A Vanderbilt study has found that established patients receiving routine clinical care in any of four outpatient clinics at Vanderbilt University Medical Center have a 64.7 percent chance, over a five year period, of receiving one or more of 56 drugs with known pharmacogenetic associations — drugs whose labels include special information for people who carry one or another common genetic variants.
These so-called medical home patients were found to have a 12 percent chance of receiving four or more of these 56 drugs.
The study, the first to assess the potential benefits of prospective genetic testing for drug safety, will appear in an upcoming issue of Clinical Pharmacology and Therapeutics (and is available now on the journal’s web site). The question posed in the study concerns the degree to which VUMC patients in general, including patients who currently may not need even a single drug, would benefit from a one-time genetic test measuring their odds for detrimental responses to a range of drugs.
Led by Jonathan Schildcrout, Ph.D., associate professor of Biostatistics and Anesthesiology, and Josh Denny, M.D., M.S., assistant professor of Biomedical Informatics and Medicine, the study also looks at life threatening or otherwise severe adverse events stemming from known gene-drug interactions. The researchers estimate that, over five years, six drugs could account for 383 such events among some 52,942 VUMC outpatients.
Those six drugs include common medicines like blood thinners and cholesterol-lowering statins. The adverse events include clinically significant bleeds, muscle toxicity, heart attacks, strokes and death.
That estimate of 383 such events is based on drug documentation found in the medical record, rates of genetic variation in the general population and associated adverse event rates found in the literature.
Those 383 events work out to a 0.7 percent chance of significant harm for the VUMC medical home population, over five years, from any of six presumably properly prescribed drugs.
“I think that’s high,” Denny said. “That’s real events in a population that wasn’t preselected for risk of medication exposure. That’s for the average person who comes into our clinic and gets seen multiple times. At that rate, over five years a full-time primary care physician could expect to avoid 10 to 15 adverse events. It’s higher than I expected, especially when you’re just considering six drugs.”
These six are only a subset of the drugs known to pose excess risk of severe adverse events for patients who carry common genetic variants, Denny said.
It’s no accident that this study arose at Vanderbilt. No other medical center is doing more to bring pharmacogenetics into day-to-day clinical decision-making.
Under a program called PREDICT (Pharmacogenomic Resource for Enhanced Decisions in Care and Treatment), VUMC has begun piloting use of a prospective genetic test for drug responses. The test is ordered based on each patient’s likelihood, over a three-year horizon, of receiving certain drugs.
The retrospective study follows patients who, anytime between January 2005 and June 2010, were seen at least three times in any 24-month period by providers in any of four Vanderbilt outpatient areas — Primary Care, Nephrology, Cardiology or Diabetes. Using natural language processing, the researchers accurately ferreted out drug documentation in the electronic medical record for 52,942 such patients.
The list of well-documented adverse gene-drug interactions is growing apace, and during the time since the study was completed the number of drugs whose labeling bears FDA-stipulated pharmacogenomic information has grown to 112, up from the 56 that were tracked in the study.
“As the list of pharmacogenetic effects grows, including serious effects, and as we learn about alternative treatments for those at increased risk, so strengthens the argument for a preemptive genotyping program,” Schildcrout said.
For more information about PREDICT, providers can call 875-4363 (875-GENE). Patients can call 343-9188, or toll-free to (866)-886-2478.
