Pharmacogenomics: The New Rx
How University of Chicago researchers are aiming to translate genetic findings into personalized medicine
Patients know the warning: "Contact your doctor immediately if you experience any of the following symptoms." It's the standard prescription-drug language, typically accompanied by a discomfiting list of possible side effects qualified by terms like "in rare cases."
Take the cholesterol-lowering simvastatin, the country's second most-prescribed medication. Roughly 1 percent to 2 percent of those who take it develop myopathy, a painful muscle injury that can lead to kidney complications and even death. And too high a dose of the blood thinner warfarin can lead to internal bleeding, while too little may result in a blood clot. "The way we practice medicine now is a bit antiquated, in that it's primarily trial and error," said University of Chicago Medicine hematologist-oncologist Peter H. O'Donnell, MD'03, assistant professor of medicine. "In a sense, you're treating every patient with the same disease in the same way. We can do better than that."
O'Donnell is on a team, led by hematologist-oncologist Mark J. Ratain, MD, that is trying to take medicine to the next level.
"Drugs can do more harm than good," said Ratain, the Leon O. Jacobson Professor of Medicine and director of the University of Chicago Medicine Center for Personalized Therapeutics (CPT). Recipient of a 2011 Translational Research Professorship from the American Society for Clinical Oncology's Conquer Cancer Foundation, Ratain has written extensively on individualizing cancer care based on genotype.
"What we're moving toward," he explained, "is tailoring drug therapies to maximize the benefit and minimize the risk."
"It's really the next frontier of pharmacology," said O'Donnell, who trained under Ratain in the University of Chicago's fellowship in clinical pharmacology and pharmacogenomics, one of the few such programs in the country.
With the recent addition of Yusuke Nakamura, MD, PhD, considered one of the world's top geneticists, as the center's deputy director and an ambitious project to make genetic testing a routine part of the doctor's visit, the center is leading the charge to translate pharmacogenomics from the lab into the clinic. Ratain's CPT team also includes associate directors Nancy J. Cox, PhD, professor of human genetics, and M. Eileen Dolan, PhD, professor of medicine; and assistant director Michael Maitland, MD, PhD, assistant professor of medicine.
From Pharmacology to Pharmacogenomics
When the human genome was first sequenced a decade ago, it opened up the promise of personalized medicine like nothing before it. "As genomics became more accessible, we got into looking at variability at a genetic level," said Ratain. The ability to peer inside an individual's DNA quickly gave birth to pharmacogenomics, the study of how someone's genotype can influence his or her drug response. As researchers have discovered, small differences in pieces of one's genetic code can predict whether someone is likely to benefit from a medication -- or suffer dangerous side effects.
Unlike disease-risk genetics, which focuses on whether individuals' genotypes make them more susceptible to developing, say, breast cancer or Alzheimer's, pharmacogenomics concentrates exclusively on genetic markers that interact with and respond to medications.
The Center for Personalized Therapeutics' flagship pharmacogenomics initiative is known as The 1200 Patients Project and is the largest-scale clinical study of its type to date. The study, said O'Donnell, is designed to pinpoint the most effective ways to incorporate published pharmacogenomic data into medical decision-making. investigators are collecting DNA samples from 1,200 patient participants. These are being tested for hundreds of DNA markers potentially impacting drug responses or side effects and then a personalized pharmacogenomics profile is created for each participant. That information is then immediately accessible to the patients' physicians through a secured informatics database.
"This is really the first broad approach at implementation," explained Ratain. The project is aimed at laying the groundwork for more tailored care, lowering medical costs and improving outcomes. It meets a particularly pressing need in oncology, where, as Ratain and O'Donnell wrote in Molecular Oncology this past January, the window between therapeutic and toxic for drugs "is often narrow, the need for favorable drug response is often acute and the consequences of drug toxicity can be life-threatening."
Ratain, for example, got his start in the field studying irinotecan, a drug used to fight colorectal cancer. Irinotecan causes severe diarrhea and neutropenia, a decrease of white blood cells, in 20 percent to 35 percent of patients. In the early 2000s, Ratain and colleagues discovered that patients with a common genetic variation in their UGT1A1 enzyme were most likely to suffer the deleterious effects. Their findings led directly to the Food and Drug Administration's 2004 decision mandating that the label for irinotecan be amended with pharmacogenomic data alerting physicians to the increased risk. Ratain's team also developed and patented a test that determines which patients are likely to have a serious adverse reaction to the drug.
