Abstract
Introduction: Warfarin is used in a wide range of doses and requires frequent INR monitoring with respective dose adjustment. Intra-individual variability of warfarin dose is determined by the individual genotype and CYP2C9 and VKORC1 gene polymorphisms.
Pharmacogenetic algorithms could be used to predict the daily dose of warfarin even before the initiation of warfarin treatment. Aim of study: To assess frequency of CYP2C9 and VKORC1 polymorphism in the Czech population and to compare empirical daily dose of warfarin with the dose predicted by three previously published pharmacogenetic algorithms.
Methods: CYP2C9 (alleles *1, *2 and *3) and VKORC1 (haplotypes A and B) genotyping was performed in 1,972 patients. Accuracy of warfarin daily dose prediction was assessed in a cohort of 280 patients with complete relevant clinical data and on a stable dose of warfarin.
Results: The heterozygous form of the variant genotype of CYP2C9 (reduced warfarin metabolism) was present in 11.6% of patients in our cohort, the homozygous form was found in 1.1%. VKORC1 haplotype A/A (lower sensitivity for warfarin) was present in 14% of the cohort.
Standard expected mean dose of warfarin was used by patients with no (29.2%) or 1 variant allele (41.5%). Coefficients of determination (R2) of the respective assessed algorithms were: Anderson 21.9%, Gage 23.8% and Sconce 58.4%.
Conclusion: The algorithm by Sconce et al provided the highest agreement between the predicted and empirical daily dose, with 4-fold higher probability that the predicted dose will be ?20% of the empirical dose compared to other assessed algorithms. Pharmacogenetic algorithms were found to be useful in patients of all body weight categories and in patients older than 80 years but not in patients younger than 40 years.