CurrentUnitedHealthcarePolicy CS152PA.AC

Molecular Oncology Testing for Solid Tumor Cancer Diagnosis, Prognosis, and Treatment Decisions (for Pennsylvania Only)

State-specific UnitedHealthcare policy (Pennsylvania only) defining medical necessity criteria for various molecular and gene expression profiling tests for solid tumors (breast, lung, prostate, thyroid, uveal melanoma) and listing tests/procedures considered proven/medically necessary vs unproven/not medically necessary; includes extensive applicable CPT/PLA/algorithmic U-codes and molecular pathology codes for reference.

Policy Summary

PayerUnitedHealthcare

PolicyMolecular Oncology Testing for Solid Tumor Cancer Diagnosis, Prognosis, and Treatment Decisions (for Pennsylvania Only)

Policy CodePolicy CS152PA.AC

Change TypeCPT/proprietary code additions; supporting info updated

Effective DateFeb 1, 2026

Next Review Date

Key ActionUse gene expression and molecular assays only in indications supported by the policy and document clinical factors and how results will influence management.

POLICY UPDATE CHANGES

Updated list of applicable CPT codes to reflect quarterly edits; added 0578U, 0585U, 0586U, 0592U, and 0597U.

Updated Supporting Information and archived previous policy version CS152PA.AB.

5Covered Indications explicitly listed (breast, lung, prostate, thyroid, uveal melanoma)

manyPLA/U codes listed in part

multipleTests categorized as unproven/not medically necessary (partial list)

CPT/Proprietary Codes Added

multipleTypes of evidence summarized (RCTs, systematic reviews, observational)

Coverage Summary

This is a Pennsylvania‑specific UnitedHealthcare Community Plan medical policy (Policy CS152PA.AC) that defines when molecular and gene expression profiling (GEP) and related molecular oncology tests are considered medically necessary versus unproven/not medically necessary for diagnosis, prognosis, and treatment decisions in solid tumor cancers. The policy explicitly covers GEP/CGP and related testing in the following tumor types: breast, lung (NSCLC), prostate, thyroid, and uveal melanoma. The coverage stance is mixed: certain tests and specific indications are considered proven and medically necessary (for example, validated breast GEPs in defined early ER+/HER2‑ settings; NSCLC targeted panels/ctDNA in appropriate contexts; selected prostate GEPs for risk stratification and Decipher post‑prostatectomy), while many other assays and broader applications (including numerous multi‑cancer detection, MRD, and large unvalidated CGP uses) are categorized as unproven and not medically necessary due to insufficient evidence.

Selected high-level stats

Covered indications count5

Major breast GEAs discussedOncotype DX, Prosigna (PAM50), MammaPrint

CPT/HCPCS molecular assay codes listed25+ (policy lists many CPT/PLA/U and HCPCS codes; multiple 0005U–0597U and 81445–81599 entries)

Tests categorized as unprovenMultiple (e.g., ExoDx Prostate, MyProstateScore, Signatera, Galleri, PancreaSeq, CancerTYPE ID, MRD assays)

CPT/Proprietary Codes Added5 (0578U, 0585U, 0586U, 0592U, 0597U)

Medical‑Necessity Criteria

Breast Cancer Gene Expression Profiling (GEP) - Medically Necessary

Breast Cancer Gene Expression Profiling (GEP) — Medically Necessary

Use of one of the following GEP tests: MammaPrint, Oncotype Dx Breast, Prosigna (PAM-50), Breast Cancer Index (BCI), EndoPredict to inform treatment decisions in individuals with invasive breast cancer.

ONE of

Newly diagnosed (within the last 6 months)
- Lymph node negative (including lymph nodes with micrometastases no greater than 2 mm) or 1-3 positive ipsilateral axillary lymph nodes.
- No distant metastases.
- Hormone receptor-positive (estrogen receptor positive, progesterone receptor positive, or both).

Coding — Applicable Molecular, PLA/U and CPT/HCPCS Codes

Applicable PLA/U codes (selected examples listed in this part)mixed

0005U	Oncology (prostate) gene expression profile by real-time RT-PCR of 3 genes (ERG, PCA3, and SPDEF), urine, algorithm reported as risk score.
0011M	Oncology, prostate cancer, mRNA expression assay of 12 genes (10 content and 2 housekeeping), RT-PCR test utilizing blood plasma and urine, algorithms to predict high-grade prostate cancer risk.
0012M	Oncology (urothelial), mRNA, gene expression profiling by RT-qPCR of five genes ... utilizing urine, algorithm reported as a risk score for having urothelial carcinoma.
0018U	Oncology (thyroid), microRNA profiling by RT-PCR of 10 microRNA sequences, utilizing fine needle aspirate, algorithm reported as a positive or negative result for moderate to high risk of malignancy.
0022U	Targeted genomic sequence analysis panel, non-small cell lung neoplasia, DNA and RNA analysis, 23 genes, interrogation for sequence variants and rearrangements.
0037U	Targeted genomic sequence analysis, solid organ neoplasm, DNA analysis of 324 genes, interrogation for sequence variants, gene copy number amplifications, gene rearrangements, microsatellite instability and tumor mutational burden.
0045U	Oncology (breast ductal carcinoma in situ), mRNA, gene expression profiling by RT-PCR of 12 genes ... algorithm reported as recurrence score.
0179U	Oncology (non-small cell lung cancer), cell-free DNA, targeted sequence analysis of 23 genes ... report of significant mutation(s).
0211U	Oncology (pan-tumor), DNA and RNA by next-generation sequencing, interpretative report for SNVs, CNAs, TMB, MSI, with therapy association.
0244U	Oncology (solid organ), DNA, comprehensive genomic profiling, 257 genes ... utilizing FFPE tumor tissue.

1–10 of 13

1/2

Applicable CPT / Molecular pathology codesCPT

81445	Solid organ neoplasm, genomic sequence analysis panel, 5-50 genes, interrogation for sequence variants and copy number variants or rearrangements; DNA analysis or combined DNA and RNA analysis.
81449	Solid organ neoplasm, genomic sequence analysis panel, 5-50 genes; RNA analysis.
81455	Solid organ or hematolymphoid neoplasm or disorder, 51 or greater genes, genomic sequence analysis panel...
81456	51 or greater genes, RNA analysis.
81457	Solid organ neoplasm, genomic sequence analysis panel, interrogation for sequence variants; DNA analysis, microsatellite instability.
81458	Solid organ neoplasm, genomic sequence analysis panel ... copy number variants and microsatellite instability.
81459	Solid organ neoplasm, genomic sequence analysis panel ... tumor mutation burden, and rearrangements.
81462	Solid organ neoplasm, genomic sequence analysis panel, cell-free nucleic acid (e.g., plasma), interrogation for sequence variants; DNA analysis or combined DNA and RNA analysis, copy number variants and rearrangements.
81463	Solid organ neoplasm, genomic sequence analysis panel, cell-free nucleic acid, DNA analysis, copy number variants, and microsatellite instability.
81464	Solid organ neoplasm, genomic sequence analysis panel, cell-free nucleic acid, tumor mutation burden, and rearrangements.

1–10 of 12

1/2

Listed applicable CPT codes for molecular oncology and multigene assaysCPT

81456	Solid organ or hematolymphoid neoplasm or disorder, 51 or greater genes, genomic sequence analysis panel, RNA analysis.
81457	Solid organ neoplasm, genomic sequence analysis panel; DNA analysis, microsatellite instability.
81458	Solid organ neoplasm, genomic sequence analysis panel; DNA analysis, copy number variants and microsatellite instability.
81459	Solid organ neoplasm, genomic sequence analysis panel; DNA or combined DNA/RNA analysis, copy number variants, MSI, TMB, rearrangements.
81462	Cell-free nucleic acid genomic sequence panel (plasma); DNA or combined DNA/RNA analysis.
81463	Cell-free nucleic acid genomic sequence panel; DNA analysis, CNV, MSI.
81464	Cell-free genomic panel; DNA or combined DNA/RNA analysis, CNV, MSI, TMB, rearrangements.
81479	Unlisted molecular pathology procedure.
81504	Oncology (tissue of origin), microarray gene expression profiling >2000 genes.
81518	Oncology (breast), 11-gene mRNA RT-PCR assay; reported as % risk for metastatic recurrence and EET benefit.

1–10 of 24

1/3

HCPCS codeHCPCS

G0327

Colorectal cancer screening; blood-based biomarker.

