Consider the rat-poison-turned-blood-thinner warfarin, also known as Coumadin. Too much and a patient can bleed to death, too little and he can throw a clot and have a stroke.
Given the considerable consequences of improper dosing for patients-and health insurers-researchers are looking intently into how a person’s genetic underpinnings determine how drugs are metabolized.
And it turns out Indiana is becoming not only a hotbed of “pharmacogenomics” research, but also a trailblazer in finding practical ways to use it on the practitioner level.
“We’re looking at a paradigm shift in health care,” said Michael D. Kane, lead genomic scientist at Bindley Bioscience Center at Purdue University. Kane recently helped found Genomic Guidance LLC, which plans to bring low-cost DNA screening on the clinical level to help predict how a person is likely to respond to anticoagulants and other drugs.
The field has been researched at Purdue and Indiana universities for years, but came into public view last year when pharmacy benefits manager Medco Health Solutions Inc. announced it would add a pharmacogenomics research center at its $140 million automated pharmacy under construction in Whitestown.
Medco will call it the easier-to-pronounce Personalized Medicine Research Center. With collaboration from leading university researchers in Indiana and elsewhere, the center will study genetic and clinical drug testing and medical claims data. The center will be overseen by
Felix Frueh, a genomics authorityformerly at the U.S. Food and Drug Administration.
The New Jersey-based pharmacy benefits manager envisions patient files including a genetic profile related to how quickly the patient metabolizes a particular drug. That data could be referenced when a person is prescribed a blood thinner, pain reliever or cancer drug.
A physician could be told, “Did you know your patient is a poor P2C metabolizer?” said Medco spokeswoman Ann Smith, referring to an enzyme the body uses to metabolize a drug. Genetic information “can be used against a myriad of drugs.”
Medco and Mayo Clinic already are studying genetic considerations in warfarin dosing. And Medco and Burlington, N.C.-based LabCorp are studying breast cancer patients who take tamoxifen.
Some pharmacogenomics research also involves identifying whether genetics make some people more likely to suffer side effects when they take two kinds of drugs, or whether one drug cancels out the other.
For example, Indiana University pharmacogenomics researchers in recent years found that some women with a particular genotype may not receive the full benefit of the breast cancer drug tamoxifen while taking antidepressant paroxetine-better known as Paxil.
The focus on patient genotypes is intensifying at Eli Lilly and Co. under the moniker “tailored therapies.” For example, subsidiary ImClone Systems has zeroed in on genes of certain patients for which its Erbitux cancer drug is most effective.
It’s all part of a trend away from the focus on blockbuster drugs to developing versions of a particular drug tailored to patients with varying metabolism types, said David Johnson, president and CEO of BioCrossroads, the Indianapolis-based life sciences industry group.
One of the best examples is the breast cancer drug Herceptin, Johnson said. Its maker, San Francisco-based Genetech Inc., focused on a population of women with a special receptor-a relatively small population of women. But for this group, the drug has performed exceptionally well in turning off cancer cell production.
“Everybody is looking at whether we could really deliver more drugs like that,” Johnson said.
And the drug that works tends to be the one insurers will pay for, he noted.
Progress, at a cost
But all of these tailored therapies require capturing a patient’s genetic information, and the costs and logistics may prove more difficult than some of the science involved in your average single-nucleotide polymorphism.
A study published early this year in the Annals of Internal Medicine on using pharmacogenomic information in warfarin patients concluded that, at current costs, routine genotyping before warfarin dosing is not cost-effective.
The University of Cincinnati Medical Center, Massachusetts General Hospital and Washington University in St. Louis concluded there would be “minimally improved” survival rates. Yet costs would rise dramatically, at $172,000 per “quality-adjusted life year” versus the “generally accepted societal threshold” of $50,000.
Purdue scientist Kane argues that test cost assumptions in the study are high. The company he co-founded in January, Genomic Guidance, is working on a one-time DNA screening for anticoagulants that will provide results in less than 90 minutes and cost less than $30.
“None of this is going to happen until it’s a value proposition for everybody” in health care, Kane said. He contends that Genomic Guidance’s on-site DNA screening system and integrated data management will meet that test.
Genomic Guidance is based at Ohio Northern University, in Ada, Ohio, but the software to provide doctors and pharmacists with predictive drug efficacy and toxicity based on a patient’s genotype was developed by Kane and Purdue information technology experts John Springer and Nick Iannotti. Their “GeneScription” system was developed under a grant from Microsoft Corp.
But cost isn’t the only obstacle-so are privacy concerns. Genetic testing is still a dicey area, with some detractors concernedthat insurance companies will use information to determine one’s predisposition for certain diseases and jack up rates or reject those in higher risk categories.
Public concern has been so strong that Congress acted. Last May, then-President Bush signed into law the Genetic Information Nondiscrimination Act, which prohibits U.S. insurance companies and employers from discriminating on the basis of information derived from genetic tests.
Dodging privacy concerns
But the kind of DNA screening Kane is looking at does not attempt to determine disease risk. It looks at whether a person harbors one or more single nucleotide polymorphisms shown to alter the rate of drug metabolism and the efficacy of prescription drugs.
People concerned about privacy won’t have to have their results placed in medical records. They could walk out of the clinic with a paper printout or have the information placed on an electronic data card.
Kane said the system could be placed first in a pharmacy, with the genetic information considered by the pharmacist while conducting the usual review to see whether the medicine will interact with other medicines the patient is taking.
If the genetic screening software indicates the prescribed dosing might not be appropriate, the pharmacist could alert the patient or doctor.
It remains to be seen how physicians and other medical providers already weary of health insurers’ telling them how to practice medicine would react to such pharmacogenomic guidance. •