Quantitative prediction of midazolam-ketoconazole drug -drug interaction

Aroonrut Lucksiri, Purdue University

Abstract

Ketoconazole (KTZ) is commonly used by pharmaceutical companies to characterize the worst-case drug interaction for CYP3A substrates under development. Doses of 200 mg and 400 mg of KTZ have been used in human drug interaction studies with midazolam (MDZ). However, large variations in the magnitude of interaction are observed. No single study has compared the inhibitory effect of 200 mg vs. 400 mg of KTZ. Thus, it is unclear whether the worst-case drug interaction study requires 400 mg of KTZ or if a 200 mg dose would suffice. CYP3A5 genotype may in part explain individual variability in metabolism and drug-drug interactions exhibited by CYP3A substrates due to its polymorphic and racially diverse expression. Twenty-four healthy volunteers completed this three-phase, randomized, crossover study. Intravenous (IV) and oral (PO) MDZ were administered on consecutive days alone (control) or on day 6 (IV) and 7 (PO) of KTZ 200 mg or 400 mg daily. Serum samples were assayed for MDZ and KTZ by HPLC-MS. CYP3A5 genotype was determined using allele specific real-time PCR (CYP3A5*3 and *6) and allelic discrimination real-time PCR (CYP3A5*7). The extent of CYP3A inhibition by KTZ was dose dependent with a significantly greater effect following 400 mg KTZ compared to 200 mg KTZ. The baseline systemic and oral clearance values were significantly greater in the group with at least one copy of CYP3A5 functional allele as compared to the group with no functional allele. There was no difference in the extent of interaction after 200 mg or 400 mg KTZ daily between the groups with no and at least one CYP3A5 functional allele. However, the ratios between control and treatment phases of the 1'OH MDZ area under the concentration vs. time curve (AUC) over MDZ AUC among expression groups was significantly different after 200 mg but not after 400 mg KTZ. A model defining KTZ and MDZ pharmacokinetics and employing a stochastic approach for a quantitative drug interaction prediction was developed. The interaction model predicted the extent of interaction reported in published articles. Further research is needed to define the interaction and the variance models.

Degree

Ph.D.

Advisors

Hall, Purdue University.

Subject Area

Pharmacology|Pharmaceuticals

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS