The Right Fit
"All things are poison and nothing is without poison. Only the dosage distinguishes the killer from the cure." (loose translation from the original German)—Paracelsus, Swiss scientist (1493–1541)
If that sounds crazy to you, consider the case of the blood thinner warfarin. Now one of the most widely prescribed drugs in the world, warfarin was originally marketed—and is still commonly used—as a rat poison.
Killer or cure? The difference is in the dose.
But it gets more complicated: A safe dose for one person might be dangerous—even lethal—for another.
So how do doctors know how much of a medicine to prescribe? Essentially, they make an educated guess. Then, depending on how well the patient responds, they might adjust the dose.
Unfortunately, for very sick patients—or for very strong drugs—the delay caused by this trial-and-error process can be harmful or even life-threatening.
Julie Johnson, a clinical pharmacist at the University of Florida in Gainesville, hopes to speed things up, getting the right prescription to each patient right away. To do this, she focuses on genes.
"The hope is that through a person's genetics, we can minimize the trial-and-error process and quickly identify the drug therapy that will work best for that person," Johnson says.
The ultimate goal is to enable doctors to tailor prescriptions for each patient.
This area of research is called pharmacogenetics or pharmacogenomics. Johnson's team is one of a dozen groups that are part of a nationwide pharmacogenetics research network (see "Genes, Disease and Drugs").
Steering in the Right Direction
Before she landed in her current career, Johnson went through her own trial-and-error process. She was raised in rural Ohio, where her parents had a small farm.
While growing up, she was very active in 4-H and showed beef cattle every summer at the county and state agricultural fairs. She even won the Grand Champion award for her steer when she was a senior in high school.
Although raising cattle may seem far afield from medical research, it actually taught Johnson skills that help her excel in the laboratory, says Deanna Kroetz, a fellow scientist (and pharmacogenetics network member) with whom Johnson has been close friends for more than 30 years.
"Julie has been doing long-term projects and setting goals since she was a kid. It contributes to how she works on things, how she thinks," says Kroetz.
As a girl, when Johnson thought about what she wanted to be when she grew up, she looked to the careers of her family and neighbors. She considered being a kindergarten teacher, veterinarian, hospital pharmacist or drugstore owner. For one reason or another, none of these was a good fit. Eventually, she considered being a faculty member in a college of pharmacy.
While studying pharmacy in college, she took an elective class in research.
"Much to my surprise, I really, really liked it," she says. "It fit. It was intellectually stimulating and allowed me to address clinically important questions."
What's Genetics Got to Do With It?
Johnson's continued interest in medical issues led her to focus on the pharmacogenetics of cardiovascular drugs.
In addition to determining whether you will be tall or short, black-haired or blond, your genes influence how your body responds to medicines.
But genes aren't the only factor. Your age, weight, lifestyle and other characteristics also play a role.
Here's how it works. When you swallow a pill, it lands in your stomach and soon moves to the small intestine. From there, it is absorbed into nearby blood vessels, then carried to your liver.
Among its many functions, the liver is your body's main toxic wasteprocessing plant. It is chock-full of enzymes that metabolize drugs, alcohol and toxins, changing these substances into new chemical forms.
Within the liver is a large family of enzymes known as cytochrome P450s, or CYPs (pronounced "sips"), which together are responsible for metabolizing about 75 percent of all medications.
Some CYP enzymes change toxic compounds into harmless ones. Others chemically alter substances to prepare them for elimination in urine or feces. Still others convert drugs into their active form.
There are five main CYP enzymes involved in drug metabolism, each of which comes in many variations.
Every person has a unique combination of CYP enzymes, genetically selected from countless possibilities. Whether your combination is advantageous, pharmacologically speaking, depends on which medications you take.
Take a CYP
Take for example CYP2D6, an enzyme that is responsible for processing about one-fourth of all prescription drugs. This enzyme has more than 100 versions, based on tiny differences in the genes that code for it. Depending on which versions you inherit, your CYP2D6 enzyme activity could be normal, superfast or nonexistent.
Why should you care what kind of CYP2D6 activity you have? Because it could make a big difference in how well medicines work for you.
Say you broke a bone or had surgery. A doctor might prescribe codeine for your pain. In order to work, codeine needs CYP2D6 to transform it into the potent painkiller morphine.