As part of their ongoing research, Ratain and O'Donnell are actively seeking out genetic biomarkers that shed light on how cancer patients will respond to different treatments. A recent University of Chicago study identified single nucleotide polymorphisms (SNPs), sections of variation in DNA's building blocks, that play a role in predicting head and neck cancer patients' response to platinum-based chemotherapy. The study, by O'Donnell; head and neck cancer specialists Everett E. Vokes, MD, physician-in-chief and chair of the Department of Medicine, and Ezra Cohen, MD; and several investigators from the CPT team and Department of Human Genetics, was published in the May 2011 Translational Research.
Cancer-fighting drugs are just one group among hundreds of medications that have been linked to genetic biomarkers during the past decade. Warfarin, one of the most researched drugs in the field, is another. Patients with a particular genetic variation will eliminate the anticoagulant more slowly from their system and, thus, require a lower dosage. There's even a commercially available test to find out whether a patient falls into that group, noted O'Donnell.
So why aren't more physicians using genetic screening before prescribing such drugs?
Even though genetic testing has existed for more than a decade (and become increasingly inexpensive), roadblocks have largely kept it out of the clinic. Among the most significant barriers: lack of knowledge and lack of accessibility.
"There is tremendous potential for this information to improve patient care. My colleagues at UChicago have recognized that one of the bigger hurdles is just to get clinicians familiar with information that already is available," Maitland said.
"Because this field is still relatively new," O'Donnell said, "a lot of physicians aren't familiar with how to use pharmacogenomic information." A 2009 Medco Health Solutions survey of more than 10,000 U.S. physicians found that 98 percent agreed with the central premise that patients' genetic makeup influences drug therapy, but a mere 10 percent felt sufficiently informed about pharmacogenomic testing. Only 13 percent reported ordering a genetic test in the prior six months.
"Even if a physician orders the test and wants to consider it, they might not know what to do with the result," O'Donnell said. "So a major part of The 1200 Patients Project is really trying to translate this information into something clinically meaningful."
There's also the practical issue of expediency. Right now, said O'Donnell, "even if you do the test, it could be days -- even weeks -- before you get results. But your patient needs a prescription now. Waiting on a genetic test isn't practical."
That's why The 1200 Patients Project emphasizes pre-emptive testing. When a patient enrolls in the study, he or she gives a one-time DNA sample via a simple blood test. The sample is analyzed for hundreds of gene sections known to influence either positive drug responses or negative side effects, and then the patient's information is stored in the secured database.
A physician considering prescribing simvastatin for high cholesterol, for example, might consult the system and find out the patient has a genotype with an increased risk for developing simvastatin-induced myopathy. Knowing that, the physician could choose from other cholesterol-lowering drugs to treat that particular individual.
In addition to highlighting drug risks, pharmacogenomics can help identify drug benefits. For example, a physician might use genetic testing and discover that someone with heart disease carries a marker predicted to respond well to a specific type of beta-blocker.
To create the informatics database, O'Donnell and his team conducted an extensive review of a decade's worth of pharmacogenomics findings, filtering out all but those backed up by the most rigorously designed studies. While a handful of medical centers around the country are doing similar projects, the University of Chicago's is the widest ranging, assessing patients for literally hundreds of possible drug interactions.
"Our scope and our size are unique," O'Donnell told Nature Medicine in November. The project encompasses many drugs commonly prescribed in the primary care setting, including various medications for hypertension and cholesterol. On the roster: hydrochlorothiazide, atorvastatin, amlodipine, simvastatin and others.
Open to any adult receiving ongoing care at the University of Chicago Medicine, the multiyear study has already enrolled more than half of its projected 1,200 participants since launching in January 2011. With a pilot group of 12 physicians whose diverse specialties include cardiology, primary care and oncology, the project, O'Donnell said, "really spreads across all disciplines, all drug classes, all diseases."
The idea, added Ratain, "is to do this across the medical center. That's our mission."
The initiative already has made significant strides toward personalized treatments through its custom-designed informatics. Bringing together a patient's genetic test results and relating those to clinical, published knowledge about pharmacogenomic relevance, explained O'Donnell, the portal gives physicians instant access to drug risks or benefits specific to an individual's genotype.
"If everyone could be tested up front for these genetic markers," said O'Donnell, "then physicians could reference them immediately and use them to inform prescriptions. What we've done, essentially, is encapsulate patient-specific pharmacogenomic information in a 30-second summary that translates the genetic findings into what they mean clinically. We've included all that information in the system."
The hope is that by providing busy clinicians with this information quickly and concisely, their patients will begin to see direct benefits from this genetic research.
This story originally ran in the Fall 2012 issue of Medicine on the Midway, a publication of the University of Chicago Medicine and Biological Sciences Division.
» Read the latest issue.