Can genetic testing help doctors better prescribe antidepressants? There’s
quite a debate.

(Photo illustration by Joe Raedle/Getty Images)

By Ilana Marcus <>
March 31

Grit alone got Linda Greene through her husband’s muscular dystrophy, her
daughter’s traumatic brain injury, and her own mysterious illness that
lasted for three years and left her vomiting daily before doctors
identified the cause. But eventually, after too many days sitting at her
desk at work crying, she went to see her doctor for help.

He prescribed an antidepressant and referred her to a psychiatrist. When
the first medication didn’t help, the psychiatrist tried another — and
another and another — hoping to find one that made her feel better.
Instead, Greene felt like a zombie and sometimes she hallucinated and
couldn’t sleep. In the worst moment, she found herself contemplating

“It was horrible,” she said. She never had suicidal thoughts before and was
terrified. She went back to her primary care doctor.

In the past, when Jeremy Bruce, Greene’s physician in Cincinnati, treated
patients for depression, he followed the same steps for almost everyone:
start the patient on one antidepressant and switch to another until
something helped. Sometimes, before they found the right treatment, the
patient would leave his practice to find a new doctor.

“They would usually be very angry,” Bruce said.

But about three years ago, Bruce tried a new approach.

Linda Greene and her husband. She tried many antidepressants before her
doctor suggested genetic testing to find a medicine that worked for her.
Doctors increasingly use information about genes to evaluate potential risk
for some diseases and to determine the best drug treatment. But using
pharmacogenetics to help treat depression remains controversial. (Family

For patients who weren’t responding well after trying one or two different
antidepressants, he started sending samples of their DNA to a company that
says it can use an individual’s genetics to match them with the
antidepressants most likely to work for them. Bruce said the test’s
recommendations seemed to help some of his patients, so now he offers the
test to any patient with depression — before they even try the first

“Psychiatric medicines make people feel horrible if you choose the wrong
one,” Bruce said. “And they feel great if you choose the right one.”

In Greene’s case, the genetic report she got back put all the medications
she had tried on a list of drugs unlikely to work for her, so she stopped
taking them. She went to a different psychiatrist, who used the test
results to prescribe something deemed better for her genetic makeup and
says the meds seem to be working.

More than a third of American adults take prescribe drugs linked to

Doctors increasingly use information about genes to evaluate potential risk
for some diseases, such as BRCA genes that are linked to breast cancer, and
to determine the best drug treatment for diseases, including acute leukemia
and HIV. But using pharmacogenetics to help treat depression remains

Doctors such as Bruce say they have seen promising patient results, but
others say there is not enough solid evidence to show that pharmacogenetics
can work for the complexities of mental health treatment. Some lab tests
have shown relationships between genes and the way a drug physically
affects the body, but studies on whether using that information leads to
better results for patients have been inconclusive.

Bruce Cohen, director of the Program for Neuropsychiatric Research at
McLean Hospital, a Massachusetts psychiatric treatment and research center
affiliated with Harvard Medical School, says studies so far haven’t shown
that genetic testing to choose antidepressants leads to collectively better
outcomes for patients.

“The differences are very small, and there’s no reason to think you
couldn’t have done better by just following standard protocols, which are
free,” he said. According to the National Institutes of Health, genetic
testing can cost anywhere from $100 to more than $2,000.

Cohen says most of the genetic tests check for variations in genes related
to metabolism, affecting how a person’s body processes a drug. Faster
metabolism can lead to lower drug levels in the body, while slower
metabolism can cause higher levels. But such factors as age, diet and other
substances in the body also have a major influence on the processing of the

“Metabolism is only a part of drug response, and even it is more determined
by nongenetic factors,” Cohen said. “I’m not saying that drug metabolism
doesn’t matter. It’s a question of degrees, it’s a question of how much of
somebody’s response is determined by metabolism as opposed to other aspects
of what happens to drugs when you take them.”

The Clinical Pharmacogenetics Implementation Consortium
<>, an NIH-funded international organization, ranks the
strength of various gene-drug pairings based on reviews of published
research and provides prescribing guidelines. The guidelines advise on how
to use genetic information that is already available, rather than which
circumstances call for genetic tests to be ordered.

Consortium co-founder Mary V. Relling, who holds a doctor of pharmacy
degree and chairs the pharmaceutical sciences department at St. Jude
Children’s Research Hospital, agreed that the elements Cohen listed affect
drug metabolism, but for some gene-drug combinations, she said the genes
dominate all other factors.

