Genomic testing for warfarin therapy: Why a promising approach stalled out

On August 16th, 2007, an FDA press conference on warfarin made history. Lead speaker Lawrence J. Lesko, PhD, at that time director of the agency's office of clinical pharmacology, spoke with passionate conviction about the FDA label change recommending genomic testing for new-start warfarin therapy.

"Today is a significant event," he said, "for those who foresee a day when medical care will be tailored to fit the unique genetic makeup of every single one of us."

The anticipated rush by anticoagulation prescribers to embrace genomic testing was encouraging, and yet while test orders immediately soared, the numbers fell flat in nearly the same record time, deflating the hopes of FDA and many others.

The barometer of adoption

"This is an important topic to pursue," said Edward Abrahams, PhD, president of the Washington, D.C.-based Personalized Medicine Coalition. "When the FDA relabels a drug, it doesn't necessarily lead to a change in the medical process, even if the evidence suggests it should."

Abrahams cited one warfarin therapy observational trial showing that "by doing genetic testing, they had 30% fewer hospital admissions from adverse events." However, he added, the nature of observational methodology rendered it "not as clear-cut."

Since the initial FDA warfarin label change, Abrahams has watched the debate over whether or not to test focus on the level and quality of evidence.

In "Pharmacogenomics of warfarin: Uncovering a piece of the warfarin mystery," a 2009 article published in the American Journal of Health-System Pharmacy by Michael P. Gulseth, PharmD, BCPS, FASHP, Gloria R. Grice, and William E. Dager, the authors examine the validity of genomic testing evidence across diverse scenarios.

While Gulseth offered sound argument supporting genomic testing in warfarin therapy, he concluded that there is "simply not yet enough compelling evidence from large, randomized trials" to make "genotype-based therapy the standard of care."

According to Abrahams, the performance of large-scale randomized controlled trials for an inexpensive product such as warfarin, which has been in use since 1954, would be too expensive, and it is unlikely that any company will seek to undertake that investment.

Inevitably, Abrahams said, this indicates a need to alter perceptions of what is "the acceptable level of evidence," in a genomic age.

Where do we go now?

While Gulseth is "firmly in the camp of pharmacogenomic testing," he acknowledged that it is a big leap from testing patients and knowing their genetic variances to using that data for new-start warfarin prescribing.

"I don't think we need to go down that road just yet," he said. "Practitioners need to be trained, and they may have a false sense of security from test results. If someone thinks [based on testing] they should delay warfarin therapy, it is not a good idea." He added, "The pharmacogenomic algorithms do a good job of incorporating variables, but people get fixated on algorithms and don't adjust the dose when it is needed, for example in acutely ill patients."