If you have nonexistent CYP2D6 activity (along with about 5 percent of Americans), codeine will do nothing for you because your body can't convert it into morphine. Your pain will just keep throbbing.
But having too much CYP2D6 activity could have even worse consequences. A few years ago, a 2-week-old baby boy died from a morphine overdose.
Doctors were baffled: How could this happen to a healthy, wellcared-for baby who was not on any medicines?
Investigators discovered that the baby received morphine through the breast milk of his mother. She was taking codeine for surgical pain associated with the delivery.
The amount of morphine in the baby's blood was about 50 times higher than what is typical for babies whose mothers take codeine while breastfeeding.
The boy's mother, as it turned out, was an ultra-rapid metabolizer of codeine. Her body readily turned codeine into morphine, which then passed into her breast milk and accumulated in the body of her newborn.
Yet being an ultra-fast metabolizer can be beneficial if you are taking tamoxifen, a breast cancer drug.
"Hooray for CYP2D6 Ultra Rapid Metabolizer!!!" posts a blogger known as Fearless, who at the age of 37 was diagnosed with invasive breast cancer.
As Fearless so clearly understands, CYP2D6 is necessary to change tamoxifen into its active form, endoxifen. For people with little or no CYP2D6 activity, tamoxifen doesn't do much good.
But for those who are ultra-rapid metabolizers, tamoxifen is efficiently transformed into endoxifen. Or, as Fearless puts it, "Hooray!"
Still, as is often the case, genes don't tell the full story. Some foods and medications affect drug-metabolizing enzymes as significantly as do individual genetic variations.
A classic example is grapefruit juice. A substance in it can increase the potency of some drugs to dangerous levels. But for other drugs, the compound can have the opposite effect, preventing absorption so the drug cannot benefit the patient.
Working Out Warfarin
Now back to warfarin, the rat-poisonturned-blood-thinner drug. Individual responses to warfarin are affected by genetics.
Every year, an estimated 2 million Americans start taking warfarin, mostly to prevent blood clots that could cause a stroke or heart attack.
Warfarin is a touchy drug. If the dose is too high, a person could bleed to death. If it is too low, a potentially fatal blood clot could form.
And here's the kicker: the ideal dose varies widely—one person may require 10 times more than another—so it's nigh impossible to get every prescription right the first time.
How do doctors even know where to start?
Typically, they begin with a generic dosage adjusted for factors like the patient's weight, age and gender. Then they wait up to a week, check the patient's blood for its clotting ability, and tweak the dosage as needed.
They repeat these steps for a few weeks until they've found the optimum dosage. The patient then remains on the final, stable dosage (with regular tests to check that it's still the right fit).
Fortunately, doctors have been doing this for decades and have carefully worked out the technique. But Johnson and her colleagues think there's a better way through pharmacogenetics.
Johnson discussed this idea with other scientists from the pharmacogenetics research network and from the online knowledge base PharmGKB. They all knew that to fully investigate whether pharmacogenetics could improve warfarin dosing, they would need a worldwide effort.
So they created the International Warfarin Pharmacogenetics Consortium. The consortium is made up of about a hundred researchers on four continents.
The scientists already knew that variations in two genes, CYP2C9 and VKORC1 (an enzyme that activates vitamin K), could influence warfarin's effectiveness. But no one was really sure whether knowledge of a patient's CYP2C9 and VKORC1 variations could help doctors arrive at the optimal dose of warfarin more quickly. That's what the consortium set out to determine.
By combining their data, consortium members had access to anonymized information from about 5,700 patients on stable dosages of warfarin. The patients came from around the globe: Taiwan, Japan, Korea, Singapore, Sweden, Israel, Brazil, Britain and the United States.
This kind of study—one that includes people of different races, ethnicities and lifestyles—is essential to draw conclusions that are applicable to a wide range of people.
From this vast pool of data, the consortium members created a computer program to predict the ideal warfarin dosage for each patient based on his or her genetic variations and clinical information like age and body size.
Then the scientists checked their predictions against the actual dosage for each patient. (These stable dosages had been established the traditional way—they were initially based on standard clinical factors, then adjusted until they were optimal.)
Voila! The genetically based computational predictions were closer to the stable dosages than were the starting dosages obtained using the standard, best-guess method.
The computer program performed especially well for patients at the low or high ends of the dosing range. This got the scientists' attention, because nearly half of the people on warfarin are at the extremes of the range, and they are the ones most susceptible to dangerous bleeding or clotting.