Updated Applicable CPT / Proprietary CodesCPT

0578U	Proprietary laboratory test code (added to applicable CPT list per revision)
0585U	Proprietary laboratory test code (added to applicable CPT list per revision)
0586U	Proprietary laboratory test code (added to applicable CPT list per revision)
0592U	Proprietary laboratory test code (added to applicable CPT list per revision)
0597U	Proprietary laboratory test code (added to applicable CPT list per revision)

Provider Actions — Authorization, Documentation, Billing & Denial Risks

Documentation Required

Pennsylvania-only application

This policy applies only to services in Pennsylvania. Requests for services that fall outside the Pennsylvania-specific criteria will be evaluated case-by-case and may be subject to Pennsylvania Exceptions and Pennsylvania Code, Title 55, Chapter 1101. Refer to federal, state, and contractual benefit terms when determining coverage and prior authorization requirements. The payer reserves the right to modify the policy and its operational requirements.

Documentation Required

Breast GEP pre-test discussion for extended endocrine therapy

For breast cancer gene expression profiling (GEP) used to inform decisions about extended adjuvant endocrine therapy, the ordering clinician and the patient must have a documented discussion prior to testing about the potential results and how the results will be used to guide decisions regarding extended hormonal therapy. Document that the test result is intended to influence the decision about extended endocrine therapy.

Background and Evidence Summaries

The policy emphasizes limitations on ordering multiple overlapping predictive GEPs for the same tumor, and restricts use of large NGS/CGP panels (>50 genes) except where aligned with FDA‑cleared/approved companion diagnostic policy or specified indications. Liquid biopsy (cfDNA/ctDNA) and comprehensive genomic profiling (CGP) are addressed with tumor‑ and context‑specific applicability (e.g., tissue or plasma‑based panels for NSCLC, tumor‑informed approaches for MRD are considered investigational). The evidence base summarized in the policy includes randomized controlled trials, large prospective trials and prospective‑retrospective studies (e.g., MINDACT, TAILORx, NRG/RTOG ancillary analyses), systematic reviews and HTA assessments (e.g., Hyams 2024, Hayes assessments), and observational and registry studies informing clinical utility and practice impact.

Clinical utility and evidence summaries for breast cancer gene expression assays

Summaries of high-level evidence, guideline positions, and assay-specific findings related to breast gene expression assays and their clinical utility.

Evidence summaries (breast GEAs): Oncotype DX (21-gene RS): prognostic for distant recurrence and overall survival in early ER+/HER2- breast cancer; high-level evidence for prognostication; only assay with high LOE supporting prediction of chemotherapy benefit from large prospective trials (TAILORx and others). Oncotype DX: evidence supports utility in node-negative (N0) and limited node-positive (N1, 1-3 nodes) disease; mixed/limited evidence for chemotherapy benefit prediction in N1. Prosigna (PAM50): provides risk of distant recurrence at ten years in postmenopausal women with N0 or N+ HR+ breast cancer; not intended for N2. MammaPrint and BluePrint: classify high vs low risk; trials (MINDACT, RASTER, NBREaST II, DETERMIND) show prognostic information and influence on adjuvant/neoadjuvant decisions; mixed results for EET prediction. EndoPredict and BCI: prognostic for late recurrence and selection for extended endocrine therapy (BCI), but evidence levels vary. Systematic reviews indicate varying LOE across assays; only Oncotype DX and MammaPrint had large prospective trials. Guidelines (ASCO, NCCN, ESMO, NICE) provide conditional recommendations aligning with these assay-specific findings.

