In vitro cytochrome P450 inhibition data and the prediction of drug-drug interactions: Qualitative relationships, quantitative predictions, and the rank-order approach

Advances in the area of in vitro drug metabolism have been made such that data can be reliably used to predict in vivo DDIs. Qualitatively, and as expected, more potent inhibitors in vitro tend to be the drugs that cause DDIs in vivo, but potency alone is in itself not quantitatively predictive. Application of an approach whereby an in vivo drug interaction study is always conducted for the most potently inhibited CYP enzyme and, on the basis of the outcome of that study, subsequent interaction studies for less potently inhibited enzymes are considered, referred to here as the rank-order approach (Fig 1), appears to be reliable. Out of 21 drugs, only 3 failed this approach (with the application of strict criteria), and in each of these 3 cases, the drug interaction that would have been missed was not of high magnitude. However, most importantly, with use of a combination of in vitro inhibition data with an estimate of unbound hepatic inlet concentration of the inhibitor, the f m(CYP) of the probe drugs, and for CYP3A probes, the contribution of the intestine, predictions of in vivo DDIs can be made with reasonable reliability. This still requires acceptance of several assumptions that may be inappropriate for some compounds (eg, no impact of active hepatic uptake, in vitro inhibition data are not confounded by nonspecific binding, and the drug is not a mechanism-based inactivator or inducer); nevertheless, the conclusion that in vitro data can be used in the quantitative estimation of in vivo DDIs is tenable. Finally, it is advocated that in vitro data used in predictions of DDIs be generated with the highest-quality biochemical and analytic methods attainable. The in vitro inhibition data used in these analyses were gathered by use of validated assays that use sensitive analytic methods and low microsomal protein concentrations (0.01-0.2 mg/mL). Previous attempts to predict in vivo DDIs with in vitro data gathered from the scientific literature have not been as successful, likely because of the variations in biochemical and analytic methods used to determine the inhibition constants. Our ability to use in vitro inhibition data in the prediction of in vivo DDIs as presented in this commentary represents a significant advance in this field. Copyright © 2005 by the American Society for Clinical Pharmacology and Therapeutics.