Dosing algorithms for vitamin K antagonists across VKORC1 and CYP2C9 genotypes

Background The multicenter, single-blind, randomized EU-PACT trial compared the safety and efficacy of genotype-guided and non-genetic dosing algorithms for acenocoumarol and phenprocoumon in patients with atrial fibrillation or deep vein thrombosis. The trial showed no differences in the primary outcome between the two dosing strategies. Objectives To explore possible reasons for the lack of differences between trial arms by performing a secondary analysis of EU-PACT data in order to evaluate the performance of both dosing algorithms across VKORC1-CYP2C9 genetic subgroups. Patients/Methods Anticoagulation control measured according to an International Normalized Ratio (INR) below (INR of < 2), within (INR of 2-3) and above (INR of > 3) the therapeutic range was compared across VKORC1-CYP2C9 subgroups. Owing to a low number of patients in each subgroup, trials for acenocoumarol and phenprocoumon were combined for analysis. Results Four weeks after therapy initiation, genotype-guided dosing increased the mean percentage of time in the therapeutic INR range (PTIR) in the VKORC1 GG-CYP2C9*1*1 subgroup as compared with the non-genetic dosing (difference of 14.68%, 95% confidence interval [CI] 5.38-23.98). For the VKORC1 AA-CYP2C9*1*1 subgroup, there was a higher risk of under-anticoagulation with the genotype-guided algorithm (difference of 19.9%; 95% CI 11.6-28.2). Twelve weeks after therapy initiation, no statistically significant differences in anticoagulation control between trial arms were noted across the VKORC1-CYP2C9 genetic subgroups. Conclusions EU-PACT genetic-guided dose initiation algorithms for acenocoumarol and phenprocoumon could have predicted the dose overcautiously in the VKORC1 AA-CYP2C9*1*1 subgroup. Adjustment of the genotype-guided algorithm could lead to a higher benefit of genotyping.