Differential outcomes from metabolic ratios in the identification of CYP2D6 phenotypes--focus on venlafaxine and O-desmethylvenlafaxine

Abstract

Background: Venlafaxine (VEN), a well accepted anti-depressant, is metabolized through the cytochrome P 450 (CYP) 2D6 isozyme to form O-desmethyvenlafaxine (ODV). Due to the involvement of CYP2D6, the formation of ODV is influenced by genetic polymorphism. We used standard tools of assessment to explore the phenotypic distribution in a retrospective manner using the pharmacokinetic (PK) data of VEN and ODV obtained from several bioavailability/bioequivalence (BA/BE) studies in healthy subjects using the reference formulation.

Methods: Four single oral dose, open-label, randomized crossover BA/BE studies of VEN (doses: 37.5-150 mg) were performed in 141 healthy subjects. Plasma samples were collected over a period 72 h post VEN administration. The samples were analyzed for VEN and ODV using a validated LC/MS/MS assay with a limit of quantification 2.073 ng/mL for VEN and 3.973 ng/mL for ODV. PK parameters (C(max), T(max), AUC (0-t), AUC(0-infinity), t(1/2)) were computed using the noncompartmental approach. AUC metabolic ratios of VEN/ODV and ODV/VEN were computed for all subjects and were subjected to normality test procedures to tease out phenotypic distribution.

Results: ODV/VEN and VEN/ODV AUC metabolic ratios were evaluated for standard normal distribution and outliers to determine phenotypic distribution. Use of the VEN/ODV AUC metabolic ratio, arranged in a rank order, resulted in a distribution that distinguished poor metabolizers (PM) and extensive metabolizers (EM). The application of the ODV/VEN AUC metabolic ratio showed a unique distribution that distinguished ultra metabolizers (UM) and extensive metabolizers. By using both metabolic ratios, 141 healthy subjects were classified as follows: PMs = 18, EMs = 118, UMs = 5. Regardless of the formulation or dose size used, the plasma concentration-time profiles for both VEN and ODV were distinct amongst the three phenotypes identified in this work.

Conclusions: The use of VEN/ODV and ODV/VEN AUC metabolic ratios suggested quantitative differences. The data support the use of ODV/VEN but not VEN/ODV metabolic ratio for the identification of UM phenotypes of VEN. The derived metabolic ratios of ODV/VEN from this work were in line with other studies that used both phenotypic and genotypic correlation strategies for VEN.