Pharmacogenetics of Warfarin Therapy

A substantial burden of cardiovascular disease results from thrombosis that manifests clinically as acute coronary syndrome, stroke, and venous thromboembolism. Although antithrombotic drugs have been shown to reduce thrombotic events, there exists considerable variability in response to many of these drugs that affects the relative efficacy and safety in individual patients. Clinical, environmental, and genetic factors contribute to variability in drug response. A recent trial (1) has provided important insights into how pharmacogenomics offers an opportunity for personalized medicine through the identification of medically actionable variants to help determine an optimal dosing regimen for warfarin. Vitamin K antagonists, such as warfarin, are the most frequently prescribed oral anticoagulants used worldwide and are indicated for patients across the spectrum of cardiovascular disease, including those with or at risk for venous thromboembolism (deep venous thrombosis and pulmonary embolism), atrial fibrillation (AF),2 and mechanical heart valves. Warfarin's mechanism of action is the prevention of γ-carboxylation of the vitamin K-dependent coagulation factors prothrombin (factor II), factor VII, factor IX, and factor X. Although warfarin and other vitamin K antagonists are highly effective in reducing thromboembolism, their use is limited by multiple genetic, food, and drug interactions, as well as a narrow therapeutic index that necessitates frequent monitoring and dose adjustments, resulting in considerable bleeding risk and inconvenience. Even with frequent monitoring and dose adjustments, patients taking vitamin K antagonists spend about a third of the time outside the standard target therapeutic range international normalized ratio (INR), 2.0–3.0. A substantial portion of interindividual variability in the metabolism and response to warfarin is explained by genetic polymorphisms in cytochrome P450 family 2 subfamily C member 9 (CYP2C9)3 and cytochrome P450 family 4 subfamily F member 2 (CYP4F2) genes, which are involved in the metabolism of warfarin and vitamin K epoxide reductase complex subunit 1 (VKORC1) gene, whose product is the molecular target of warfarin and is the rate-limiting step in vitamin K recycling. The US Food and Drug Administration revised the warfarin product label in 2007 to reflect that knowing a patient's CYP2C9 and VKORC1 genotype could help optimize warfarin dose selection. It certainly makes intuitive sense that a pharmacogenetic-based dosing algorithm should optimize the time to reach and stay in a therapeutic range, which translates into improved clinical outcomes. Unfortunately, recent clinical trials assessing whether incorporating genotype into initial warfarin dosing to improve time in therapeutic range have provided mixed results (2–4). In addition, none of the trials had adequate sample size or duration of follow-up to examine clinical outcomes. The recently published randomized clinical Genetic Informatics Trial (GIFT) of warfarin to prevent deep venous thrombosis helps fill in some important gaps in our data regarding the efficacy of genotype-guided warfarin therapy (1). The GIFT trial was designed to determine whether genotype-guided dosing improved the safety of warfarin initiation in patients ≥65 years being given warfarin for elective hip or knee arthroplasty. The trial randomized 1650 individuals to a genotype-guided or clinically guided warfarin dosing on days 1 through 11 of therapy and to a target INR of 1.8 or 2.5 in a 2 × 2 factorial design. This publication (1) reports only the genotype-guided vs clinically guided warfarin dosing strategy. Patients were genotyped for CYP2C9, CYP4F2, and VKORC1 polymorphisms. The primary endpoint was the composite of major bleeding, INR ≥ 4, venous thromboembolism, or death. Patients underwent screening lower-extremity duplex ultrasonography approximately 1 month after arthroplasty. A total of 87 patients (10.8%) in the genotype-guided group vs 116 patients (14.7%) in the clinically guided warfarin dosing group met at least 1 component of the primary endpoint (relative risk, 0.73, 95% CI, 0.56–0.95; P = 0.02). The percentage of time in therapeutic range (TTR) improved by 3.4% from a mean of 51.3% with clinically guided warfarin dosing to 54.7% with genotype-guided dosing. This trial has several strengths in comparison with previous trials in this field. The GIFT trial was larger, used genotype-guided dosing for 11 days compared with 5 or fewer days in prior trials, and incorporated the CYP4F2 polymorphism in the genotype algorithm. These factors allowed it to be powered for an endpoint that included clinical events instead of just TTR, a surrogate endpoint. However, it must be noted that approximately 65% of the events were for INR ≥ 4. No individual component of the primary endpoint reached statistical significance, except for INR ≥ 4, although the relative reductions were all consistent and favored the genotype-guided regimen. The authors did, however, perform a post hoc analysis that demonstrated a 1.4% reduction in the rate of symptomatic adverse major clinical events (major bleeding, symptomatic deep venous thrombosis, or pulmonary embolism), but it did not reach statistical significance (P = 0.51). The GIFT trial provides additional evidence that genotype-guided warfarin dosing likely not only improves TTR but also has an impact on clinical events. Although larger than prior trials, the GIFT trial was still underpowered to definitively evaluate the clinical utility of whether routine genotype-guided warfarin dosing should be embraced. Larger trials are needed. An additional question is whether nonvitamin K antagonist oral anticoagulants (NOACs) should be preferentially used instead of warfarin. A genetic analysis from one of the phase 3 trials comparing one of the NOACs with warfarin for stroke prevention in AF suggests that genotyping may have a role to play in optimizing anticoagulation selection (5). In total, 14348 patients in the Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation–Thrombolysis in Myocardial Infarction 48 (ENGAGE AF-TIMI 48) trial, which randomized patients to warfarin or the factor Xa inhibitor edoxaban, were genotyped for variants in CYP2C9 and VKORC1. Among patients treated with warfarin, compared with healthy responders, patients with CYP2C9 and VKORC1 genetic variants that made them sensitive and highly sensitive warfarin responders spent a greater proportion of time overanticoagulated and were at higher risk of bleeding. In such patients, edoxaban was particularly beneficial in reducing bleeding. These data demonstrate that genotyping could be used to identify patients who are especially good candidates for treatment with NOACs or, if they are treated with warfarin, might benefit from more careful dose selection and monitoring. Antithrombotic drugs play a critical role in our therapeutic armamentarium. Although clinical, environmental, and genetic factors are important contributors to the variability in drug response, the emergence of pharmacogenomics offers the opportunity for personalized medicine through the identification of medically actionable variants. Despite ample promise, no cardiovascular pharmacogenetic association is routinely tested for in clinical practice. The reasons for this are varied and include evidential, logistical, and financial barriers. The data from GIFT and ENGAGE AF-TIMI 48 trial suggest that the main benefit of pharmacogenetic testing for CYP2C9 and VKORC1 genetic variants may not be to improve warfarin dosing, but rather to identify a subset of patients who would derive the most benefit from the more expensive NOACs. In countries where NOACs have already replaced warfarin for almost all new initiations of anticoagulation, there is little to gain from pharmacogenetic testing for warfarin. But in resource-limited countries where vitamin K antagonists still dominate, such a strategy may be cost-effective. atrial fibrillation international normalized ratio time in therapeutic nonvitamin K antagonist oral anticoagulant Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation–Thrombolysis in Myocardial Infarction 48. cytochrome P450 family 2 subfamily C member 9 cytochrome P450 family 4 subfamily F member 2 vitamin K epoxide reductase complex subunit 1.