The Art and Science of Medicine
Volume XXVI Number 1
It's no secret that people respond differently to drugs--the arthritis medication your neighbor calls a wonder drug may do nothing for you or your uncle. As a result, physicians and their patients have long known they may have to slog through some of their own "clinical trials" before they find a drug or dose that works.
Now, the deciphering of the human genome promises an era in which drugs and patients can be much better matched in advance--the era of pharmacogenetics.
Or, as David Flockhart prefers to call it, the era of precision prescribing.
David Flockhart, chief of the Division of Clinical Pharmacology in the Indiana University School of Medicine's Department of Medicine, is helping make the Medical School a leader in pharmacogenetic research and the clinical practice accompanying that research.
He arrived at the IUSM in the summer of 2001 from Georgetown University, bringing with him seven researchers, swiftly establishing IU as a site in the Pharmacogenetics Research Network organized by the National Institute of General Medical Sciences, part of the National Institutes of Health.
He also swiftly set about creating pharmacogenetics programs and fellowships in pediatrics, psychiatry, cancer, and bioethics at the School of Medicine. It's an effort prompted by a longstanding interest in how drugs act differently in different people.
The way drugs work
Whenever we take a medicine, our bodies go to work on it. Often the drug must first be broken down and converted to a useful form called a metabolite. Then it must be delivered to places in the body where it's needed, and eventually it must be eliminated. How each of our bodies does this depends on our individual genetic makeup.
There are genes that affect whether and how fast a drug is metabolized, genes that play roles in transporting drugs to the appropriate sites in the body, genes that determine whether the cells have the necessary receptors to latch on to the drugs once they arrive, and more.
Take codeine, for example, the world's most commonly prescribed opiate painkiller. Once codeine enters the body, it is broken down by an enzyme--CYP 2D6--and converted into morphine. The morphine brings the pain relief.
But not always. About seven of every 100 people have a gene that makes a form of CYP 2D6 that won't convert codeine to morphine. These individuals get nausea more than pain relief. They may complain that they're not feeling better and be accused of being fakers and drug abusers.
In another example, some children being treated for leukemia can suffer life-threatening shortages of bacteria-fighting white blood cells because they have an altered form of a gene that makes an important enzyme, called thiopurine methyltransferase. For situations such as these, pharmacogenetic tests are already available, says Flockhart: "The actual science behind pharmacogenetics is more compelling than it has ever been." Which is why, he says, the test for the enzyme should be standard care. But many physicians aren't aware of the test, and health insurers often balk at the $300 to $500 cost.
The time has come, Flockhart believes, to start doing clinical trials to determine whether--and which--pharmacogenetic tests should become routine. "In 10 years, we will have done a series of prospective trials that prove to all--scientists, government, educators, insurance and medical plan people--that pharmacogenetics tests are medically important and economically worth it, that they save lives or money," he says.
Better drugs, basic right
Flockhart came to issues of drug safety and pharmacogenetics through a process that might be called geographic. A native of Edinburgh, Scotland (where his family still lives), Flockhart got his doctorate in pharmacology at the Welsh National School of Medicine. He came to the United States for postdoctoral work at Vanderbilt University, then decided to go to medical school "to make the stuff relevant."
Flockhart attended medical school at the University of Miami, a school and a city whose populations were, he notes, as ethnically diverse as Scotland's was not.
"You're sensitized to ethnicity in Miami," he says. He was also sensitized to human rights issues there, after meeting people from Latin America whose families were suffering at the hands of despots.
After arriving at Georgetown University in Washington, D.C., for his residency, Flockhart's interest in human rights pushed him in two directions. He served several years as a board member of Amnesty International USA, and he began to study how race and ethnicity affected responses to drugs. Eventually, he says, the time pressures of laboratory work forced him off the AI board, but he "took that energy and threw it into research."
