Predicting transporter‐mediated drug interactions: Commentary on
Concurrent administration of multiple probe substrates (e.g., cytochrome P450 (CYP) enzymes), the so‐called “cocktail” approach, has been used to assess DDIs in humans,4 especially when a perpetrator drug affects multiple elimination pathways. Notwithstanding the potential interactions among probe substrates and challenges associated with analytical assays, this approach has several advantages. First, it is an efficient in vivo screening tool to study DDI potential of new molecular entities as perpetrators because interaction effect on multiple enzymes and/or transporters could be simultaneously studied rather than resorting to the conduct of multiple single drug‐pair studies. Second, the approach could minimize the confounding influence of interindividual and intraindividual pharmacokinetic variability over time. Third, it can be used to study certain DDIs in vivo when in vitro DDI assessments are challenging, as is the case for DDIs associated with certain herbal products or therapeutic proteins. Finally, in addition to DDI evaluation, such cocktail approaches can be used to mechanistically study effects of other patient factors, such as organ impairment, on multiple pathways.
A well‐selected and adequately validated cocktail approach can be a robust technique to demonstrate selectivity, and to estimate the magnitude of individual changes in a victim drug phenotype in the presence of a perpetrator drug or other patient factors. Moreover, such an approach can facilitate the early characterization of a drug candidate's DDI potential, and provide guidance for the safe conduct of larger clinical trials to identify proper comedication restrictions.4 Depending on the specificity of the substrate drugs and robustness of the cocktail, further in vivo DDI studies may be needed to provide quantitative exposure changes for dosing recommendations in labeling.
In the past, most cocktails developed to predict DDIs have focused on the involvement of CYPs. For example, CYP substrates with proper specificity and sensitivity (e.g., midazolam for CYP3A) have been included and validated in various cocktail combinations. In contrast, very limited information is available for cocktails that could be used to predict DDIs associated with transporter interactions. Ideal transporter probe substrates would be those that are minimally metabolized and are predominantly transported by a single transporter pathway, because the pharmacokinetics of such substrates are “sensitive” to decreased transporter function associated with either pharmacological or genetic inhibition. A major challenge for transporter probe substrates, however, is their general lack of selectivity for a particular transporter pathway; that is, most transporter substrates are also substrates for multiple other transporters and/or enzymes.3 For example, simvastatin is a substrate for both organic anion‐transporting polypeptide (OATP)1B and CYP3A. Therefore, the DDI results using one substrate may not be extrapolated to predict the perpetrator's effect on other substrate drugs. Another challenge is that “ideal” probe substrates cannot always be dosed together with other probe substrates due to the potential for pharmacokinetic and/or dynamic interactions. Therefore, doses of substrates in the cocktail and the tolerability and safety associated with their usage in humans need to be carefully evaluated. In addition, analytical assay challenges (e.g., assay sensitivity and specificity, assay interference, and compatibility of assay conditions) cannot be ignored because multiple moieties (substrates plus their metabolites) need to be assayed.