UnitedHealthcare Pharmacogenomics Coverage Update

Coverage Summary

Payer: UnitedHealthcare. Policy: Pharmacogenomics Testing (Policy Number: MMP391.16), Effective Date: 2026-03-01. Scope: pharmacogenetic/pharmacogenomic testing (single-gene, multi-gene panels, and combinatorial tests) to determine how genes affect drug response.

Coverage stance: Mixed. Tests that meet an actionability standard (CPIC guideline level A or B or listed FDA pharmacogenomic/pharmacogenetic associations with therapeutic recommendations) may be considered reasonable and necessary when results would directly impact medication management; other tests or panels with insufficient analytical validity, clinical validity, or clinical utility are considered not reasonable and necessary.

Example covered/provisionally covered single-gene test: TPMT (CPT 81335) — CPIC level A and listed in FDA pharmacogenomic associations with dosing recommendations. Examples called out as not reasonable/necessary: urinary biomarker panel Foundation PI (FPI) and commercial psychiatric panels such as Psych HealthPGx and Genomind Professional PGx Express CORE.

Medical-Necessity Criteria

Covered Indications / Medical Necessity

Pharmacogenetics testing will be considered reasonable and necessary if:

Covered Indications / Medical Necessity - TOP

Clinical need and actionability: The patient has a condition where clinical evaluation has determined the need for a medication that has a known Gene-drug interaction(s) for which the test results would directly impact the drug management of the patient's condition
CPIC level A or B OR FDA table of pharmacogenomic associations or FDA pharmacogenetic associations
Evidence standard for actionability: The test meets evidence standards for genetic testing as evaluated by a scientific, transparent, peer-reviewed process and is determined to demonstrate actionability in clinical decision making by Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline level A or B; OR is listed in the U.S. Food and Drug Administration (FDA) table of known Gene-drug interactions where data support therapeutic recommendations or a potential impact on safety or response or the FDA label
Preserve note about CPIC level A or B or FDA pharmacogenomic associations

Specific Gene Examples with Coverage Implications

Policy commentary and guidance on certain genes/tests:

Specific Gene Examples with Coverage Implications - TOP

TPMT genotype testing: CPIC level A — testing recommended; CPIC recommends adjusting starting doses of thiopurines (mercaptopurine, azathioprine, thioguanine) based on TPMT genotype
CYP1A2: genotype polymorphisms do not have a clinically meaningful effect on the pharmacokinetics of rucaparib (not actionable)
CYP3A4: insufficient evidence to provide recommendations for dosing statins (CPIC level C: no recommendation)
COMT: no therapeutic recommendations for dosing opioids based on COMT genotype (CPIC level C: no recommendation)
TYMS: no recommendations provided for capecitabine and fluorouracil (CPIC Provisional Level D: no recommendation)

Not Covered / Non‑Reasonable and Necessary

Not Covered / Non-Reasonable and Necessary

Tests considered not reasonable and necessary when:

Not Covered / Non-Reasonable and Necessary - TOP

Lack of validation: Genetic testing where either Analytical Validity, Clinical Validity, or Clinical Utility has not been established
Foundation PI urinary biomarker tests: Urinary biomarker laboratory tests for chronic pain (e.g., Foundation PI) are not reasonable and necessary
Psychiatric panels not reasonable: Psych HealthPGx Panel and Genomind Professional PGx Express CORE panels are not reasonable and necessary for pharmacogenomic testing due to insufficient evidence of efficacy

This policy reflects a mixed coverage stance: listing of procedure codes or tests in the policy is for reference and does not guarantee coverage; benefit coverage is determined by the member-specific plan and applicable laws.

