Skip to Content

Still Waiting for Personalized Medicine

Pharmacogenomics promises to let doctors choose drugs and dosages based on tests of your genetic profile. But just try taking a test.
November 1, 2006

Nausea, fatigue, dizziness, dry mouth, insomnia. For people like me, who seem susceptible to every side effect possible, the tiny type on ads for new drugs is required reading. NyQuil puts me into a half-conscious delirium. Codeine makes me throw up. And back in college, when my doctor prescribed Wellbutrin to help me quit smoking, I experienced blurred vision and the worst headaches of my life.

Given that my troubled history with medication is shared by my mother and sister, I have long suspected a genetic basis for my sensitivity. So like many others, I have over the last few years eagerly anticipated the benefits of pharmacogenomics–a field whose researchers aim to let doctors tailor prescriptions to their patients’ genetic makeups. It’s one of the most tantalizing promises of the genomic era: quick and easy tests that tell you which drugs to take or what dose is right for you.

A few tests have been developed for specific diseases, such as cancer–most notably a genetic test that predicts which lung cancer patients will respond to some medications. But a new product, marketed by the Swiss pharmaceutical giant Roche and approved by the U.S. Food and Drug Administration in January 2005, now has the potential to begin making pharmacogenomics broadly accessible. Called the AmpliChip CYP450 assay, it uses genetic analyses to ascertain how quickly people metabolize certain drugs, thus predicting who is most likely to experience unpleasant or even toxic side effects.

When two people take the same dose of a drug, their bodies may metabolize it so differently that the amount of it that can act on its target varies tremendously. Some people may have an especially efficient form of an enzyme that breaks down a drug; others may have a less functional version. The AmpliChip test works by detecting specific variations in genes that code for two important drug-metabolizing enzymes, CYP2D6 and CYP2C19. These enzymes help break down 25 percent of all drugs, including the most commonly prescribed drugs in the United States, such as antidepressants, blood pressure medicines, cough medicines, and painkillers.

People with genetic variations that give them less efficient versions of the enzymes, known as poor metabolizers, could have high levels of a drug in their body for a longer period, increasing the potential for side effects. People with functioning copies of the genes are called extensive metabolizers, while people with extra copies of the functioning CYP2D6 gene are labeled ultrarapid metabolizers. “This kind of information is something every doctor seeing patients should know about,” says Julio Licinio, a psychiatrist at the University of Miami’s Miller School of Medicine and the editor of The Pharmacogenomics Journal.

CYP2D6 variations occur relatively frequently in the gene pool, though the incidence differs by population group. About 5 to 10 percent of Caucasians have a genetic profile that makes them poor metabolizers of drugs broken down by CYP2D6, whereas among Asians, the proportion is more like 15 to 20 percent. And almost 30 percent of people from North Africa and the Middle East are ultrarapid metabolizers, meaning they may need higher than standard doses of various drugs.

In addition to helping people avoid side effects, the AmpliChip might save lives. Some older antidepressants, for example, can cause cardiovascular problems if ingested in high enough doses. And while most drugs are administered in their therapeutically active form, some drugs, such as the breast cancer drug tamoxifen, must be metabolized to be effective, so they may fail to work in poor metabolizers.

“Adverse [drug] events cost the ­public-­health system one to two billion dollars per year,” says Lawrence Lesko, director of the Office of Clinical Pharmacology at the FDA. “One of the [results] is going to be prospective use of devices like AmpliChip.” He adds that in the future, doctors or pharmaceutical companies could be held liable if a patient is not given a genetic test and experiences serious side effects. “I think we’re going to see a huge consumer interest in pharmacogenomics,” says Lesko. “And I think we’re going to see more tests.”

Comical Confusion

Maybe. But it’s clear that the consumer boom has not yet begun. For routine blood tests such as those for high cholesterol or anemia, you stroll to a lab down the hall from the doctor’s office, surrender some blood, and get a call with the results a few days later. But as I quickly learned, the sheer novelty of the AmpliChip test makes matters far more difficult. For one thing, the test is enormously expensive–I paid $1,360, though prices vary depending on the lab–and is not covered by most insurance companies. And many doctors are unfamiliar with the test and may be reluctant to order it. When I asked my doctor to have me tested, he quickly dismissed the idea. He had never heard of the test; besides, he told me, he could figure out my optimal dose of various drugs the old-fashioned way, by trial and error.

