radiolabeled compounds in vitro, the validation study for the determination of GSH conjugate formation using S-35-GSH by radio-HPLC was useful to predict metabolic activation. Following oral administration of 20 mg/kg of the radiolabeled compounds to rats, radioactivity that covalently bound to plasma and liver proteins was determined. The in vivo maximum covalent binding level in liver based on the free fraction of plasma area under the concentration curve (AUC) and in vitro covalent binding rate was found to correlate well (r(2) = 0.79). Therefore, this model for in vitro covalent binding studies in human and rat and in vivo rat studies might be useful in predicting human metabolic activation of compounds.