Skip to content

Improvement Needed in Genetic Test–Guided Treatment of Major Depressive Disorder

Key findings

  • This multicenter, randomized, controlled, double-blind trial compared antidepressant therapy guided by a pharmacogenomic test versus treatment-as-usual in 304 outpatients with moderate to severe major depressive disorder
  • No significant difference in depression severity over eight weeks was identified between the two arms
  • Similarly, secondary measures of depression did not distinguish the intervention arm from the control arm
  • An exploratory analysis determined that significantly fewer participants worsened in the assay-guided group than in the treatment-as-usual group
  • Another exploratory analysis indicated that individuals prescribed therapy concordant with pharmacogenetic test recommendations are more likely to remit than those whose therapy was somewhat concordant or nonconcordant

Functional variations in a small number of pharmacokinetic genes influence the levels of medications in blood, including more than 80% of medications commonly used in psychiatry. Genetic tests are now commercially available to help psychiatrists individualize drug selection for each patient, but single-blind, randomized, controlled trials of these tests in major depressive disorder (MDD) have had mixed results.

To better characterize the potential benefit of a pharmacogenomic assay for patients with MDD, Roy Perlis, MD, MSc, director of the Center for Quantitative Health at Massachusetts General Hospital, and Maurizio Fava, MD, psychiatrist-in-chief and director of the Division of Clinical Research in the Mass General Research Institute, and colleagues performed a multicenter, randomized, controlled, double-blind (subject and rater) trial. According to their report in Depression and Anxiety, depression severity with assay-guided treatment was no better than with treatment-as-usual.

Study Methods

Investigators at 21 centers were trained on the interpretation of genetic testing results and the relevance of each genetic variant to pharmacotherapy. The Genecept Assay (version 2.0) incorporates 45 variants of seven pharmacokinetic cytochrome P450 genes and 12 variants of 11 pharmacodynamic or other genes.

The 304 participants were outpatients, ages 18 to 75, who had a primary diagnosis of nonpsychotic MDD and had either not responded to or tolerated one prior trial of a standard antidepressant for their current depressive episode. They were also required to score >18 on the Structured Interview Guide for the Hamilton Depression Rating Scale 17-item version (SIGH-D) at both the screening visit and a baseline visit one week later.

All participants provided a DNA sample at the screening visit. They were randomly assigned 1:1 to:

  • Assay-guided treatment (AGT)—Assay results were provided to the treating investigator at baseline; the investigator could use the results to select antidepressant therapy but was not required to do so
  • Treatment-as-usual (TAU)—Investigators treated patients without knowing the genetic test results

Primary and Secondary Analyses

For the primary outcome, change from baseline in SIGH-D, there was no significant difference detected between AGT and TAU at week 8.

Likewise, there was no difference between AGT and TAU in:

  • Response (≥50% reduction from baseline in SIGH-D)
  • Remission (SIGH-D ≤7)
  • Quick Inventory of Depressive Symptomatology (Self-Report) score
  • Clinical Global Impression–Severity score

Post Hoc Analyses

There was no evidence that AGT led to poorer outcomes than TAU. In fact, the exploratory analysis indicated that significantly fewer participants worsened in the AGT group.

In another post hoc analysis, a third-party reviewer without knowledge of outcomes categorized the prescribed treatment as clearly concordant with the assay's guidance or somewhat/mixed, nonconcordant or unable to categorize. In both study arms combined, patients treated in clear concordance with assay results were more likely to remit (34%) than patients whose treatment was discordant or of somewhat/mixed concordance (19%; P = 0.005).

Interpreting the Findings

An irony of precision medicine is the frequent difficulty in identifying the precise population in which an intervention will have the greatest benefit. The positive findings in this study are preliminary and post hoc, but they suggest the future possibility of identifying high-risk individuals with MDD who would benefit most from genetic testing.

Explore research in the Department of Psychiatry

View the Department of Psychiatry's Guide to Mental Health Resources


In nondepressed young adults who had a family history of depression, higher amygdala activity on functional MRI was significantly correlated with lower resilience.


Calls, texts and movement patterns captured by individuals' smartphones over the course of two weeks provided sufficient information to predict severity of SAD symptoms with moderately strong accuracy.