When I finally got the opportunity to take the test (thanks to a new doctor and a subsidy from Technology Review), I encountered an almost comical level of confusion. Though my new, younger physician was more open to the idea, I was the first of her patients to ask about the AmpliChip, and she had no idea how to order it. I found prescribing directions on the Roche website, and a list of the few labs in the country that actually offer the test. When I called the closest one of them, a 20-minute drive from Boston, the person on the other end of the phone said she didn’t know what I was talking about and hung up.

Calls to both Roche and the lab’s headquarters resolved the confusion, and a few days later I arrived at a medical strip mall with a prescription from my doctor and specific instructions for the phlebotomist–“Tell her to draw seven milliliters of blood into a lavender-top tube and store it at room temperature”–lest the confusion persist. When I asked the phlebotomist how long I would have to wait in order to get the results, she replied that she had no idea. I was her first patient to get the test.

A week later, I drove to my doctor’s office. She showed me a single-page document with a chart listing some of the drugs metabolized by the relevant enzymes, as well as definitions of poor and ultrarapid metabolizers. Near the top of the page was a small box containing the sum total of my results: “CYP2D6, extensive metabolizer, CYP2C19, extensive metabolizer.” Much to my surprise, I am totally normal. My DNA sequences encoding both enzymes contain none of the known variants that would render them less effective in metabolizing drugs like codeine and the ingredients in NyQuil.

Did this mean I had imagined the side effects of various drugs? Were the nausea and headaches really a kind of negative placebo effect? After consulting with several experts, I still don’t know the answer. Walter Koch, head of research at Roche Molecular Diagnostics, explains that a complex network of factors can influence an individual’s response to drugs, including “age, gender, diet, hormone levels, concurrent medications, and inherited variations [in genes other than those tested for by the AmpliChip].”

It’s also possible that I possess a rare mutation in CYP2D6 or CYP2C19, one the AmpliChip test doesn’t look for. Although the AmpliChip detects the majority of known clinically relevant mutations in these genes, new variants of the genes are still being discovered, according to Miami’s Licinio. Ultimately, my conundrum points out a limitation of diagnostic testing. “These tests are just one of the pieces of information that should be part of a patient’s history, along with your age, your parents’ medical history, and other factors,” says Lesko.

Who Pays?

But at more than a thousand dollars a test, the AmpliChip is not just another easily gained piece of information. For most patients, genetic testing remains expensive and exotic, both economically and logistically inaccessible. So what will it take for pharmacogenomics information to become a standard part of our medical charts? Experts say educating physicians is one of the biggest obstacles. “When most doctors were in medical school, pharmacogenomics was not part of their training,” says Licinio.

Already, though, more and more physicians do want to use these tests. David Mrazek, chair of the psychiatry and psychology department at the Mayo Clinic in Rochester, MN, uses them routinely in clinical practice. He says the benefit of pharmacogenomics in psychiatry is clear: people vary enormously in their response to antidepressants and antipsychotics, both of which can cause troublesome side effects. Some patients spend weeks or months or even years trying different doses of various drugs to find the one that brings the most relief with the fewest problems. By testing for genetic variants in drug-metabolizing enzymes, Mrazek is able to save his patients much of that trial and error. “If a patient is a poor metabolizer of Prozac, I’ll start them on a different drug,” he says.

The real hurdle, then, will be financial. Insurance companies still consider the AmpliChip experimental and are unlikely to cover it until large clinical trials prove that it can both help patients and cut costs. Such studies, which are already under way for psychiatric disorders, will also help determine how best to use the test.

Whatever the economic and insurance considerations, however, the advent of genetic tests like AmpliChip seems all but inevitable. And for patients, that is a good thing.

Emily Singer is the biotechnology and life sciences editor of Technology Review.

Keep Reading

Most Popular

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

It’s time to retire the term “user”

The proliferation of AI means we need a new word.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Sam Altman says helpful agents are poised to become AI’s killer function

Open AI’s CEO says we won’t need new hardware or lots more training data to get there.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.