The influences of CYP2D6 genotypes and drug interactions on the pharmacokinetics of venlafaxine: exploring predictive biomarkers for treatment outcomes.

ABSTRACT

RATIONALE Two biomarkers concentration ratio of O-desmethylvenlafaxine/venlafaxine and concentration sum of venlafaxine + O-desmethylvenlafaxine were adopted to indicate venlafaxine responses, but neither is validated. OBJECTIVES: To evaluate the ability of two biomarkers in reflecting venlafaxine pharmacokinetic variations, and to further examine their relationship with venlafaxine treatment outcomes. METHODS: Two well-defined influencing factors CYP2D6 genotypes and drug interactions were enriched into a three-period crossover study to produce venlafaxine pharmacokinetic variations In each period, healthy CYP2D6 extensive metabolizers (EM group; n = 12) and CYP2D6*10/*10 intermediate metabolizers (IM group; n = 12) were pretreated with clarithromycin (CYP3A4 inhibitor), or nothing (control), or clarithromycin + paroxetine (CYP3A4 + CYP2D6 inhibitors), before administration of a single-dose of 75 mg venlafaxine. Both biomarkers were evaluated (1) for their relationship with the influencing factors in healthy volunteers and (2) for their relationships with the venlafaxine responses/adverse events reported in two patient studies. RESULTS: Significant venlafaxine pharmacokinetic variations were observed between the EM and IM groups (geometric mean ratio [95 % CI] of area under the curve, 3.0 [1.8-5.1] in the control period), and between the control and clarithromycin + paroxetine periods (4.1 [3.5-4.7] and 2.0 [1.7-2.4] in the EM and IM group, respectively). O-Desmethylvenlafaxine/venlafaxine was superior to venlafaxine + O-desmethylvenlafaxine to reflect the influencing factors. In the patient studies, O-desmethylvenlafaxine/venlafaxine > 4 showed high precision in predicting venlafaxine responders/partial-responders (92 %) and patients without venlafaxine-related adverse events (88 %); the O-desmethylvenlafaxine/venlafaxine < 4 and venlafaxine + O-desmethylvenlafaxine > 400 ng/ml combination showed higher precision (100 %) than O-desmethylvenlafaxine/venlafaxine < 4 alone (65 %) in predicting venlafaxine non-responders. CONCLUSION: We propose using O-desmethylvenlafaxine/venlafaxine for CYP2D6 phenotyping, and O-desmethylvenlafaxine/venlafaxine with venlafaxine + O-desmethylvenlafaxine for predicting venlafaxine treatment outcomes in future prospective studies.