Pharmacogenomic (PGx) Testing: What a DNA Medication-Response Test Shows, What the FDA Says, and Its Limits

A pharmacogenomic test reads genes that change how your body handles certain drugs. For a short list of gene-drug pairs the evidence is strong enough to change a prescription. For most medicines it is not, and the FDA has warned against tests that claim otherwise.

By Kymata Health Editorial Published Last checked

This guide

Blood Testing & Biomarkers
Question
What does a pharmacogenomic (PGx) DNA test tell you about your medications?
Sources
12 cited
Quick answers
4 questions
Clinician review
None; sources cited inline

Researched and written by Kymata Health Editorial; sources cited inline. Drawn from the studies and documents cited below. Educational only, not medical advice.

Key takeaways

5 points
  • FDA lists gene-drug pairs where the data support prescribing changes, such as HLA-B*57:01 and abacavir or CYP2C19 and clopidogrel, but says listing a pair does not mean it advocates testing first1.
  • Screening for HLA-B*57:01 before abacavir cut immunologically confirmed hypersensitivity from 2.7% to 0% in a randomized trial of 1,956 people7.
  • Across 6,944 patients, prescribing guided by a 12-gene panel lowered clinically relevant adverse drug reactions (odds ratio 0.70)8.
  • In depression the benefit is small: in a VA trial of 1,944 patients, remission was higher over 24 weeks but not at week 24 itself10.
  • FDA authorized a direct-to-consumer PGx report in 2018 with a warning not to use it to change medication on your own2, and warned a lab in 2019 for unsupported drug-response claims3.

What the test reads

A PGx test, usually a cheek swab or saliva sample, looks for variants in genes that affect drug metabolism, drug transport or the risk of specific reactions. Reports usually sort you into groups such as poor, intermediate, normal or ultrarapid metabolizer for enzymes like CYP2D6 and CYP2C191.

That label is a prediction about drug levels, not a verdict on whether a medicine will work. FDA notes that genotype is only one of many factors that affect drug concentrations and response, and that most associations it lists have not been evaluated for their effect on clinical outcomes1.

Where the evidence is strong

FDA's Table of Pharmacogenetic Associations, updated Sept. 10, 2026, separates pairs where data support therapeutic recommendations from pairs with safety or response signals and pairs that affect drug levels only1. Examples from the first group:

  • Abacavir (HIV): do not use in people positive for HLA-B*57:011. In the PREDICT-1 trial, screening eliminated immunologically confirmed hypersensitivity (0% vs 2.7%)7.
  • Clopidogrel (heart and stroke): CYP2C19 intermediate and poor metabolizers form less active drug and may have higher cardiovascular risk; FDA says to consider another P2Y12 inhibitor1, and the CPIC guideline strengthened its advice for intermediate metabolizers in 20225.
  • Codeine and tramadol: CYP2D6 ultrarapid metabolizers can reach dangerous active-metabolite levels; both are contraindicated in children under 121.
  • Citalopram: CYP2C19 poor metabolizers have a maximum recommended dose of 20 mg because of QT prolongation risk1.
  • Thiopurines (azathioprine, mercaptopurine): TPMT or NUDT15 poor metabolizers face a higher risk of bone-marrow suppression and generally tolerate 10% or less of the usual mercaptopurine dose1.

What CPIC guidelines do and do not do

The Clinical Pharmacogenetics Implementation Consortium (CPIC) publishes peer-reviewed guidelines, 29 at last count, that tell clinicians how to use genetic results they already have; they deliberately do not address whether a test should be ordered4.

CPIC's 2023 antidepressant guideline gives prescribing recommendations for CYP2D6, CYP2C19 and CYP2B6, and concludes that the data for SLC6A4 and HTR2A do not support their use in antidepressant prescribing6. If a report ranks your antidepressants using those genes, it goes beyond the guideline.

Does testing improve outcomes?

For side effects, the best evidence is the PREPARE study: 6,944 patients in several European countries got either genotype-guided prescribing with a 12-gene panel or standard care. Among patients with an actionable result, clinically relevant adverse drug reactions occurred in 21.0% vs 27.7% (odds ratio 0.70)8. It was open-label, and nearly all participants were of European, Mediterranean or Middle Eastern ancestry8.

For depression, results are modest. In GUIDED (1,167 patients), the primary measure, symptom improvement at week 8, did not differ significantly (27.2% vs 24.4%), though response and remission were higher9; several authors were employees of Myriad Genetics, a genetic testing company9. In the VA's PRIME Care trial (1,944 patients), testing reduced prescriptions with predicted drug-gene interactions, and remission over 24 weeks was higher (risk difference 2.8%), but the difference at week 24 was not significant10. The lead author reported nonfinancial support from Myriad Genetics10.

The FDA's position

In 2018 FDA authorized 23andMe's pharmacogenetic reports for 33 variants as a consumer test to inform conversations with a clinician, while stating that it does not determine whether a medication is right for you and that results should be confirmed with clinical pharmacogenetic testing before any medical decision2. FDA required a warning that consumers should not use the results to stop or change any medication2.

In 2019 FDA sent a warning letter to Inova Genomics Laboratory for marketing tests that claimed to predict responses to named drugs, including antidepressants and opioids, without FDA review3. FDA said that after its 2018 safety communication most firms removed specific drug names from their labeling and reports3.

