Drug Interaction Studies: Study Design, Data Analysis, and Implications for Dosing and Labeling

Abstract

One of the most effective ways in which regulatory agencies communicate with sponsors and guide drug development is through the issuance of guidances or guidelines. These can be issued domestically in a given region such as the United States by the Food and Drug Administration (FDA) or internationally through the International Conference on Harmonization. Currently, there are over 400 final or draft guidances that can be found through the FDA website. The development of guidances proceeds through a process known as Good Guidance Practices, which is intended to assure that there is an appropriate level of meaningful public participation in the development of guidance. In the past 10 years, clinical pharmacology guidances covering important areas have been issued, including pharmacokinetic data in patients with renal and hepatic impairment, dose-response studies, and drug-drug interactions.

No clinically relevant CYP3A induction after St. John’s wort with low hyperforin content in healthy volunteers

“…Regulatory authorities suggest that in case of a negative drug interaction with the most sensitive substrate, it can be presumed that less sensitive substrates would also be unaffected [22]. If study results with the most sensitive substrate indicate clinically significant interaction, additional studies with less sensitive substrates are recommended [22]. This pharmacokineticbased approach assumes that the pharmacokinetics of midazolam would be changed to a greater extent than the pharmacokinetics of other CYP3A substrates potentially coadministered in future patients.…”

“…To study the relevance of inhibition or induction by an investigational drug, oral midazolam is suggested to be the most sensitive substrate for CYP3A [22]. Regulatory authorities suggest that in case of a negative drug interaction with the most sensitive substrate, it can be presumed that less sensitive substrates would also be unaffected [22].…”

“…To evaluate if pharmacokinetic changes could have clinical relevance, pharmacokinetic drug-drug interactions are suggested to be treated as equivalence problems [22]. No-effect limits of 0.80-1.25 are often used by default.…”

“…The relative differences in pharmacokinetic parameters after comedication with St. John's wort compared with baseline were evaluated as mean and 95% confidence interval (CI). Moreover, 90% CI of the geometric mean ratio of the treatment phase over the control phase was calculated to evaluate differences in pharmacokinetic parameters in the absence and presence of St. John's wort [22]. No-effect limits of 0.80-1.25 have often been used by default.…”

“…If the investigational drug is suspected to be an inducer of CYP3A, an initial clinical study using oral midazolam as the most sensitive substrate for CYP3A is suggested. If this study were to show negative results, it could be presumed that less sensitive substrates would be unaffected, and no further studies would be needed [22]. In an exploratory study using midazolam as probe substrate, six St. John's wort preparations with different compositions of main constituents, especially hyperforin, have been found to induce CYP3A to different extents [19].…”