Genetic testing rarely makes a difference in antidepressant selection. Here are the exceptions.
STUDY: Baune BT, Pharmacogenomics 2026
STUDY TYPE: Narrative review
FUNDING: Independent
Background
Nearly two-thirds of patients with major depressive disorder don’t respond to their first antidepressant. Pharmacogenomic testing promises to fix that by matching patients to the right drug using their DNA. The idea is sound, but the commercial products have outrun the evidence.
The Study
- Critical narrative review of pharmacokinetic evidence, randomized controlled trials, and meta-analyses of pharmacogenomic-guided antidepressant prescribing.
- Focused on adult major depressive disorder, with selected adolescent data.
- Priority given to prospective trials and guideline-supported drug-gene pairs.
Results
The strongest evidence targets how the body metabolizes antidepressants, not how it responds to them.
Among the metabolic genes, CYP2D6 and CYP2C19 have guideline-backed evidence for prescribing decisions involving citalopram, escitalopram, sertraline, paroxetine, fluvoxamine, venlafaxine, vortioxetine, and tricyclic antidepressants. A third gene, CYP2B6, is relevant for sertraline specifically.
Two genes commonly included in commercial panels — SLC6A4 and HTR2A — don’t support prescribing recommendations. The associations have never produced a reliable signal in meta-analyses.
Meta-analyses show small-to-moderate benefits for pharmacogenomically guided prescribing overall, though the gains often disappear with adequate blinding. Patients with prior treatment failures, dose-limiting side effects, or difficult-to-treat depression are more likely to benefit from the tests. The PRIME Care trial — the largest real-world study, run in Veterans Affairs clinics — found testing reduced drug-gene mismatches but produced only modest remission gains that didn’t hold at 24 weeks.
Practice Implications
- Pharmacogenomic testing is useful for a few genes and a few medications.
- The metabolic genes tell us about serum levels, much like drug interactions do, and checking the serum level is a more direct measure.
—Chris Aiken, MD
Director, Psych Partners
Editor in Chief, Carlat Psychiatry Report