A PGx panel designed and run inside a single lab is a laboratory developed test. FDA's 2024 rule that would have treated such tests as medical devices was vacated by a federal court on March 31, 2025, and FDA formally reverted the regulation on Sept. 19, 202512.

Limits and privacy

A panel only reports the variants it tests for, and FDA's table covers a limited set of drugs and genes1. A normal result does not mean a medicine is safe for you: FDA notes that genotype is only one of many factors that affect drug levels and response1.

Genetic data carry some protection: GINA bars health insurers from using genetic information for coverage or premiums and bars most employers from using it in employment decisions, but it does not cover life, disability or long-term care insurance, and it does not apply to employers with fewer than 15 employees11.

Who might benefit, and how to use a result safely

Testing is most useful before or while starting a drug that has an FDA-labeled or CPIC-guided gene interaction, such as clopidogrel, abacavir, codeine, certain antidepressants or thiopurines14. ClarityX, for example, sells cheek-swab PGx panels that a physician network orders; we found no FDA clearance for them, so treat the report as information for your prescriber.

Never stop, start or change a dose on the strength of a PGx report; FDA says patients should not adjust medications without consulting their prescriber1. Bring the report to the clinician or pharmacist who manages the drug, ask which findings have CPIC or FDA-label support, and expect that a result from a consumer test may need confirming with a clinical lab2. People taking antidepressants, opioids, cancer drugs or diabetes medicines should be especially careful; FDA has warned that changing a drug or dose based on unproven test claims could cause serious harm3.

Where to buy

ClarityX

  • At-home cheek-swab PGx panels: Mindwell ($399, mental-health medications) and Max Rx ($499, 285+ medications), checked Oct 13, 2026. A physician network orders each test and a pharmacist reviews results. It is a laboratory-developed test, not FDA-cleared. Treat the report as something to discuss with your prescriber or pharmacist, ask which findings have CPIC or FDA-label support, and do not stop, start or change a medication on your own.

    See ClarityX PGx tests (paid link, opens in new tab)

Quick answers

Questions

4 answers

Can a DNA test tell me which antidepressant will work?

No test can promise that. CPIC gives dosing advice for CYP2D6, CYP2C19 and CYP2B6 results6, and trials found small, mixed gains in remission910.

Is a PGx test FDA-approved?

Not as a rule. The 23andMe reports FDA authorized in 2018 were the first direct-to-consumer PGx test it authorized2, and FDA's 2024 rule to regulate laboratory developed tests as devices was vacated in 202512.

Should I stop a medicine if my report flags it?

No. FDA says not to adjust medications without your prescriber1, and the 23andMe authorization required a warning against stopping or changing medicines based on results2.

Can my employer or health insurer see or use my results?

GINA bars health insurers and employers with 15 or more employees from using genetic information, but not life, disability or long-term care insurers11.

More in Blood Testing & Biomarkers

Keep reading

5 pages

Supplements and vitamins: every pageSupplement & Nutrient EncyclopediaBrowse all topics

Evidence-graded supplement stacks

References

Every study and document cited in this guide, numbered as in the text.

Sources

12 sources
  1. 1Table of Pharmacogenetic Associations. U.S. Food and Drug Administration, 2026.
  2. 2FDA authorizes first direct-to-consumer test for detecting genetic variants that may be associated with medication metabolism. U.S. Food and Drug Administration, 2018.
  3. 3FDA issues warning letter to genomics lab for illegally marketing genetic test that claims to predict patients' responses to specific medications. U.S. Food and Drug Administration, 2019.
  4. 4CPIC Guidelines. Clinical Pharmacogenetics Implementation Consortium, 2026.
  5. 5Clinical Pharmacogenetics Implementation Consortium Guideline for CYP2C19 Genotype and Clopidogrel Therapy: 2022 Update. Clin Pharmacol Ther, 2022. PMID 35034351
  6. 6Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for CYP2D6, CYP2C19, CYP2B6, SLC6A4, and HTR2A Genotypes and Serotonin Reuptake Inhibitor Antidepressants. Clin Pharmacol Ther, 2023. PMID 37032427
  7. 7HLA-B*5701 screening for hypersensitivity to abacavir. N Engl J Med, 2008. PMID 18256392
  8. 8A 12-gene pharmacogenetic panel to prevent adverse drug reactions: an open-label, multicentre, controlled, cluster-randomised crossover implementation study. Lancet, 2023. PMID 36739136
  9. 9Impact of pharmacogenomics on clinical outcomes in major depressive disorder in the GUIDED trial: A large, patient- and rater-blinded, randomized, controlled study. J Psychiatr Res, 2019. PMID 30677646
  10. 10Effect of Pharmacogenomic Testing for Drug-Gene Interactions on Medication Selection and Remission of Symptoms in Major Depressive Disorder: The PRIME Care Randomized Clinical Trial. JAMA, 2022. PMID 35819423
  11. 11Genetic Discrimination. National Human Genome Research Institute, 2022.
  12. 12Laboratory Developed Tests. U.S. Food and Drug Administration, 2025.

Dietary supplements are not intended to diagnose, treat, cure or prevent any disease, and these statements have not been evaluated by the FDA. Prescription options need a licensed clinician. Talk to your clinician or pharmacist before starting anything new, especially if you take prescription medicines, are pregnant or breastfeeding.

Educational only. Not medical advice. Some links earn us a commission: kymatahealth.com/affiliate-disclosure.