Evidence-based Conclusions and Implications

Evidence-based conclusions and implications

Summary findings and evidence certainty statements derived from systematic reviews, RCTs, and evaluations referenced in this part:

Evidence-based conclusions and implications - TOP

Key evidence points: Pharmacogenetic panel testing (psychiatry): aggregate evidence shows some short-term improvements in response/remission (8-12 weeks) in MDD in several meta-analyses and RCTs but overall evidence has heterogeneity, risk of bias, variable panel composition, and inconsistent long-term outcomes.
Include nested 'Key evidence points' branch
Urinary biomarker evidence: Urinary biomarker (Foundation Pain Index, FPI) tests: evidence is low/very low quality and clinical utility/outcomes are inconclusive; studies include retrospective observational cohorts and small RCTs with limitations (selection bias, single-center, industry funding, lack of longitudinal outcomes).
Multiple studies (Pope 2021; Gunn 2020; Amirdelfan 2020; Peabody 2020) reported associations with PROMIS-29 or practice changes but with methodological limitations.

Coding

Non-Covered / Referenced Procedure CodesCPTNot Covered

0031U	CYP1A2 (cytochrome P450 family 1, subfamily A, member 2) (e.g., drug metabolism) gene analysis, common variants (i.e., 1F, 1K, 6, 7)
0032U	COMT (catechol-O-methyltransferase) (drug metabolism) gene analysis, c.472G > A (rs4680) variant
0033U	HTR2A, HTR2C gene analysis panel (Deleted 12/31/2025)
0117U	Pain management, analysis of 11 endogenous analytes, LC-MS/MS, urine, algorithm reported as a pain-index score (Foundation PI SM)
0173U	Psychiatry genomic analysis panel, includes variant analysis of 14 genes (Psych HealthPGx Panel)
0175U	Psychiatry genomic analysis panel, variant analysis of 15 genes (Genomind Professional PGx Express CORE)
81230	CYP3A4 gene analysis, common variant(s) (e.g., 2, 22)
81346	TYMS (thymidylate synthetase) gene analysis, common variant(s) (e.g., tandem repeat variant)

Provisional Coverage / Covered ExampleCPTCovered

81335

TPMT (thiopurine S-methyltransferase) gene analysis, common variants (e.g., *2, *3)

Diagnosis Codes (Associated/Referenced for CPT 81335)ICD-10

C91.00	Acute lymphoblastic leukemia not having achieved remission
C91.01	Acute lymphoblastic leukemia, in remission
C91.02	Acute lymphoblastic leukemia, in relapse
C91.10	Chronic lymphocytic leukemia of B-cell type not having achieved remission (Deleted 07/12/2025) (Effective 01/01/2024)
C91.11	Chronic lymphocytic leukemia of B-cell type in remission (Deleted 07/12/2025) (Effective 01/01/2024)
C91.12	Chronic lymphocytic leukemia of B-cell type in relapse (Deleted 07/12/2025)
C91.30	Prolymphocytic leukemia of B-cell type not having achieved remission (Deleted 07/12/2025)
C91.40	Hairy cell leukemia not having achieved remission
C91.50	Adult T-cell lymphoma/leukemia (HTLV-1-associated) not having achieved remission
C91.60	Prolymphocytic leukemia of T-cell type not having achieved remission

1–10 of 20

1/2

Provider Actions & Billing Guidance

Documentation Required

Clinical need and linkage to actionable gene-drug interaction

Provider must document that the patient has a condition requiring a medication with a known gene–drug interaction and that test results would directly impact drug management. Testing must be linked to an actionable source: CPIC guideline level A or B OR an entry in the FDA table of pharmacogenetic/pharmacogenomic associations that supports therapeutic recommendations or potential impact on safety/response.

Actionability standard: CPIC level A or B or FDA pharmacogenetic/pharmacogenomic table linkage
Example: TPMT (CPIC level A; listed in FDA Table of Pharmacogenomic Associations)

Documentation Required

Analytic/Clinical Validity and Clinical Utility requirement

Providers and laboratories must ensure and be able to demonstrate that the test meets standards of Analytical Validity (AV), Clinical Validity (CV), and Clinical Utility (CU) to satisfy the reasonable and necessary standard. Documentation of validation/verification and evidence supporting clinical utility should be retained and made available on request.

