Winters had always been rough for Diane Smith.

Since her mom died a decade ago, they’ve been all but unbearable. Her depression would get progressively worse until the anniversary of her mother’s death: Valentine’s Day.

This year — its heavy snowfall forcing her to shovel constantly — was shaping up to be particularly bad. But, somehow, the despair never came.

“I would have been hospitalized at that point because I would have been so despondent,” said the 61-year-old Bend resident, “and I just got through it miraculously.”

Last spring, her doctor performed a test that turned up a genetic mutation he believes was making her antidepressant less effective. He put her on a supplement designed to augment the medication.

A plethora of tests exist that identify, usually through a mouth swab, whether patients’ unique genetic makeups affect how well specific medications work for them and whether they could produce dangerous side effects. This emerging sector of medicine is called pharmacogenetics, or pharmacogenomics. Although it’s been slow to catch on among medical providers, research on the subject is robust.

More than 100 Food and Drug Administration- approved medications list relevant genetic factors that could affect their safety or effectiveness. The FDA lists on its website more than 200 genetic biomarkers and drug pairings and their potential consequences for patients.

Dr. Richard MacDonell, Smith’s doctor and a primary care physician in Bend who offers pharmacogenetic testing, said traditional medicine works by trial-and-error: You take patients through several different types of medications for their condition, tweak the dose here and there, until you find one that works without terrible side effects. The process can take months. A look at their genetic makeup, however, can narrow things down much faster.

“The question is, ‘Can we be smarter about it and tailor it instead of just shotgunning?’” MacDonell said.

Varying evidence levels

Not every gene and drug combo has the same amount of research supporting its link. Some pharmacogenetic test kits offer to test for genes whose effects on medications aren’t well established in medical literature. That’s not to say there isn’t potential benefit, however.

Smith’s genetic biomarker and medication, for example, are not on the FDA’s list. The idea is that some people with depression also have a mutation in their methylenetetrahydrofolate reductase gene, which helps the body produce folate. Without enough folate, antidepressants don’t work properly. MacDonell puts patients on a prescription folate supplement called Deplin. The FDA approved Deplin for boosting the effectiveness of antidepressants, but it does not consider the link between MTHFR and antidepressants conclusive. Some studies support the link.

Nonetheless, MacDonell has put several patients with depression on the supplement Smith takes. He can’t prove it’s working, but said their symptoms have improved.

“I claim that the Deplin helped,” he said. “Can you prove it? No.”

Dr. Deepak Voora, an assistant professor of medicine with Duke University and a scientist in its Center for Applied Genomics & Precision Medicine, said it’s one thing to make an association between a drug and a genetic mutation, and another to show that having that information actually improves outcomes for the patient. Many combos have the former, but not the latter.

“That’s something that really limits their uptake because physicians want to see that, ‘Hey, if I use this fancy new test, is it going to improve how I manage this patient in some way, beyond a theoretical argument?’”

Risk rating

Voora calls the HIV drug Abacavir the “poster child” for pharmacogenetics because its link to a specific biomarker is among the most well understood. Patients with HIV are frequently considered for the medication, but if their genetic tests reveal they have the biomarker, called HLA-B, their doctor will choose a different drug to avoid the risk of serious side effects.

“Abacavir is kind of, I’d say, the shining example of pharmacogenetics because it’s got really good data behind it,” he said. “We’ve got a variety of other drug-gene pairs where the data is less robust than Abacavir.”

MacDonell uses a test from a San Diego-based company called Millennium Health, which works by sending in a cotton swab that’s swiped inside the patient’s mouth. Doctors order tests for the specific genetic biomarkers they’re interested in. Each biomarker lists the drug or drugs it may affect.

Millennium produces a read-out that divides the medications in question into three categories based on their genetic impact: low, moderate or high. (On other tests, this might be green, yellow or red.) Low genetic impact means the drug is unlikely to cause any adverse reactions, whereas high means the patient is at high risk of having an adverse reaction to the drug. In the latter case, the doctor either takes the patient off of the drug, tries a lower dose or monitors closely for side effects.

“It absolutely does not mean to pull someone immediately off of that medication if it’s in high risk; it simply means to put a flag around it and to really watch it and monitor to see if things change in terms of that patient’s condition,” said Kelly Olson, Millennium’s associate director of clinical affairs.

The tests have become cheaper in recent years — from thousands of dollars down to hundreds — and a single test is able to cover much more ground these days than in the past, when many could only test for one genetic biomarker.

For someone without insurance coverage, Millenium’s test costs $249, Olson said. Several insurers refuse to cover Millennium’s and other pharmacogenetic tests. Moda Health, for example, considers several of the tests, including Millennium’s “experimental and investigational” and says there is not enough scientific evidence to support their use. Neither PacificSource Health Plans nor Regence BlueCross BlueShield of Oregon cover Millennium’s test.

Learning curve for docs

MacDonell learned about the tests when a fellow physician, who has since retired, told him they had worked wonders on his patients with depression.

“I was naturally pretty skeptical,” he said.

For many doctors, MacDonell included, the tests can be intimidating to dive into because they’re not geneticists. It can be a steep learning curve. Many test manufacturers, Millennium included, offer training for providers who use them.

Dr. Ann Moyer, co-director of Mayo Clinic’s Personalized Genomics Laboratory, said the practice is slowly catching on, but it does take a while to fully understand.

“I think in general people are pretty interested in it, but I think it is difficult to just start incorporating because of its newness for them,” she said.

For now, MacDonell uses the tests narrowly — only for patients such as Smith, whose antidepressants aren’t always effective.

Smith, who has struggled with depression since her 20s, said she thinks the improvement in her mental health is no fluke. Her husband even pointed it out to her. She’s considering spreading the word to family members who might have the same genetic mutation.

She added, “I’d have to credit Deplin, without having any further explanation of why I would be doing so much better this year than in previous ones.”

— Reporter: 541-383-0304,

tbannow@bendbulletin.com

16418926