Dan Roden, M.D., left, and C. Michael Stein, M.B., Ch.B., are studying how genetic information can improve dose prediction for the blood-thinner warfarin. (photo by Dana Johnson)
Genetic data helps improve warfarin dosing
Genetic information improves dose prediction for the blood-thinning drug warfarin, a multi-center research consortium including investigators at Vanderbilt reports this week in The New England Journal of Medicine.
Warfarin is challenging to prescribe because individual responses vary widely and the wrong warfarin dose can be dangerous — too much can cause serious bleeding; too little can cause life-threatening blood clots.
The current findings “represent a step forward for personalized medicine — the idea of using a combination of clinical and genetic data to tailor therapies to 'fit' individual patients,” said study contributor Dan Roden, M.D., assistant vice chancellor for Personalized Medicine.
One question now is whether using genetic data to predict warfarin dose “makes any difference” in terms of clinical outcomes, said C. Michael Stein, M.B., Ch.B., associate director of the Division of Clinical Pharmacology, who also contributed to the study.
A large clinical trial, supported by the National Institutes of Health and set to begin next month, will test the impact of gene-based warfarin dosing on patient outcomes and determine if a more tailored prescribing approach leads to clinically meaningful improvements in patient care. Vanderbilt will participate in the trial.
More than 2 million people in the United States take warfarin, the generic name for Coumadin, to prevent blood clots after a heart attack, stroke or major surgery, or because they have atrial fibrillation, artificial heart valves or deep venous thrombosis.
But warfarin dosing is tricky. Because patients' responses to the drug vary widely, it's difficult at the beginning of therapy — when the risk of complications is highest — to find the dose that will prevent clotting without causing serious bleeding, for example.
To start warfarin therapy, doctors generally select a “standard” dose, or they modify the dose to account for clinical and demographic factors (such as age, weight and other medications) that affect a patient's response to warfarin.
After a few days, they measure the patient's blood-clotting time and make dose adjustments.
This trial-and-error process takes time, and patients who are outside the “therapeutic window” are at increased risk for bleeding and clotting complications.
Scientists have known for several years that genetic variations — particularly in two genes called CYP2C9 and VKORC1 — also affect individual responses to warfarin. In 2007, the Food and Drug Administration updated warfarin labeling to notify doctors that genetic testing could improve initial dose estimates for individual patients. But the new labeling didn't include instructions about how to use the genetic data.
“When the genetic testing came along, there was an impetus to figure out how we optimize all the information we know about what affects warfarin dosing,” Stein said. “To do that, you need data from a huge number of patients — a very broad spectrum of patients with different backgrounds.”
To explore methods for using genetic information in warfarin dosing, investigators from 21 research groups in nine countries formed the International Warfarin Pharmacogenetics Consortium (IWPC).
The IWPC members contributed clinical and genetic data for a total of 5,700 warfarin-treated patients.
Using a subset of these patients, the investigators developed formulas for determining warfarin dose using clinical data alone or using both clinical and genetic data. Then, they compared the predicted warfarin dose from their formulas — or a fixed dose — to the actual stable therapeutic dose in another subset of the patients.
They found that the formula that included both clinical and genetic data more accurately predicted the stable warfarin dose than the clinical-alone formula or the fixed-dose approach, especially for patients at the low or high ends of the dosing range — patients who are at increased risk for bleeding or clotting complications.
“This work provides physicians with information that will help them determine the appropriate dose of warfarin for their patients,” Stein said.
The consortium's formula is available on the Web site www.warfarindosing.org.
Other Vanderbilt participants in the IWPC include Marylyn Ritchie, Ph.D., Ute Schwartz, M.D., and Daniel Kurnik, M.D. The NIH supported the research.
Stein is the Dan May Professor of Medicine. Roden is the William Stokes Professor of Experimental Therapeutics and the director of the Oates Institute of Experimental Therapeutics.