Definitions

Term	Definition
Comparative Genome Hybridization (CGH)
CGH is a technology that can be used to detect genomic copy number variations (CNVs). Tests use labeled tumor and reference DNA hybridized on arrays and differences in probe intensity are measured by scanner.
Comprehensive Genomic Profiling (CGP)
A type of next-generation sequencing test able to detect all classes of genomic alterations, including cancer biomarkers, from a single sample.
Gene Expression Profiling (GEP)
A laboratory test that analyzes messenger RNA (mRNA) patterns to determine gene activity; also referred to as gene expression testing, gene expression classifier testing, or gene expression assay.
Liquid Biopsy
Testing on bodily fluid to identify tumor-derived cells or circulating tumor DNA/RNA used for detection, treatment selection, or monitoring.
Measurable Residual Disease (MRD)
Very small number of cancer cells or cell contents detectable during/after treatment, often using PCR, flow cytometry, or NGS; indicates residual disease risk.
NGS
Next Generation Sequencing: techniques that can quickly analyze multiple sections of DNA simultaneously.
Risk group definitions (Prostate)
Very Low/Low/Favorable Intermediate/Unfavorable Intermediate/High/Very High risk prostate cancer as defined per NCCN v1.2025 clinical and pathological criteria.
Multigene analysis
For this policy, generally refers to a gene panel containing five or more genes; may include DNA or RNA analyses and algorithmic classifiers.
GEO/GEA
Gene expression assay — tests that measure expression patterns of defined genes to provide prognostic or predictive signatures.
DRFI
Distant recurrence-free interval.
DMFS
Distant metastasis-free survival.
BCS
Breast-conserving surgery.
GEP / GEA / GEC / GC
Gene expression profiling / gene expression assay / gene expression classifier / genomic classifier — assays that measure expression of multiple genes to provide prognostic or predictive information.
ctDNA / cfDNA
Circulating tumor DNA / cell-free DNA detected in plasma used for liquid biopsy molecular profiling.
GEC
Genomic expression classifier or genomic classifier (test that analyzes gene expression to predict cancer behavior).
GPS
Genomic Prostate Score (Oncotype DX Prostate), a 17-gene assay to predict adverse pathology and guide risk stratification.
EPI
ExoDx Prostate IntelliScore, a urine exosome gene expression assay to assess risk of high-grade prostate cancer.
Bethesda categories
Cytology classification system for thyroid FNA results (e.g., III = AUS/FLUS; IV = FN/SFN; V = Suspicious for malignancy; VI = Malignant).
NPV/PPV
Negative predictive value / Positive predictive value — test performance metrics influenced by disease prevalence.
CGP
Comprehensive genomic profiling — broad NGS-based panel to identify genomic alterations and actionable targets.
MCD
Multi-cancer detection test — assays designed to detect signals from multiple cancer types via biomarkers in body fluids (e.g., cfDNA methylation).
TTMV-HPV DNA / ctHPVDNA
Circulating tumor DNA or tumor tissue-modified viral HPV DNA measured in plasma for detection or surveillance of HPV-related cancer.
MRD
Molecular residual disease — low levels of tumor-derived DNA detected after definitive therapy indicating residual disease risk.
WGTA
Whole genome and transcriptome analysis combining DNA and RNA sequencing for tumor profiling.
Liquid biopsy / ctDNA
Cell-free DNA testing (circulating tumor DNA) for detection of tumor-derived genomic material in blood; referenced as emerging but with uncertain sensitivity, false-positive rates and predictive value for early-stage disease.

Selected Evidence & References

Selected studies, systematic reviews, and HTA cited in policy:

• TAILORx (Sparano et al./TAILORx updates): 21‑gene Oncotype DX—large prospective trial demonstrating prognostic power and informing chemo benefit analyses.

• MINDACT (Piccart et al.): randomized phase 3 demonstrating MammaPrint can identify genomically low‑risk individuals among clinically high‑risk women; ~46% of high clinical risk women might avoid chemo.

• Hyams et al. 2024 systematic review: high‑level evidence supporting prognostic capability of five common breast GEAs; Oncotype DX strongest for chemo benefit.

• Hayes assessments (multiple, 2020–2024): HTA‑style reviews covering Oncotype DX, EndoPredict, BCI, ThyroSeq v3, Afirma GSC/XA, Decipher, Prolaris, Signatera, NavDx, Epi proColon/ColoHealth, PancraGEN/PancreaSeq and others — noting evidence strengths, gaps, and rating clinical utility.

• PATHFINDER and SYMPLIFY (GRAIL/MCED studies): prospective feasibility studies of multi‑cancer detection assays reporting low cancer‑signal positive rates and variable positive predictive values; feasibility but insufficient evidence for routine screening.

• Decipher systematic reviews and ancillary RCT analyses (Feng et al., Jairath et al.): associations with metastasis, PCSM and OS across prostate disease states but limited prospective outcome evidence.

• Pancreatic cyst NGS/PancreaSeq and targeted NGS studies (Paniccia, Singhi): promising sensitivity/specificity for mucinous cysts/advanced neoplasia but further prospective validation required.

Not Covered / Investigational / Unproven Tests

Unproven / Not Medically Necessary - named tests (comprehensive list)

Consolidated list of tests and categories considered unproven or not medically necessary across multiple tumor types; each named test is a separate leaf (any operator semantics).

