Task Force on Gene Testing for Antidepressant Efficacy Concludes Tests Not Yet Ready for Widespread Use

One problem is lack of transparency regarding algorithms used by commercial entities to determine treatment recommendations on the basis of testing of genetic variations.

Individualized medicine employing pharmacogenomics to guide psychiatric treatment is the future, experts agree. But the future is not here yet.

That’s the conclusion of the APA Task Force for NovelBiomarkers and Treatments, a component of the APA Council on Research. In a report published in AJP in Advance this spring, the group described how there are insufficient data to support widespread use of pharmacogenomic tests in clinical practice to guide antidepressant treatment.

Charles Nemeroff, M.D, Ph.D., says that while genetic polymorphisms have been linked to antidepressant efficacy, many of them are typically not included in commercially available tests.

The task force was chaired by Charles Nemeroff, M.D., Ph.D., the Leonard M. Miller Professor and chair of the Department of Psychiatry and Behavioral Sciences at the University of Miami Miller School of Medicine. The task force members examined the research base supporting the claims of four major commercial pharmacogenomic tests (GeneSight, GeneCept, IDgenetix, and CNSDose). In general, these companies claim that their pharmacogenomic tests can improve the accuracy of drug choice and reduce trial and error.

The task force found the evidence supporting these claims to be problematic in at least two important ways: there are methodological flaws in supporting research that seriously undercut company claims; and there is a lack of transparency with regard to the algorithms used by the companies to derive treatment recommendations from tests looking at multiple genetic variations (known as combinatorial tests).

“We do not believe the evidence is sufficient at this time to justify the cost associated with pharmacogenomic testing, and the data simply do not reinforce or support commercial claims,” Nemeroff told Psychiatric News.

Most Tests Examine Pharmacokinetics

There are two broad domains affecting drug response that researchers believe could be influenced by an individual’s genetic profile: pharmacokinetics (what the body does to a medication, especially how quickly or slowly it is metabolized) and pharmacodynamics (what the medication does to the body’s cells, tissues, and organs).

The bulk of available commercial testing is around pharmacokinetics. Genetic variations affecting the expression of cytochrome P450 (CYP 450) enzymes are known to affect whether the body metabolizes a particular drug quickly or slowly.

For each of the four commercially offered tests, the task force assessed the research used to support proprietary claims. “In all cases, almost all of the studies are of short duration, have mostly small sample sizes, and are mostly unblinded,” Nemeroff told Psychiatric News. “The experimental and treatment-as-usual arms of the studies are typically not matched, and most of the study cohorts for most of the studies are very heterogenous, including patients with bipolar disorder as well as depression and anxiety disorders.”

He added, “A critical point that was very concerning to our group, and to the council, is that the formula by which the companies come to a conclusion that is supposed to guide the primary care clinician is proprietary, so it can’t be reviewed.”

While the makers of these tests claim the results can help guide treatment for depression, as well as other psychiatric disorders including bipolar disorder, attention-deficit/hyperactivity disorder, and posttraumatic stress disorder, Nemeroff noted that “the evidence for this is meager.”

How Well Do Tests Work?

The task force, in its paper, notes that GeneSight testing, produced by AssureRx Health, has the “most substantial evidence” of the four tests examined, with three clinical trials resulting in eight publications. GeneSight uses a drug-gene interpretative report that categorizes drugs into three “bins,” using color-coded descriptors: green, “use as directed”; yellow, “use with caution”; and red, “use with caution and with more frequent monitoring.” Once a patient’s DNA sample is received, the company can complete the analysis and send a report to the health care provider within 36 hours, according to GeneSight.

Such a quick turnaround for patients and doctors could potentially allow clinicians to quickly customize prescribing based on a patient’s phenotype. The task force noted that there are key limitations to the studies evaluating whether use of GeneSight improves treatment outcomes. As with the studies performed by the other companies of tests examined, many of the studies by GeneSight have been short in duration, included small samples of patients, and/or lacked blinding.

At APA’s 2018 Annual Meeting, a poster presented by GeneSight researchers reported the results of a large, multicenter trial that compared outcomes in patients with depression who received GeneSight testing and those who did not receive GeneSight testing. The data from this study were not available at the time the task force conducted its analysis.

The 24-week study included 1,167 patients with moderate-to-very-severe depression who had failed at least one antidepressant medication. Patients were divided into two treatment arms: the GeneSight Psychotropic test arm (n=560), in which clinicians used the GeneSight test results to guide treatment decisions; and a treatment-as-usual arm (n=607), in which psychiatrists and primary care physicians prescribed medication as they normally would without the benefit of genetic testing. Measured outcomes were response, remission, and symptom improvement on the Hamilton Depression scale (HAM-D17) at week 8.

According to a GeneSight press release, 26 percent of those in the GeneSight treatment arm achieved response to treatment compared with 19.9 percent in the treatment-as-usual group, a statistically significant response. Also, 15.3 percent of patients in the GeneSight treatment arm achieved remission compared with 10.1 percent in the treatment-as-usual group, which was also a significant difference, according to the researchers.

Nemeroff said the GeneSight study fell short of statistical significance on its primary outcome—symptom improvement on the HAM-D test. Moreover, he said the remission rates for both treatment arms (10 and 15 percent) were unusually low even for treatment-resistant depression.

Search Continues for Biomarkers to Guide More Precise Prescribing

The task force conclusion that there is insufficient evidence to support widespread use of pharmacogenetic decision support tools is shared by other experts. “The claims on company websites may be good marketing, but they are not balanced, and the time-pressured clinician or the uninformed consumer, often in distress, may be especially vulnerable to the pitch,” George Zubenko, M.D., Ph.D., Barbara Somer, M.D., and Bruce Cohen, M.D., Ph.D., wrote in an editorial in JAMA Psychiatry. “Medicine has a history of use of improperly evaluated treatments, and some persist because consumers demand help and can find clinicians who will comply.”

The task force notes in its report that serious efforts are under way to identify genes and other determinants that might guide more precise prescribing for patients with depression. These include the International Study to Predict Optimized Treatment in Depression (iSPOT-D), Establishing Moderators and Biosignatures of Antidepressant Response for Clinical Care for Depression (EMBARC), and the Predicting Remission in Depression to Individual and Combined Treatment (PREDICT) studies. The goal of such efforts is, in part, to identify biological substrates of major depression and their roles in response to antidepressant medications.

“The notion of personalized medicine in psychiatry is an important one, and I have no doubt that in the future lab tests will be helpful in identifying who is at risk for major psychiatric disorders and individualizing treatment for a given patient in your office,” Nemeroff said. “I do believe that in five years we may have a well-validated pharmacogenetic laboratory test to be used in antidepressant prescribing. But the bottom line is that we don’t feel we are there yet.” ■