“Over <> and over
<> and over
<> again, studies show that
patients who have [particular] genetic defects have more toxicity, for
example, than patients who don’t,” which could cause adverse reactions to a
drug, said Relling, who studies the clinical application of
pharmacogenetics and investigates approaches to improving drug therapy for
pediatric leukemia patients.

One aspect of the disagreement is how much clout to give different types of
studies. Relling said studies that show a strong relationship between a
gene and a drug should be enough to inform prescribing guidelines. Cohen
argues that simply observing a relationship on the genetic level is not
sufficient and that studies need to prove patients actually have better
outcomes when treatment is guided by genetics.

Relling said that while her consortium will only issue prescribing
guidelines based on strong evidence, some commercially available genetic
panels for psychiatry include genes with less established interactions.

The Food and Drug Administration in November released a statement
warning patients and doctors to be cautious about genetic testing that has
not been approved by the FDA, including those that guide the prescribing of

“The FDA is aware of genetic tests that claim results can be used by
physicians to identify which antidepressant medication would have increased
effectiveness or side effects compared to other antidepressant
medications,” said the statement by the directors of the Center for Devices
and Radiological Health and of the Center for Drug Evaluation and Research.
“However, the relationship between DNA variations and the effectiveness of
antidepressant medications has never been established. Moreover, the FDA is
aware that health care providers have made changes to patients’ medication
based on genetic test results that claim to provide information on the
personalized dosage or treatment regimens for some antidepressant
medications, which could potentially lead to patient harm.”

Greene said she wasn’t concerned the test she used to find an
antidepressant wasn’t FDA approved. She said she was desperate for help
that had eluded her and she trusted her doctor’s clinical experience. She
said the test cost her around $400.

Anthony Rothschild, a professor of psychiatry at the University of
Massachusetts Medical School and co-author of an industry-funded study
<> that
found inconclusive patient outcomes when genetic testing was used to
prescribe antidepressants, said such testing should just be considered one
tool in the depression-treating arsenal, especially in cases where patients
have not responded to medication.

“I would say to the critics, ‘Do you have any other suggestions right now
when you have a person in your office and things don’t seem to be working?’
” he said. “I think that it has a place at the present time. Maybe
something else will come along that’s better, but this is an important

Amanda Jostworth, 38, another patient of Bruce’s, started taking an
antidepressant last July, when her husband was recovering from an illness
and she struggled with the stress of his condition, her full-time job and
caring for their four kids. She tried one medication for about five months
and found all of her motivation sapped. She trains for marathons and
couldn’t bring herself to go out for a walk, let alone a run.

This year, Bruce recommended that she take the genetic test. Based on the
DNA report, he switched her to a different antidepressant. About a month
later, she said, she was already feeling better. “I’ve only been on it for
six weeks, but I feel amazing,” she said.

Others haven’t had such good results.

Following a suicide attempt at 17, Adam James, now 29, was on a variety of
different psychiatric medications through his early 20s. All of them left
him feeling terrible, he said. A therapist recommended that he take a
genetic test. When he got the results back, he saw that he had already
tried a number of the antidepressants and antipsychotics the test report
associated with low to moderate chances of poor genetic interaction.

The test also found he had genes that can lead to decreased folate levels,
but his primary care doctor said his levels were fine. “I took that as more
evidence against the test,” he said. Now, he has come off psychiatric drugs
completely. “I’ve been there, I don’t want to go down that road again,” he
said. He has been trying a ketogenic diet that he believes may be helping

Nora Whelan, 33, had her primary care doctor order the test after her
psychologist suggested it, in the hope of finding a medication to treat the
depressive symptoms of premenstrual dysphoric disorder. She wanted to avoid
the often-lengthy process of trying multiple medications before finding the
right fit.

When she and her doctor got the results, they noticed an antidepressant
that hadn’t worked well for her back in college was supposedly a good match
for her genetics. She tried a different recommended medication, but after a
few weeks, her symptoms were worse than ever. Now, she’s off that
antidepressant and relying on a medication she had already been taking for
anxiety, vitamin supplements, yoga and diet changes to address her symptoms.

She said the test’s guidance could possibly be helpful for other patients,
even though it didn’t work for her. It’s hard to know which medication will
be effective since everyone reacts differently.

Ultimately, she said, “they’re all basically a shot in the dark.”