Genetics Informatics Trial (GIFT) of Warfarin to Prevent DVT

The overall objective of the Genetics-InFormatics Trial (GIFT) of Warfarin to Prevent DVT is to elucidate novel strategies to improve the safety and effectiveness of warfarin therapy. With this study we directly respond to Health and Human Services (HHS) priorities to advance the field of personalized medicine and to prevent venous thromboembolic (VTE) disease. In 2007, the Honorable Mike Leavitt, Secretary of HHS, announced the Personalized Health Care Initiative and wrote that a key goal was, "… to use our personal genetic information to tailor treatments more effectively to each patient."(1) Recently, President Obama and Francis Collins (Director of the NIH) have made precision medicine a national priority.(2) Previously, the Acting Surgeon General issued a Call to Action to reduce the number of cases of VTE in the United States.(3) To facilitate precision dosing strategies for VTE prevention, we have made publically available a non-profit, web application, www.WarfarinDosing.org. A public version of www.WarfarinDosing.org estimates warfarin doses for the initial 5 days of warfarin therapy. The version being evaluated in GIFT provides doses for the initial 11 days of warfarin therapy. Aim 1: To determine how pharmacogenetic-based warfarin therapy affects the safety and effectiveness of warfarin therapy. The intensity of anticoagulant therapy is measured by the International Normalized Ratio (INR). During initiation, the INR often falls outside the therapeutic range. INRs that are too low predispose patients to VTE while supratherapeutic INR values increase risk of bleeding.(4, 5) Previously, the FDA approved the label change of warfarin/Coumadin™ to recommend considering lower initial doses in patients known to have certain polymorphisms in genes affecting warfarin metabolism and sensitivity.(6) However, whether this strategy improves the safety and effectiveness of warfarin therapy in general is unknown. In particular, how this strategy affects subgroups with and without the genetic variants of interest is also unknown. Hypothesis 1: Pharmacogenetic therapy decreases the composite risk of a non-fatal VTE, non-fatal major hemorrhage, death, or INR ≥ 4.0 in all patients, and/or in the subgroup of patients whose pharmacogenetic and clinical predicted therapeutic maintenance doses differ by \> 1.0 mg/day. Based on our meta-analysis of prior trials(7), we anticipate 80% power to simultaneously detect a 32% relative risk reduction in the composite outcome for Aim 1 (as measured by a chi-square test). In the clinical arm, based on preliminary data, we anticipate that the rate of the composite outcome will be 15.7% in the clinical arm and 10.7% in the pharmacogenetic arm. We obtained these estimates because they average a rate of 13.2%, which is the rate of the composite outcome for Aim 1 observed from the initial 775 GIFT participants. The power was calculated using a two-sided alpha of 0.05 for a test of proportions, a drop-out rate of 2%, and a partitioned (two-sided) alpha with 0.044 allocated to the whole population and 0.01 to the high-risk subgroup. Because of correlation between these two subgroups, using these alphas preserves an overall type 1 error rate of 0.05. Aim 2: To determine whether warfarin therapy with a target INR of 1.8 is non-inferior to therapy with a target INR of 2.5 at preventing VTE or death in orthopedic patients. One randomized trial (PREVENT) found that a target INR value of 1.5-2.0 prevented 64% of VTE recurrence.(8) Although that trial excluded orthopedic patients, such an approach has been endorsed by the American Academy of Orthopedic Surgeons (AAOS). On page 15 of the 2007 AAOS guidelines (9) they offer the following recommendation for VTE prophylaxis around the time of joint replacement: "Warfarin, with an INR goal of ≤ 2.0, starting either the night before or the night after surgery, for 2-6 weeks." However, the AAOS grade the overall evidence for VTE prophylaxis in this population as low (level III). The AAOS guidelines conflict with the prior American College of Chest Physician (ACCP) guidelines,(10) which recommend, as one of their (Grade 1A) options (page 338 S), using an "…adjusted-dose vitamin K antagonist (INR target, 2.5; range 2.0 to 3.0)." Because lower target INR values may reduce the risk of hemorrhage and simplify warfarin management(8) we propose to test the following: Hypothesis 2: For prevention of non-fatal VTE or death, a target INR of 1.8 will be non-inferior to a higher target INR (2.5). Using a non-inferiority margin of 3% absolute risk reduction in non-fatal VTE or death and an estimated composite rate of 5.56% (based on preliminary GIFT data), we will have 83% power to detect the non-inferiority of a target INR of 1.8 in 1600 patients.