Follow the Gene, or the Guidelines?

Genesight reports guide treatment by separating meds into red (avoid), yellow (caution), and green

A simple, low-cost approach wins the race

STUDY: Nierenberg AA et al, Journal of Clinical Psychopharmacology 2026

STUDY TYPE: Randomized, unblinded, comparative-effectiveness trial

FUNDING: Patient-Centered Outcomes Research Institute; no pharmaceutical industry funding

Background

Two simple steps make a big difference for patient’s outcomes:

1. Rating their symptoms at every visit (and graphing it)
2. Following treatment guidelines

In one randomized trial, that process raised remission rates from depression 2.5-fold. Genetic testing, in contrast, has yet to make that kind of difference. This study looked at whether genetic testing could add to guideline-informed care.

The Study

• 201 adults with major depression at 5 academic medical centers.
• Randomized to pharmacogenomic testing plus guideline-informed care, or guideline-informed care alone (CANMAT guidelines).
• Followed 12 months; primary outcome was well-being, with depression severity and function as secondary measures.

Results

There was no difference on the primary outcome (wellbeing) or on secondary outcomes: depression severity, physical function, social participation, and clinical adjustments (all p > 0.35). Analyzing the data by severity, prior treatment failures, or comorbidity also showed no benefit.

Clinicians who used genetic testing often avoided medications with moderate gene-drug interactions, and this practice lead to no gains in well-being.

Limitations

Samples were limited to academic centers. COVID-19 cut the target sample roughly in half. Neither patients nor clinicians were blinded (although that unblinding usually favors genetic testing).

Practice Implications

Genetic tests look at the same enzymes that drug interactions address. Most clinicians get dozens of drug interaction alerts a day — enough to inspire the term alert fatigue — but they know only a few of them are actionable, like

• Potent 3A4 inhibitors with clozapine
• Potent 1A2 inhibitors with clomipramine
• Potent 2D6 inhibitors with atomoxetine

Those are also among the actionable gene-drug interactions endorsed by the FDA and the Clinical Pharmacogenomics Resource.

The problem is that actionable interactions are rare, and most are flagged by the FDA because of very rare complications that won’t show up in these clinical trials (like QTc prolongation). Here are some other reasons why genetic testing falls short:

1. Antidepressants have wide therapeutic windows and most are metabolized through multiple pathways, limiting the relevance of a single metabolic gene.
2. CYP genes give only a blurry view of the serum level, and we don’t check serum levels of antidepressants anyway. Not because it is expensive, but because it isn’t useful (except tricyclics).
3. Why is the view blurry? Around 30-40% of a metabolic enzyme’s activity is determined by environment, not genes.

I might order a test for a patient who has unusual reactions that make me suspect abnormal genes, like if they get worse on any drug metabolized by CYP2D6. If it’s abnormal, that’s information they can use for life, even if it doesn’t change antidepressant prescribing now. I won’t avoid medications just because there’s an abnormal metabolic gene, but I might check serum levels in those cases.

But I’d take practice guidelines over genetic testing any day. Distilling hundreds of controlled trials tells us more than a blurry test.

—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report

What’s Your Take? Share in Comments