Pharmacogenomic Testing in Bipolar Disorder

Pharmacogenomic tests such as GeneSight look at genes that affect how your body handles medications. For bipolar disorder, they are more limited than many people expect. The core mood stabilizers, lithium, lamotrigine and valproate, are mostly not cleared by the liver enzymes these panels focus on, and no gene test can diagnose bipolar disorder. A few genetic results do matter a great deal, especially the HLA markers linked to dangerous skin reactions with carbamazepine, oxcarbazepine and lamotrigine. This guide explains what the FDA and the CPIC say, and what a report can and cannot tell you.

What genetic testing cannot do

Bipolar disorder runs in families: family and twin studies estimate heritability at roughly 60–85%. But it is polygenic, shaped by many genetic variants of small effect plus life experience, and NIMH states that no one gene causes it. So there is no gene test that diagnoses bipolar disorder, in you or in your child. Our page on what causes bipolar disorder explains the genetics and environment in more depth.

Pharmacogenomic panels answer a different question: how your body may process or respond to specific drugs. Even there, they cannot tell you which medication will work. The GeneSight company itself says its results do not provide information on disease diagnosis or which medications will work best.

Why the main mood stabilizers are a poor fit for CYP panels

Most commercial panels emphasize cytochrome P450 (CYP) liver enzymes such as CYP2D6 and CYP2C19, which matter for many antidepressants. The bipolar medications people take most often are handled differently:

The FDA's Table of Pharmacogenomic Biomarkers in Drug Labeling reflects this. Lithium, quetiapine, olanzapine, lurasidone, asenapine and lumateperone do not appear on it at all. For valproate, the listing is POLG (mitochondrial disease), as a boxed warning and contraindication, along with urea cycle disorders.

HLA-B*15:02 and HLA-A*31:01: the results that matter most

The clearest genetic safety test in bipolar care concerns rare but life-threatening skin reactions, Stevens-Johnson syndrome and toxic epidermal necrolysis (SJS/TEN):

• Carbamazepine: the label has a boxed warning linking SJS/TEN to HLA-B*15:02. Risk in patients of Asian descent is about 10 times higher, and the label says to test before starting in patients likely to carry the allele and to avoid the drug if they test positive unless the benefit clearly outweighs the risk. The FDA table also lists HLA-A (HLA-A*31:01) in carbamazepine's warnings.
• Oxcarbazepine: HLA-B appears in its Warnings and Precautions.
• Lamotrigine: since October 2025, the Lamictal boxed warning lists the HLA-B*15:02 allele as a risk factor for serious rash, alongside taking it with valproate and exceeding the recommended starting dose or dose increases. Serious rash can happen without any of these.

CPIC (the Clinical Pharmacogenetics Implementation Consortium) has a guideline for HLA-A, HLA-B and carbamazepine/oxcarbazepine, rated Level A for carbamazepine (HLA-A and HLA-B) and oxcarbazepine (HLA-B). CPIC does not have an HLA-B pairing for lamotrigine; the change came through the FDA label. A negative HLA result never makes rash impossible, so the standard rash precautions still apply; see lamotrigine for bipolar disorder.

CYP2D6 and antipsychotics

CYP2D6 is where pharmacogenomics touches antipsychotic prescribing. The FDA table lists CYP2D6 for aripiprazole, aripiprazole lauroxil and brexpiprazole (in dosing sections), cariprazine (clinical pharmacology only), clozapine, iloperidone, paliperidone and risperidone, as well as fluoxetine. In plain terms, a person who processes CYP2D6 drugs very slowly may need a different dose of some of these.

CPIC rates aripiprazole and risperidone with CYP2D6 at Level B and brexpiprazole at B/C, and it has no guideline for any antipsychotic. For antidepressants, CPIC's SSRI guideline covers CYP2D6 and CYP2C19; fluoxetine pairings are Level C and are not used for recommendations. For lamotrigine, CPIC lists only UGT1A4 at Level D. For the drugs themselves, see atypical antipsychotics in bipolar disorder.

Where GeneSight fits in bipolar care

A GeneSight report sorts many psychiatric medications into categories and shows the genes behind each placement, and it includes HLA markers linked to skin reactions with certain mood stabilizers. Our guides explain what GeneSight testing can and cannot do and how to read a GeneSight report.

For bipolar disorder specifically, a report can be helpful when:

• carbamazepine or oxcarbazepine is being considered, and HLA status is not already known;
• an antipsychotic with CYP2D6 dosing guidance is planned, especially after unusual side effects at ordinary doses;
• antidepressants are part of a plan for bipolar depression, alongside a mood stabilizer;
• several medications are being combined and the prescriber wants extra information.

It adds little to choosing between lithium, lamotrigine and valproate, and nothing to the diagnosis. The FDA and CPIC positions are covered in more depth in what FDA and CPIC guidance says about pharmacogenomic testing.

Family history, ancestry and test results

Two things often get mixed up. Family history of bipolar disorder raises a person's chance of developing it (first-degree relatives had several times the risk in a large Swedish registry study), but that is about illness risk, not about how you process drugs. Ancestry, on the other hand, matters for the HLA results: the carbamazepine label ties its testing advice to patients likely to carry HLA-B*15:02, which is found mainly in people of Asian ancestry.

If a relative had a serious skin reaction to an anticonvulsant, or did unusually well or badly on a specific medication, tell your prescriber. That history can be as useful as a lab report. And if you have already had pharmacogenomic testing, bring the full report rather than a summary, because the gene pages explain why each drug was placed where it was.

Questions to ask before testing

• Which medication decision would this result actually change for me?
• Is a single-gene HLA test enough, or would a broader panel add anything?
• How will the results be stored in my record so future prescribers can see them?
• Will my insurance or the testing company cover it, and what would I owe? (The testing company bills separately.)

Never change a medication based on a report alone; the GeneSight report itself says medications should not be changed based solely on the results. See bipolar medications explained for how each drug is chosen.

At Willow & Stone

At Willow & Stone, GeneSight pharmacogenomic testing is offered when it is clinically useful, and results are reviewed alongside your history rather than on their own. Learn about bipolar disorder treatment at Willow & Stone, start with the complete bipolar disorder guide, or book a consultation.

Common Questions

Sources & Further Reading

• FDA: Table of Pharmacogenomic Biomarkers in Drug Labeling (content current 08/12/2026)
• CPIC: Guideline for HLA genotype and carbamazepine/oxcarbazepine
• DailyMed: Lamictal (lamotrigine) prescribing information (10/2025)
• DailyMed: Equetro (carbamazepine ER) prescribing information (8/2025)
• NIMH: Bipolar Disorder (NIH Pub. 25-MH-8088, 2025)
• DailyMed: Depakote (divalproex) prescribing information (3/2026)

This article is general education, not a diagnosis or advice for your situation. Do not start, stop, or change a medication without talking to the clinician who prescribes it; stopping a mood stabilizer suddenly can trigger relapse.