Unproven / Not Medically Necessary - named tests (any): NGS panels of > 50 genes or CGP unless otherwise specified; Afirma Xpression Atlas (XA); CancerTYPE ID; blood based colorectal cancer screening tests (ColoHealth, Signal-C, Guardant Shield); Decipher Bladder; DecisionDx-Melanoma; DiffDx-Melanoma; DecisionDx-SCC; DermTech PLA; Merlin (SkylineDx); myPath Melanoma; ExoDX Prostate Test; MyProstateScore; MyProstateScore 2.0; Confirmmdx; Selectmdx; TMPRSS2 fusion gene assays; Measurable Residual Disease (MRD) assays in solid tumor cancers (Invitae Personalized Cancer Monitoring, Signatera, RaDaR, Guardant Reveal, Haystack MRD); Multi-cancer early detection/screening tests (e.g., Galleri); NavDx; Oncotype DX Colon Recurrence Score; Genefx Colon (ColDx); ColonSentry; PancraGEN; PancreaSeq Genomic Classifier; Percepta GSC; Solid tumor profiling including WES/WGS/whole transcriptome (e.g., CancerVision, Tempus xE/xR, OncoExTra); Tempus Immune Profile Score (IPS); Whole genome methylation profiling.

Each listed test is considered unproven/not medically necessary unless otherwise specified by policy.

Policy stance: Tests and uses listed as unproven in the policy are considered investigational or not medically necessary because current evidence is insufficient to demonstrate that their routine use improves clinical management or patient outcomes. Where evidence is limited but evolving, the policy permits consideration of testing within clinical trials or per specific guideline/contractual exceptions; prior authorization and documented justification that test results will change management are often required for coverage decisions in Pennsylvania.

Overall Evidence‑Based Coverage Determinations

Overall Evidence‑Based Coverage Determinations (implied): The policy uses an evidence‑tier framework—tests with established clinical utility and guideline support (e.g., select breast GEPs in defined ER+/HER2‑ early breast cancer contexts; tissue or plasma multigene panels in NSCLC when clinically indicated; select prostate GEPs for initial risk stratification or post‑prostatectomy management) are covered in specified, validated clinical contexts. Other assays lacking prospective, high‑quality evidence of clinical utility (including many MRD ctDNA assays, multi‑cancer detection blood tests, broad unvalidated CGP/WES/WGS/WGTA for routine use, and numerous named proprietary tests) are considered investigational and not routinely covered; use may be limited to clinical trial settings or exceptional cases with prior authorization and supporting documentation that results will change management.

Revision History

2026-02-01material changeLatest

Applicable CPT/proprietary laboratory codes were appended to the policy's applicable codes list per 02/01/2026 revision. Updated list of applicable CPT codes to reflect quarterly edits; added 0578U, 0585U, 0586U, 0592U, and 0597U.

2026-02-01non-material changeLatest

Supporting information section was updated and the prior policy version CS152PA.AB archived. Updated Supporting Information and archived previous policy version CS152PA.AB.

Policy Summary

PayerUnitedHealthcare

PolicyMolecular Oncology Testing for Solid Tumor Cancer Diagnosis, Prognosis, and Treatment Decisions (for Pennsylvania Only)

Policy CodePolicy CS152PA.AC

Change TypeCPT/proprietary code additions; supporting info updated

Effective DateFeb 1, 2026

Next Review Date

Key ActionUse gene expression and molecular assays only in indications supported by the policy and document clinical factors and how results will influence management.

On This Page

HER2 receptor negative.

Adjuvant chemotherapy is not precluded due to any other factor (e.g., advanced age and/or significant comorbidities).

Currently receiving adjuvant hormonal therapy

Hormone receptor-positive (estrogen receptor positive, progesterone receptor positive, or both).
HER2 receptor negative.
Individual and treating physician have had a discussion prior to testing regarding the potential results of the test and determined to use the results to guide a decision regarding extended adjuvant hormonal therapy.

Use of more than one predictive GEP for the same tumor in an individual with breast cancer is unproven and not medically necessary due to insufficient evidence of efficacy; exception: BCI may be used once to evaluate extended endocrine therapy even if another GEP was previously performed to determine adjuvant chemotherapy.

GEP for other indications (including ductal carcinoma in situ - DCIS) or for treatment decisions beyond the indications above is unproven and not medically necessary.

Breast Cancer GEP - Not Medically Necessary / Unproven

The following tests or uses are unproven and not medically necessary for breast cancer: BluePrint; DCISionRT; Oncotype DX Breast DCIS Score; other GEP for DCIS or other non‑specified indications.

Lung Cancer - Medically Necessary

Multigene molecular profiling (including panels of up to 50 genes, or >50 genes only when consistent with FDA Cleared or Approved Companion Diagnostic policy) performed using tumor tissue or via liquid biopsy (cfDNA/ctDNA) is proven and medically necessary for non-small cell lung cancer when used to identify biomarkers for therapy selection.

