Sense and nonsense in the prediction of drug-drug interactions

Drug interactions are always a major concern in medicine and within the pharmaceutical industry. Fatal drug interactions have been reported, and several prominent drugs have been withdrawn from the market because of serious adverse reactions related to drug interactions. Therefore, drug interactions represent not only a medical problem for clinicians, but also an economic loss for pharmaceutical companies. Today, many pharmaceutical companies are predicting potential interactions of new drug candidates in an attempt to minimize such losses and to more effectively safeguard the welfare of patients. Can in vivo drug interactions be predicted accurately from in vitro metabolic studies? Should the prediction be qualitative or quantitative? These are the fundamental questions that industrial drug metabolism scientists must confront daily. Prediction of in vivo drug interactions from in vitro metabolic data is highly controversial, because of the complexities of factors that are involved in drug interactions. Some scientists believe that quantitative prediction of drug interaction is possible, whereas others are less optimistic, and believe that quantitative prediction is extremely difficult, if not impossible. The purpose of this review is to present and discuss the technical problems inherent in estimating in vitro Ki values and in measuring inhibitor concentration at the active-site of enzymes. Theoretic considerations are briefly reviewed, and representative examples are drawn from literature to illustrate the sense and nonsense in predicting in vivo drug interactions.