Pharmacogenomic dosing of warfarin: ready or not?

Warfarin is a medication with a narrow therapeutic index, nonlinear intrapatient pharmacokinetics, and high interpatient variability in its dose-response relationship. These characteristics create great difficulty in determining an appropriate dose; sub- or supratherapeutic doses can increase the risk of bleeding and venous thromboembolism complications. Algorithms based on nongenetic factors of patient age, gender, body weight, diseases, diet, smoking, and medication traditionally have been used to determine warfarin dose requirements. However, these formulas account for less than 20% of the variability in warfarin response. Following completion of the Human Genome Project, several genetic variants of CYP2C9 and VKORC1 have been identified that account for a greater proportion of the variability in patient response to warfarin than is explained by nongenetic factors. Moreover, algorithms that analyze both patient genetic and nongenetic factors, i.e., pharmacogenomics, in warfarin response account for 55% to 60% of the variability. This raises the prospect of enhancing the ability to predict warfarin dose requirements and, thereby, improving its safety, effectiveness, and therapy efficiency. This review evaluates the impact of combining genetic and nongenetic factors in accounting for the variability in warfarin response and the prospect that pharmacogenomic algorithms will improve warfarin dosing early in therapy, possibly achieving a more rapid attainment of the therapeutic dose, improving safety, and increasing effectiveness. The most comprehensive and widely available pharmacogenomic algorithms for estimating warfarin dose requirements when initiating therapy, www.WarfarinDosing.org, is reviewed.