Prostate Cancer GEP - Medically Necessary (GPS, Prolaris, Decipher)

Biopsy-proven, untreated, localized adenocarcinoma

Genomic Prostate Score (GPS)
- Test is ordered by a physician specializing in organ-confined prostate cancer (surgical oncology/urology, radiation oncology, or medical oncology).
- Results will be used to assist with treatment decision-making when the individual has not yet received treatment and is a candidate for either active surveillance or definitive therapy.
- Life expectancy is greater than 10 years.
- Risk group: Very Low-Risk, Low-Risk, or Favorable Intermediate-Risk Prostate Cancer (per definitions).
Prolaris Biopsy or Decipher Prostate Biopsy genomic classifier
- Test is ordered by a physician specializing in organ-confined prostate cancer (surgical oncology/urology, radiation oncology, or medical oncology).
- Results will be used to assist with treatment decision-making when the individual has not yet received treatment and is a candidate for either active surveillance or definitive therapy.
- Life expectancy is greater than 10 years.
- Risk group: Very Low-Risk, Low-Risk, Favorable Intermediate-Risk, Unfavorable Intermediate-Risk, or High-Risk Prostate Cancer (per definitions).

Use of tissue-based genomic biomarkers (GPS, Prolaris, Decipher) is intended to inform selection between active surveillance and definitive therapy; ordering clinician and intended decision context must be documented in the medical record.

Molecular screening panel tests for prostate cancer (e.g., ExoDx Prostate Test, My Prostate Score, ConfirmMDx, SelectMDx) are unproven and not medically necessary.

Decipher Prostate RP - Medically Necessary (Post-prostatectomy)

ANY of the following

Decipher Prostate RP genomic classifier to inform adjuvant treatment after radical prostatectomy when adverse pathologic features are found (e.g., high-grade disease, Gleason score ≥8, extracapsular extension, positive surgical margins, seminal vesicle invasion).
Decipher Prostate RP genomic classifier to inform adjuvant treatment after radical prostatectomy when PSA is greater than zero at any point following prostatectomy.

Results should be used in conjunction with clinicopathologic features to guide decisions regarding adjuvant radiation or systemic therapy; ordering clinician should document intended impact on management.

Prostate - Not Medically Necessary

Molecular screening panel tests for prostate cancer (e.g., ExoDx Prostate Test, MyProstateScore, MyProstateScore 2.0, ConfirmMDx, SelectMDx, TMPRSS2 fusion gene) are unproven and not medically necessary due to insufficient evidence of efficacy.

Thyroid Cancer / Indeterminate Thyroid Nodule Testing - Medically Necessary

ALL of the following

Fine needle aspiration cytology is indeterminate for follicular-patterned lesion (Bethesda III or IV).
Results of the molecular test will be used to make decisions about further surgery (e.g., to proceed with or avoid diagnostic lobectomy/thyroidectomy).
Tests considered proven and medically necessary include Afirma GSC, ThyroSeq V3, ThyGeNEXT/ThyraMIR for this indication.

Comprehensive genomic profiling (CGP) of confirmed anaplastic thyroid cancer is proven and medically necessary.

Thyroid - Not Medically Necessary / Unproven

Other molecular tests for indeterminate thyroid nodules or thyroid cancer are unproven and not medically necessary, including Afirma Xpression Atlas (XA), CGP (e.g., NeoTYPE Thyroid Profile), and use of more than one molecular profile test for the same nodule.

Uveal Melanoma GEP - Medically Necessary

ALL of the following

Individual has primary, localized uveal melanoma.
There is no evidence of metastatic disease.
Individual has not previously had DecisionDx-UM testing for the current diagnosis.

GEP (e.g., DecisionDx-UM) may be used to assist with predicting disease severity and making treatment decisions in this context.

Unproven Molecular Tests - Not Medically Necessary

Any of the following

NGS panels of > 50 genes or CGP unless otherwise specified.
Afirma Xpression Atlas (XA).
CancerTYPE ID.
Measurable Residual Disease (MRD) assays in solid tumors (tumor-informed or tumor‑naïve) (e.g., Invitae Personalized Cancer Monitoring, Signatera, RaDaR, Guardant Reveal, Haystack MRD).
Multi-cancer early detection/screening tests (e.g., Galleri).
NavDx.
Oncotype DX Colon Recurrence Score, GeneFx Colon, ColonSentry.
PancraGEN, PancreaSeq Genomic Classifier.
Percepta GSC.
DecisionDx-Melanoma, DecisionDx-SCC, DiffDx-Melanoma, DermTech PLA, Merlin (SkylineDx), myPath Melanoma.
Decipher Bladder.
Blood-based colorectal cancer screening tests (e.g., ColoHealth, Signal-C, Guardant Shield).
Solid tumor profiling that includes Whole Exome, Whole Genome, or whole transcriptome sequencing (e.g., CancerVision Inocras, Tempus xE/xR, OncoExTra).
Tempus Immune Profile Score (IPS).
Whole genome methylation profiling.

