Induction of drug metabolizing enzymes: a survey of in vitro methodologies and interpretations used in the pharmaceutical industry--do they comply with FDA recommendations?

Abstract The FDA has published guidelines by which to carry out and interpret in vitro induction studies using hepatocytes but do researchers in pharmaceutical companies actually follow these to the letter? In a survey of 30 participants in the pharmaceutical industry, 19 questions were posed regarding the species investigated, methodologies and interpretations of the data. Also addressed was the in-house decision making processes as a result of in vitro induction data. The survey showed that, although the basic methods were similar, no two researchers carried out and interpreted induction assays in exactly the same way. No single method was superior but all included enzyme activities as the major end point. Hepatocytes from animal species were used to confirm animal in vivo data but only human hepatocytes were used to predict human induction responses. If a compound was found to be positive in an in vitro induction assay, few would halt the development of the compound. The majority would consider other properties of the compound (bioavailability, clearance and therapeutic concentrations) and follow the FDA recommendation to conduct clinical drug–drug interaction studies. Overall, the results from this survey indicate that there is no standard pharmaceutical industry method or evaluation criterion by which in vitro assays are carried out. Rather than adhering to the FDA guidelines, some adapt methods and interpretation according to their own experience and need (whether screening or lead optimisation). There was general consensus that studies using human hepatocyte cultures currently provide the best indication of the in vivo induction potential of NCEs. In addition, the assessment of in vitro induction data from the literature suggest that the two-fold induction threshold and the percent of positive control criteria may not be the best methods to accurately assess the in vivo induction potential of a drug. Although the two-fold induction criterion is now obsolete, more predictive models for determining the clinical induction potential are needed. Alternative models are proposed and discussed herein.