The consortium published these discoveries last year in a major medical journal.
As the consortium's research con tinues, its strategy is being tested in a large clinical trial to determine whether a gene-based approach to prescribing warfarin will improve the effectiveness and safety of the drug for new patients. The trial is called Clarification of Optimal Anticoagulation through Genetics (COAG).
Lowering the Pressure
Most of Johnson's work focuses on drugs that treat high blood pressure, or hypertension.
In the case of warfarin, genes influence how much of the drug a patient needs. For drugs that treat hyper tension, genes influence which drug—there are dozens—would be best for each patient.
"If doctors randomly pick one of those medicines, there's only a 50 percent chance that it will work," Johnson says.
"We're trying to find out if there are genetic markers to use to pick the right drug from the outset," she continues. "Right now, it's a trial-and-error process. It can be very frustrating for patients, especially for young people."
Too often, people get fed up when doctors repeatedly change the medicines, she says. The patients feel fine—hypertension has no obvious symptoms—and they may not understand the importance of finding an effective medicine.
"So they go untreated for an extended period. And that's bad," Johnson says.
Even though they can't feel it, the extra force of blood smashing against artery walls can seriously damage internal organs. Long-term consequences include kidney failure, strokes, heart attacks, heart failure and death. Because of this, hyper tension is sometimes called the "silent killer."
The condition is hard to study because it is influenced by many factors, including genetics, diet and exercise.
"Hypertension isn't just one disease," Johnson says. "It's reflected in one way—high blood pressure—but it's really a half dozen pathophysiologies at least."
The underlying cause—or pathophysiology, as Johnson puts it—of each person's hypertension points to the best way to lower that person's blood pressure.
One type of drug for hypertension (beta blocker) works by blocking nerve signals that make the heart beat faster and harder. Another type (diuretic) increases urination and relaxes blood vessels. A third (calcium channel blocker) relaxes heart and blood vessel muscles by keeping out the calcium ions these muscles need in order to contract.
"There's probably a right medicine out there for every patient, but finding that is the trick," she says.
Johnson's research team has identified a number of gene variations, or genotypes, that influence how people respond to blood pressurelowering drugs. They have also found genotypes associated with dangerous, long-term complications from hypertension.
Bringing the research full circle, they discovered which drugs work best for people with these genotypes.
That's the beginning of success for Johnson, who hopes to help doctors move beyond their current approach toward what you might call a genotype-and-get-it-right approach.
"For me personally, the long-term goal is to guide the selection of medication for an individual patient," she says. "In that scenario, physicians would order a genetic test and based on it, they would narrow down the drug possibilities that would be best for that individual patient."
In other words, doctors would be able to look up a patient's genotype to find the right medicine and/or dose for that person—or at least find out what not to prescribe.
"We're not quite there yet," she says. "But we think we're getting close."
Maybe then, we'll be able to modify Paracelsus' words from "Only the dosage distinguishes the killer from the cure," to "Now genetics help us distinguish the killer from the cure."
Genes, Disease and Drugs
Have you ever taken a medicine that didn't work or that caused bad side effects? If so, you know that not everyone responds the same way to medications.
Recognizing the importance of understanding these differences, the National Institute of General Medical Sciences, together with other components of the National Institutes of Health, created the Pharmacogenetics Research Network (PGRN) in 2000.
Scientists in this nationwide collaboration study genes and medicines relevant to a wide range of diseases. The researchers expect that the knowledge they uncover will help doctors use genetic information to tailor treatments for each patient, in essence making drugs safer and more effective for everyone.
So far, PGRN scientists have learned important lessons about gene-based responses to drugs that treat asthma, breast cancer, childhood leukemia, depression, heart disease, high blood pressure and other conditions. Discoveries made by the network have already led to changes in the prescribing instructions for some medications.
In 2008, members of the PGRN joined forces with scientists in Japan to form the Global Alliance for Pharmacogenomics. The new research teams use state-of-the-art genomic technology to examine the sequences of thousands of genes simultaneously. This approach allows scientists to get a fix on the genes most likely to play an important role in how people respond to drugs.
You can find more information about the Pharmacogenetics Research Network at http://www.nigms.nih.gov/Research/SpecificAreas/PGRN. —A.Z.M.