These tests are considered unproven and not medically necessary due to insufficient evidence of efficacy for the listed indications.

Applicable Codes and Test Definitions; Evidence and Use Limitations

Applicable Codes and Test Definitions; Test Use Contexts and Limitations; Evidence-supported uses and limitations; Coverage-relevant evidence summary; Overall Evidence-Based Coverage Determinations

ALL of the following

Applicable codes: include CPT/HCPCS codes appropriate to the specific assay; providers must bill using the specific test code and document the test name and intended clinical use in the medical record.
Test definitions: GEP = mRNA-based gene expression assays; CGP = next-generation sequencing detecting multiple classes of genomic alterations; Liquid biopsy = cfDNA/ctDNA assays used per clinical context.
Test use contexts and limitations: assays are intended to answer specific clinical questions (e.g., need for adjuvant chemotherapy, decision about extended endocrine therapy, selection between active surveillance vs definitive therapy, post‑prostatectomy adjuvant decision-making). Tests should not be used for screening, general prognosis outside specified contexts, or when results will not influence management.
Use of tissue-based genomic biomarkers for localized prostate cancer: GPS, Prolaris, Decipher are covered when ordered by appropriate specialists, for untreated localized adenocarcinoma, documented life expectancy >10 years, and when results will inform choice between active surveillance and definitive therapy; documentation of risk group and intended management impact required.
Role of genomic classifiers in active surveillance and treatment selection: evidence indicates these assays can provide incremental prognostic information beyond clinical and pathologic features to assist shared decision-making; they should not replace established clinical criteria but complement them.
Use of biomarkers to avoid or guide biopsy: molecular assays for the purpose of avoiding diagnostic procedures (e.g., multi-cancer screening or tests not specified above) are not covered unless evidence and indication are separately listed as proven.
Use of genomic classifiers in post-prostatectomy/radiation recurrent settings: Decipher Prostate RP is covered to inform adjuvant therapy after radical prostatectomy when adverse pathologic features or persistent/biochemical recurrence (PSA >0) are present; other genomic tests in the recurrent setting are unproven unless explicitly specified.
Evidence and guideline-based positions summarized: coverage aligns with guideline recommendations supporting select GEP/CGP assays for defined indications (breast GEP for node-negative/1-3 node positive HR+/HER2- disease; lung multigene testing for NSCLC; thyroid indeterminate nodules Bethesda III/IV; prostate genomic classifiers for localized disease management and Decipher RP post‑prostatectomy).
Coverage-relevant evidence summary: randomized trials and prospective validation studies exist for certain assays (e.g., Oncotype Dx, MammaPrint, Prosigna, BCI) demonstrating prognostic and predictive utility in specified populations; for prostate classifiers, validation studies show prognostic discrimination and utility in risk stratification though impact on long-term outcomes varies; many other assays and broad CGP or MRD applications lack sufficient evidence for clinical utility and are considered investigational.
Operational notes: prior authorization may be required; providers must document clinical criteria, discussion of results use, and treating specialist ordering the test; multiple GEP tests for the same lesion are not covered except where explicitly allowed (BCI once for extended endocrine therapy).

Applicable tests: Breast Cancer Index (BCI) — allowed for evaluation of extended endocrine therapy

Prior Authorization

Companion diagnostic / FDA-cleared exceptions

Companion diagnostics and FDA-cleared/approved assays are treated differently. Multigene panels >50 genes are allowed only when used in a manner consistent with the payer's FDA Companion Diagnostic policy; otherwise >50-gene panels or CGP are considered unproven for many indications. Refer to the separate UnitedHealthcare policy on FDA Cleared or Approved Companion Diagnostic Testing for Pennsylvania for exceptions.

