Roy Fagerberg, age eighty-two, of Chapel Hill, is among more than 1.5 million Americans taking the blood thinner warfarin. The typical starting dose is five milligrams a day, but he needs only three.
Two milligrams in difference are important when the drug has also been used as rat poison for fifty years.
Finding the right dose of warfarin for each patient is time-consuming trial and error, but a new study by an international group that includes UNC Eshelman School of Pharmacy scientists suggests that looking at a patient’s genes can speed up the process.
Warfarin is tricky: the dose needed to achieve the desired anticoagulant effect can vary by up to ten times from person to person. Getting too much or too little can cause dangerous bleeding or clotting. Currently, patients start with an average dose, which is then carefully adjusted over a period of weeks or months until the correct dosage is found. Entire clinics are devoted to this pursuit.
“The promise of genetic testing is getting the right drug in the right amount to the right person at the right time every time. . . . This study is a big step in that direction with warfarin.”
— Howard McLeod, director of the UNC Institute
for Pharmacogenomics and Individualized Therapy
Now, researchers from the International Warfarin Pharmacogenetics Consortium—a global collaboration of scientists hosted by the Stanford University-based research network, PharmGKB—have shown that taking two genes into account can result in a warfarin dose that is closer to being right the first time. The study appears in the February 19, 2009, issue of the New England Journal of Medicine.
“Our analysis shows that the pharmacogenetic test is better at identifying patients who lie at either the high end or the low end of the warfarin-dose spectrum,” said Michael Wagner, PhD, a professor in the School and a member of the paper’s writing committee. “That’s important because nearly half of the people taking warfarin fall into those extreme ranges, putting them at the greatest risk for excessive bleeding or clotting.”
Variations in two genes—cytochrome P450, subfamily 2C, polypeptide 9 (CYP2C9) and Vitamin K epoxide reductase complex 1 (VKORC1)—affect how much warfarin a person needs to achieve the desired effect. In 2007 the FDA added this genetic information to warfarin’s label but did not specify how the information could be used.
To begin developing guidelines, consortium members pooled clinical and genetic data from 5,052 patients taking warfarin from twenty-one centers around the world.
Because these patients were already on warfarin, the researchers knew the appropriate dose for each person. Then they created three different dosing strategies designed to recommend an initial dose for each patient. The first was a fixed dose of five milligrams. The second method factored in clinical data such as age, weight, and medical condition. The third combined the clinical information with the patient’s genetic profile.
The strategies then were compared to see which one best predicted an initial dose that was within 20 percent of the dose a patient was taking, a threshold that the researchers considered to be a meaningful difference.
For the 46 percent of patients requiring less than three milligrams or more than seven milligrams of warfarin each day, using genetic information resulted in a more accurate dose. For patients needing the smaller dose, the pharmacogenetic method made a correct prediction nearly 50 percent of the time, compared to the clinical information alone, which was accurate only one-third of the time. For those who required a larger dose, the pharmacogenetic guideline predicted the correct dose nearly 25 percent of the time, compared to the clinical method’s seven percent success rate.
“The promise of genetic testing is getting the right drug in the right amount to the right person at the right time every time,” said Howard McLeod, PharmD, director of the UNC Institute for Pharmacogenomics and Individualized Therapyand Fred Eshelman Distinguished Professor of Pharmacotherapy and Experimental Therapeutics in the School. “With a patient like Mr. Fagerberg, who can only tolerate a small dose, getting it right the first time is even more important. This study is a big step in that direction with warfarin.”
The next step is to determine whether more accurate initial dosing results in better health for the patient, he said.
“We’ve been dosing warfarin through trial and error for more than fifty years, with good success,” McLeod said. “We currently have a trial underway to see if using pharmacogenetic dosing initially improves a patient’s health long term. After all, better health for everyone is the ultimate goal of our research.”
Warfarin, commonly known under the brand name Coumadin, was initially developed as a pesticide against rats and mice. It was approved for medical use in the 1950s and is used to prevent the formation of blood clots. Physicians write more than 30 million prescriptions for the drug each year. Warfarin interacts with many other drugs and even some foods and herbs, such as garlic, ginseng and ginger.
“This research study has made an important advance toward personalizing medicine—it uses data from countries around the world to develop a gene-based strategy for warfarin dosing that could benefit a wide range of patients,” said Jeremy M. Berg, PhD, director of the National Institute of General Medical Sciences, which partially funded the study. “This is a wonderful example of international cooperation and the results are especially valuable for the United States, since our population is so genetically diverse.”