Definitions

Definitions — brief entries:

Actionable Use: Genotype information that may lead to selection/avoidance or dose modification of therapy based on FDA label, FDA warning/safety concern, or CPIC level A or B gene–drug interaction.

Analytical Validity (AV): Assessment of a test's technical performance (sensitivity, specificity, accuracy, precision) to detect the variant.

Clinical Utility (CU): Ability of a test to provide information that informs treatment decisions and improves health outcomes.

Clinical Validity (CV): Ability of a test to classify a patient's circumstance into a diagnostic, prognostic, or predictive functional category in a defined population.

Combinatorial PGx Test: A multi-gene panel requiring a proprietary algorithm to evaluate relationships that result in drug recommendations or warnings.

Multi-Gene Panel: Laboratory test detecting variants of at least two genes ordered as a specified panel.

Single-Gene Test: Test detecting relevant variants of one gene (multiple single-gene orders are not considered a panel).

Provisional CPIC Level Status: Initial provisional assignment of CPIC levels (A–D) until sufficient review yields a final status.

Foundation Pain Index (FPI): A proprietary urine-based multi-biomarker panel intended to detect biochemical abnormalities associated with chronic pain.

Pharmacogenomic (PGx) testing: Genetic testing panels that assess variants in drug-metabolizing enzymes and other genes to guide medication selection/dosing (commonly includes CYP2C19, CYP2D6, etc.).

Evidence & Key References

Evidence / Source	Key finding or value
Morris et al., 2022 (Cost-effectiveness systematic review)
71% of 108 studies found PGx-guided treatment cost-effective (44%) or cost-saving (27%); variation by drug (clopidogrel favorable; warfarin less so).
CPIC TPMT recommendations (Relling et al., 2019; 2024 update)
CPIC recommends adjusting thiopurine starting doses based on TPMT genotype; 2024 update: start 20%-50% for TPMT/NUDT15 intermediate metabolizers depending on dose.
Peabody et al., 2020 (RCT of FPI)
Reported increased diagnosis/treatment of micronutrient/metabolic abnormalities and more non-opioid treatments after FPI results, but study design and external validity limitations.
Gunn et al., 2020 (retrospective cohort on FPI)
Retrospective analysis of 17,834 samples: 77% had ≥1 abnormal biomarker; observational design with confounding and industry involvement.
Amirdelfan et al., 2020 (cross-sectional validation of FPI)
Cross-sectional study (153 chronic pain, 334 controls) with AUROC 0.7490 distinguishing pain vs control; limited validation and outcome data.
Zhang et al., 2025 (meta-analysis, psychiatry)
Meta-analysis of 13 RCTs: improved response at 8 and 12 weeks and remission at 8 weeks for PGx-guided antidepressant treatment; heterogeneity and panel variability noted.
Xu et al., 2024 (large RCT, n=665)
RCT showed higher response/remission and fewer ADRs at 8 & 12 weeks with PGx-guided therapy; limitations: lack of blinding, single-center, 12-week follow-up.
ECRI Genetic Test Assessment (2023)
Concluded evidence is inconclusive for the Foundation Pain Index (FPI); clinical validity and health outcomes not established.

Medicare Determinations

Name	Number	Type	Effective Date
Pharmacogenomics Testing for Warfarin Response
90.1
NCD

Pharmacogenomics Testing (multiple LCDs listed)
L39073 / L39063 / L39995 / L39616 / L36021 / L36021 / L35160 / L36256 / L35025 / L36807 / L34519 / L35062 / L38288 / L38294 / L38335 / L38337 / L38351 / L38353 / L38435 / L38429 / L38394
LCD

Revision History

2026-03-01effective_dateLatest

Policy Effective Date: March 1, 2026

2026-02-11last_review

Last Committee Approval Date: February 11, 2026