As a clinical pharmacologist, Flockhart views issues of pharmacogenetics, drug safety and side effects, prescribing errors and drug interactions, as interrelated. Public interest in those issues is being driven by publicity over the sequencing of the human genome. With the genome in hand, he says, the public will expect and demand--through elected representatives, if need be--that prescription drugs be safer and more effective.
It's an expectation that may be more strongly held among patients than their physicians. Patients, says Flockhart, almost always expect there will be side effects from the drugs they are prescribed. Doctors, on the other hand, often do not, because they are focused on the therapeutic benefit they expect the drug to provide.
It's a problem with the potential to get worse, because technology has dramatically improved our ability to diagnose ailments, while the ability to prescribe well in response to those diagnoses has lagged behind. In large part, that's simply due to "more stuff," says Flockhart: there are more and more drugs.
He has two answers to this problem. First, teach medical students well about drug therapies.
"We offer it here at IU, but it's not broadly present in most medical school curricula. It's an aspect of medical education that needs a lot of thought and resources now," Flockhart says.
Second, practicing physicians need to make a conscious effort to develop what he calls a "personal formulary" of drugs. Most physicians need to be familiar with only a relatively small number of drugs they use regularly, but they should be intimately familiar with them. Then they need to develop a systematic way to determine whether they should add a new drug to their personal formulary.
"Most doctors already have a formulary in their head, they have 75 percent of the knowledge they need. They just need to educate themselves the rest of the way," says Flockhart.
Pursuing precision prescribing
In the meantime, Flockhart and his colleagues are adding substantially to the further education physicians need.
One of Flockhart's research efforts involves studying how the body makes use of tamoxifen, which is prescribed to many women to reduce breast cancer risks. Many breast cancer cells have receptors that are stimulated by the female hormone estrogen. The goal of tamoxifen therapy is to block those cells from latching on to estrogen.
Flockhart and his colleagues have developed more sensitive tests of the activity of tamoxifen and its metabolites--the converted forms of the drug that appear to do the real work of preventing cancer. "We believe we understand the metabolism of tamoxifen better than anybody," he says.
According to Flockhart, his group has discovered a new metabolite of tamoxifen, one they believe is the primary actor in the body. This newly discovered metabolite is created, however, by the same enzyme involved in codeine metabolism, which suggests that perhaps 7 percent of women taking tamoxifen are not gaining anything from it, other than a false sense of security.
And there's more: Women who take tamoxifen often are prescribed antidepressant drugs to combat hot flashes. Flockhart is looking at the possibility that these same drugs may be blocking the action of the active tamoxifen metabolite.
Work is also underway in other School of Medicine pharmacogenetic initiatives, such as in pediatrics.
Drugs in general have not been adequately studied in children, and the details of how genetics affects the activity of drugs in young patients are particularly ripe for research, say IU visiting professors of pediatrics Kathleen Neville and Jamie Renbarger.
Renbarger is initially attempting to find genotypes that affect how children respond to vincristine, a chemotherapy drug used to treat a variety of pediatric cancers. Neville is conducting similar studies in sickle cell disease, looking at ways genetic factors affect how children respond to codeine used to treat the pain of sickle cell crises.
In psychiatry, Christopher McDougle, chair of the Department of Psychiatry, notes that current psychiatric drugs are effective in up to 70 percent of patients, depending on the drug and disease. But that still leaves nearly a third of all patients who don't benefit. The goal, then, is a genetic test that would predict if and how a patient will respond to a particular drug. "Wouldn't it be nice if we could do that?" he muses. "I think eventually we will get there, but I wouldn't want to be overly optimistic."
Even if identifying the appropriate genes to determine a patient's reaction to a psychiatric drug is easier than finding the genes associated with the disease itself, it still will be a tough chore, McDougle says.
Flockhart notes that eventually there is likely to be a standard set of 100 or 200 specific changes in genes that are relevant for prescribing drugs. "But in order to get to that point," he adds, "we need to decide what that standard set is."