Genetics May Help Fine-Tune Warfarin Dosage
Formula that includes DNA data proves faster, more accurate in early tests
WEDNESDAY, Feb. 18, 2009 (HealthDay News) -- A DNA test could make prescribing the widely used anticoagulant warfarin safer by predicting the lowest effective dose, a new study finds.
The U.S. Food and Drug Administration estimates that 2 million Americans take warfarin to prevent potentially dangerous blood clots that can cause strokes. Reasons range from implantation of an artificial heart valve to the abnormal heartbeat called atrial fibrillation. But establishing and maintaining the ideal dose is done by trial and error, with weekly and monthly monitoring.
"In a very large group of patients, we showed that the use of genetics helps predict the right dose, and this may take the dangerous guesswork out of prescribing warfarin," said study co-author Dr. Russ Altman, a professor of bioengineering, genetics and medicine at Stanford University School of Medicine. This "may be one of the first widely used examples of pharmacogenetics -- the use of genetic information to make better prescribing decisions," he said.
Because an effective dose of warfarin can vary as much as 10-fold from one person to the next, doctors start most people on a low dose and increase it until blood tests indicate the dose is correct. That process can take months.
For the study, which is published in the Feb. 19 issue of the New England Journal of Medicine, researchers collected demographic and genetic data on people at risk for stroke, heart disease and other problems related to blood clotting. The ideal doses of warfarin for these individuals had already been established by trial and error.
In addition, they looked at variants of two genes: CYP2C9, which affects how the liver activates and excretes warfarin, and VKORC1, which activates vitamin K, essential for blood clotting.
They used the data to create an algorithm that would predict the ideal dose of warfarin in a study of more than 4,000 people.
The researchers found that the algorithm predicted the ideal dose of warfarin better than other approaches.
Specifically, the algorithm's predictions were, on average, within about 8.5 milligrams of a person's ideal weekly dose. Demographic and clinical data alone predicted doses of about 10 milligrams off the ideal weekly dose, and a method that used a fixed daily dose, or about 35 milligrams a week, was off by an average of 13 milligrams, the researchers found.
"Warfarin has a very narrow therapeutic range," said co-author Teri Klein, a senior research scientist in genetics at Stanford. "Overdosing and underdosing of warfarin results in patients either being at much greater risk for excessive bleeding [overdosing] or clotting [underdosing]," she said.
"We found, based on this population, that 46 percent of the population would benefit by including their genotypes for these two genes," Klein said. "By including the genotypes along with clinical and demographic information, initial dosing is much closer to the ideal dose, thus limiting potential adverse events."
Altman said a trial now is needed to see if people who are prescribed warfarin based on genetic testing actually do better than those whose dosage is not determined with the aid of genetics.
"More broadly," he said, the study's finding "tells us that the day when we routinely use genetic information to personalize the use of drugs is coming and may be almost here for warfarin."
Dr. Richard C. Becker, a professor of medicine in the divisions of cardiology and hematology at Duke University School of Medicine, said that the researchers appear to be on the right track but that more testing is needed before the approach becomes standard practice.
"The safe and effective use of medications based on patient-specific needs is an achievable goal in the early 21st century," Becker said. "Genetic testing added substantially to traditionally employed daily dose prediction methods, particularly among patients with low dose requirements."
The combination of patient-specific characteristics and genetic variations was able to minimize the likelihood of recommending excessively high and potentially dangerous doses of warfarin, Becker said.
"While the potential benefit of warfarin management algorithms that employ genetic testing has not been established, and therefore cannot be recommended in daily clinical practice, accumulating evidence is provocative and supports further research," he said.
Another expert cautioned that using genetics to establish drug doses is still a long way off, however.
"There is an important clinical need to develop better means of determining appropriate initial dosing of warfarin for each individual patient," said Dr. Gregg C. Fonarow, a professor of cardiology at the University of California, Los Angeles.
"However, as this study is retrospective, it could not evaluate outcomes, and recent clinical trials have failed to confirm the expected benefits of other pharmacogenetic algorithms for warfarin dosing [so] prospective clinical trials are required to determine the value, if any, of the findings in this study," he said.
The U.S. National Library of Medicine has more on pharmacogenomics.