YouTube Facebook LinkedIn Google+ Twitter Xinginstagram rss  

Predicting Dose Response

By Kevin Davies

“All substances are poisons; there is none which is not a poison. The right dose differentiates a poison...”
— Paracelsus (1493-1541)

Sept. 13, 2007 | Last month, the FDA issued an important ruling regarding the labeling of warfarin (Coumadin), one of the most commonly prescribed drugs in the United States — but one where dosage is critical. The drug’s label will be revised to note the value of genetic testing in assessing patient dosage, which typically ranges from 1 to 10 mg/day.

“This means personalized medicine is no longer an abstract concept but has moved into the mainstream,” said FDA’s director of clinical pharmacology, Larry Lesko.

Warfarin is a potent anti-coagulant that was first synthesized in 1948 as a rat poison. The drug gained notoriety as a blood thinner when President Eisenhower took it in 1955. Today, more than 2 million Americans suffering from deep-vein thrombosis, heart attacks, and strokes take the drug. By one estimate, warfarin has saved more lives than any other single medication except penicillin.

But dosage mishaps result in about 40,000 ER visits per year (only insulin accounts for more adverse drug hospitalizations), and occasional fatalities. A recent Wall Street Journal article, recounting the saga of a St. Louis woman administered warfarin to treat blood clots in her lungs, concisely described the problem: “The medicine probably helped save her life. Then it almost killed her.” (Genetic tests performed after she was rushed back to hospital due to internal bleeding revealed she required a much lower dose than normal.)

We now know that warfarin activity is influenced by a pair of genes. In 2002, researchers found that two variants in the cytochrome P450 2C9 gene lower the metabolism of the drug, thus posing an increased risk for bleeding. Polymorphisms in the gene for the biochemical target of warfarin — vitamin K epoxide reductase complex 1, VKORC1 — also impact dosage.

In a report just out in Blood, Brian Gage and colleagues at Washington University in St Louis have produced an algorithm (see that factors genotypic data and medical factors to calculate and refine warfarin dosage. In a study of 92 orthopedic patients, Gage found that genotyping (particularly CP2C9) was important in predicting warfarin response, but so too were smoking status, blood loss during surgery, and other medical factors. Other trials are underway, including CROWN (CReating an Optimal Warfarin Nomogram), organized by Brigham & Women’s Hospital, which aims to develop a drug-dosing algorithm that is “superior than the current practice of educated guesses and trial and error.”

Warfarin joins four other drugs with newly introduced genetic language in the label — the chemotherapeutic agents 6-mercaptopurine, azathioprine, and irinotecan, and the ADHD drug atomoxetine. However, the FDA has not gone so far as to stipulate mandatory genotyping in the label’s black box. There has been considerable pushback from healthcare providers and insurers who are worried about liability, cost, and want to see proof — in the form of prospective clinical trials — of the effectiveness of the gene tests.

Even Gage, who says that the gene tests are most valuable in making dosing decisions for about a third of warfarin patients, concedes this is not unreasonable.

Lesko and the FDA are to be commended for actively pushing pharmacogenomics awareness in the medical community. There will be many more drug labeling issues confronting the agency and medical community in the coming years. But the tests are still not widely covered by health insurance firms, which deem them too “experimental.”

Ironically, the day of the FDA’s warfarin decision, the Washington Post reported a twist in the investigation of the cause of death of Jolee Mohr, an Illinois woman receiving an experimental adeno-associated virus (AAV) treatment for arthritis developed by Seattle-based Targeted Genetics. The Post revealed Mohr’s immediate cause of death as a massive fungal infection that seemingly capitalized on her compromised immune system.

Whatever the primary cause of Mohr’s infection proves to be — there is no proven link as yet to the AAV vector, and she was taking several other drugs at the time — it’s another setback for gene therapy, which has been beset with problems since the first such trial in 1990. Even recent successes in treating about a dozen patients suffering from rare immunodeficiencies have been tainted by several cases of leukemia among the treated patients.

Whether it’s a proven lifesaver or a leading-edge therapeutic, there is still so much about the body’s response to drugs that we do not fully understand. 500 years after Paracelsus, we’re still grappling to predict dose response.

Subscribe to Bio-IT World  magazine.

Email Kevin Davies.

Click here to login and leave a comment.  


Add Comment

Text Only 2000 character limit

Page 1 of 1

For reprints and/or copyright permission, please contact Angela Parsons, 781.972.5467.