There's A Gene Test That Tells You Which Antidepressant Will Work Best

One woman’s unusual experience highlights the potential of precision medicine.

by Emily Glover
Nov 18 2016, 5:00pm

Michayla Sullivan doesn't remember pleading with her mom to chop off her hair, convinced it was on fire. It's just one of several incidents that she no longer recalls. In its place are hazy recollections of conversations about group homes and medications. The pills always seemed to do more harm than good, she thought, but she trusted her doctors knew what was best for her.

Not long before then, Sullivan had been a law student at the University of Notre Dame. Like many of her sleep-deprived, overworked peers, she often felt lethargic and generally down. She went to a clinic and—without much more than a brief consultation—walked out with a prescription for an anxiolytic, an anti-anxiety medication. If she didn't feel better, she was told, she could come back and try something something else.

This is a familiar process for the estimated 43 million American adults who are currently being treated for some form of mental illness: Just take whatever the doctor gives you. Wait a few weeks. Don't feel better? Get another script. "Generally you wait at least six weeks and also try to get to the maximum dose if tolerated by the patient before switching or adding another medication," says Celia Trotta, a board-certified psychologist in New Jersey.   

For most people, this process of trial and error works. But for an estimated 10 to 30 percent of patients being treated for depression, the outcome is less predictable—and the consequences can be dire. Side effects of common medications like Ativan, Xanax and Busparcan include extreme exhaustion, hallucinations, and confusion.

For Sullivan, that initial medication was only the first in a line of drugs that weren't just ineffective—they were actually harmful. Within one week of picking up the prescription, her anxiety progressed to the point of active paranoia. Another week later, Sullivan's parents removed her from school and took her to the hospital on the advice of another doctor. "This just kind of began two years of them not really understanding what was wrong with me," she says. 

The following months were filled with a stream of new doctors and medications—no one was able to provide a proper diagnosis. "Some of the symptoms fit, but none of them were ever quite right. So they kept trying new medications, which caused new varieties of what I now know were side-effects," she says. She experienced hallucinations, and became unable to care for herself. "I looked like I had mental illnesses that I didn't actually have, so they would try to treat them."

Even now, those two years are a blur. The only evidence she went on entire trips with her family are the pictures they have. She may introduce herself to someone only to learn they've met before. The days blended together as she slept for long stretches. She considers these gaps in her memory to be something of a good thing—what was going on wasn't something anyone would want to remember.

"My parents and I were starting to look at group homes for long-term care," she says. "They were changing their will and talking to my sister about providing for me because it looked like I was going to be unable to care for myself the rest of my life."

That was until Sullivan's doctor recommended a pharmacogenetic test called GeneSight, which involves taking a swab of the patient's mouth to look for genes that break down active ingredients in common medications. The doctor wasn't overly optimistic that it would solve Sullivan's problems, considering none of his previous patients were shown to have any genetic abnormalities. But Sullivan wasn't like most others, as the results proved, according to Sullivan: "It turns out I can't metabolize any antidepressants, any anxiolytics, at al," she says. "And what was diagnosed as depression in the first place was probably hypothyroidism exacerbated by law school."

With those results in hand, Sullivan's doctor began weaning her from the medications, which Sullivan says took several months because she had been "essentially overdosing on them" for the better part of two years. But even the worst of the withdrawal symptoms were worth it to her because, as she said, she is now a "normal, functioning member of society," which was something that seemed out of the question just a few years ago.

Prescribing medication is far from an exact science, as Sullivan's experience illustrates. "If a medication does not seem to be working, options include increasing the dosage, switching medications, or adding a second medication," Trotta explains. The makers of pharmacological tests like GeneSight, as well as other, similar companies like PGXL Laboratories and YouScript, hope to help doctors find the precise level of medication that works for each patient's unique physiological makeup. 

Research is still ongoing into the accuracy and effectiveness of these tests. In the meantime, physicians seem to be gradually buying into their promise. A recent study in Psychiatry Research found that 80 percent of doctors who used a pharmacogenetic test believed it would become "common standard in psychiatric drug treatment." They've also received federal endorsement: In 2015, President Obama launched the Precision Medicine Initiative, noting that genetic tests "can help identify the best drug for each patient and speed effective treatment." 

The potential upside of pharmacogenetic tests may be strongest for patients diagnosed with mental illnesses. In one study of 165 patients—which was funded by Assurex—patients who had genetically guided medications saw an average 70 percent improvement in depressive symptoms within eight weeks. Another independently funded study published in The Primary Care Companion for CNS Disorders in 2015 analyzed data from 685 patients with mood or anxiety disorders. The researchers found 87 percent of the patients "showed clinically measurable improvement" three months after their physicians used commercially available genetic tests, though the researchers note they didn't have a treatment-as-usual control group.

While promising, it may be too early to tell if these tests are as effective as they seem, says James Evans, professor of genetics and medicine at the University of North Carolina. "There have been small studies that indicate that some of these tests may guide therapy. My general, feeling though, is that it's really not ready for primetime," he explains. "I want to make sure that when I do a test on a patient that there's good science behind that show it to be effective. Otherwise we're practicing medicine through anecdotes."

For someone who is one of those anecdotes, it's a different matter. Says Sullivan, "I don't think I'm being melodramatic when I say that test saved my life." 

Image: Stephanie Santillan

mental illness
Gene testing