Pharmacogenetics on Trial: The Utility of Genotyping in Warfarin Management

Clarification of Optimal Anticoagulation Through Genetics (COAG); Genetics Informatics Trial (GIFT) of Warfarin to Prevent DVT

COAG – NCT00839657; GIFT– NCT01006733

COAG – National Heart, Lung, and Blood Institute (NHLBI)

GIFT – Washington University School of Medicine

COAG – Bristol-Myers Squibb

GIFT – Intermountain Health Care, Inc., University of Utah, Hospital for Special Surgery, New York, and NHLBI

COAG –16 centers throughout the United States

GIFT – Washington University School of Medicine (MO); Hospital for Special Surgery (NY); University of Utah, Intermountain Medical Center (UT)

COAG – Stephen E. Kimmel, MD, MSCE, University of Pennsylvania

GIFT – Brian F. Gage, MD, MSc, Washington University School of Medicine

COAG – 1,600

GIFT – 1,238

COAG is a phase 3, double-blind, randomized trial to compare the efficacy of a warfarin-dosing algorithm based on genotype and clinical data with a dosing algorithm based on clinical data only. Patients in the experimental arm will receive initial dosing of warfarin for the first three to four days of treatment, as determined by an algorithm that uses clinical data and evaluation of vitamin K epoxide reductase complex-1 gene (VKORC1) and cytochrome P450 2C9 gene (CYP2C9) polymorphisms. A dose adjustment will be made after three and/or four doses of warfarin using a dose revision algorithm that incorporates both clinical and genetic variant information. Warfarin dosing for the comparator arm will be guided by an algorithm that only includes clinical information. The primary outcome is the percentage of time patients spend in the therapeutic INR range (2-3) during the first four weeks of therapy. Secondary outcomes include the occurrence of an INR > 4 or a serious adverse clinical event during the first four weeks.

GIFT is a phase 3, double-blind, randomized trial designed to evaluate whether the addition of genotyping will reduce the risk of VTE and severe bleeding associated with warfarin management in patients > 65 years old receiving warfarin prophylaxis after hip or knee arthroplasty. The study contains two arms with different target INRs: 1.8 and 2.5. Within each arm, warfarin dosing guided by genotyping and clinical information will be compared with dosing guided by clinical information alone. Dosing algorithms are available at www.WarfarinDosing.org. Primary outcomes will be evaluated over a four- to six-week time frame and include a non-fatal venous thromboembolic event, non-fatal hemorrhage,  death from any cause, or INR > 4. Secondary outcomes include percent of time in therapeutic INR range, or time to first laboratory event (INR > 1.5 + target INR).

The challenge of dosing warfarin to achieve an INR of 2-3 is well-known to most hematologists. Accurately predicting the maintenance dose could improve the time within therapeutic range and potentially lead to safer administration of warfarin. A substantial body of literature demonstrates that polymorphisms in the VKORC1 and the (CYP2C9) affect warfarin dose requirements. Polymorphisms in these genes account for approximately 50 percent of the variance in dose requirements. These observations suggest that prediction of maintenance dosing could be improved by testing VKORC1 and CYP2C9 genetic variants. This hope has been buttressed by relatively small studies indicating increased time in therapeutic range and reduced episodes of bleeding with the use of VKORC1 and CYP2C9 genotyping. COAG and GIFT will test the hypothesis that guiding warfarin management using algorithms that incorporate genotyping information will improve the efficacy and safety of warfarin.

There is considerable interest in demonstrating tangible health-care benefits from the substantial investment made in deciphering the human genome. Pharmacogenetics is viewed as an area in which this objective may be achieved, and warfarin genotyping has been dubbed the poster child of pharmacogenetics. Warfarin is commonly used, inexpensive, effective, and has few side effects. Its Achilles’ heel, however, is its narrow effective and safe therapeutic range and the problem of severe bleeding associated with overdosing. The data that polymorphisms in VKORC1 and CYP2C9 influence response to warfarin are compelling and raise the possibility that systematic patient genotyping will inform improved warfarin management. COAG and GIFT are designed to address this possibility.

As with any trial testing algorithms of warfarin dosing, results from these tightly controlled studies may not be reproduced in the real world. The cost-effectiveness of genotype-guided warfarin management and the willingness of insurers to pay for testing are also uncertain. In addition to these hypothetical concerns, skeptics have raised two criticisms with regard to the approach itself. The first is a logistical argument: By the time the genotyping information becomes available for a particular patient, the INR response will havecaptured the relevant information on warfarin sensitivity. The second argument is that even if genotyping improves the initial time in therapeutic range, it may not have a significant impact on bleeding outcomes. COAG will directly address the first criticism, and GIFT will address the second criticism. The outcome of these trials will likely determine whether genotyping has an important role in the future of warfarin management, and it may have broader implications for the nascent field of pharmacogenetics.