Panels >50 genes: allowed only per FDA CD policy or when specified in this policy

Documentation Required

Document receptor status, staging, intended decision & assay-specific populations

Document tumor and patient features required to support use of prognostic or predictive assays. Record receptor status (ER/PR/HER2), nodal stage, menopausal status (if relevant), clinical nodal and metastatic stage, and the specific intended clinical decision the test result will influence (e.g., adjuvant chemotherapy, extended endocrine therapy, local therapy selection). Recognize that each assay has an intended population and analytic claims described by the test manufacturer — order assays only for the population for which the assay is validated.

Impacted codes: CPT proprietary codes added for tests (see billing rules); include test name and specimen source in documentation
Document: receptor status, nodal stage, menopausal status, intended clinical decision

Denial Risk

Routine ordering not supported

Ordering genomic tests routinely or reflexively without documentation of clinical indication or discussion of how results will affect management may be denied. Prior authorization may be implied by policy context and failure to follow documentation or prior authorization requirements increases denial risk.

Routine/reflex ordering without documented indication or counseling may be denied
Prior authorization: follow payer operational requirements

Billing Rule

Applicable CPT code updates (effective as of revision)

CPT/proprietary codes were added effective with the current revision date. Use these codes for billing newly described proprietary assays and ensure the effective revision date is used for claims.

Added CPT/proprietary codes (effective with this revision): 0578U, 0585U, 0586U, 0592U, 0597U
Effective revision date: 2026-02-01

Documentation Required

Use assays for prognosis and treatment decision support with documented indication

Use assays that provide prognostic or treatment decision support only when there is a documented clinical indication that the result will influence management. Document how results influenced treatment planning, including shared decision-making with the patient.

Documented indication required; record impact on treatment plan and patient counseling
Follow guideline-supported assay selection (e.g., MammaPrint, Oncotype DX Breast, Prosigna, BCI, EndoPredict for indicated breast scenarios)

Documentation Required

Use validated assays and quality assurance; interpret in clinical context

Select validated assays and use quality-assured laboratories. Prefer tests and platforms supported by clinical guidelines and by analytic/clinical validation. Results must be interpreted in the clinical context and integrated into the shared decision-making process with the patient.

Use guideline-supported tests where available; ensure laboratory validation and quality programs are in place
Interpret results with clinical context and document shared decision-making

Documentation Required

Prefer panel-based testing for NSCLC

For non-small cell lung cancer (NSCLC), prefer panel-based NGS or CGP (up to 50 genes per policy rules) to comprehensively identify actionable genomic alterations. Limit larger panels (>50 genes) to situations consistent with the companion diagnostic/FDA CD policy or when specifically indicated and documented.

Prefer panel-based NGS/CGP for NSCLC (≤50 genes unless FDA CD exception applies)
Document intended clinical use (targeted therapy selection, clinical trial eligibility)

Documentation Required

Selective ordering of genomic assays and CUP guidance

Order genomic testing selectively — avoid routine use for indications without evidence of clinical utility. For CUP, consider NGS/CGP when results may identify actionable targets or clinical trial options; use molecular profiling judiciously and document the anticipated management impact.

CUP: consider NGS/CGP to identify actionable targets or trial eligibility; document expected influence on therapy
Avoid routine prognostic GEP for cutaneous melanoma; perform oncogenic mutation testing in primary cutaneous melanoma per guidelines; biopsy-based prognostic testing for uveal melanoma (e.g., DecisionDx-UM) allowed when criteria met

Documentation Required

Counseling and informed consent for molecular testing

Provide pre- and post-test counseling and informed consent as appropriate for molecular testing. Explain limitations, potential incidental findings, possible need for confirmatory testing, and implications for treatment, surveillance, or familial risk. Document counseling in the medical record.

Counseling/informed consent should cover test limitations, incidental findings, follow-up testing, and impact on management
Document pre-/post-test counseling and shared decision-making

Documentation Required

Melanoma: avoid routine prognostic GEP for cutaneous melanoma; uveal melanoma biopsy/GEP allowed

Avoid using multi-gene prognostic GEP tests for cutaneous melanoma routinely; these are considered unproven. For melanoma, perform oncogenic mutation testing for primary cutaneous melanoma when guideline-supported. For uveal melanoma, biopsy and prognostic GEP (e.g., DecisionDx-UM) are allowed for primary localized disease when criteria are met and no prior DecisionDx-UM testing for the current diagnosis has been done.

Do not routinely order DecisionDx-Melanoma, DiffDx-Melanoma, DermTech PLA, Merlin, myPath Melanoma for prognostic use
Oncogenic mutation testing and uveal melanoma GEP: follow policy-listed indications

Documentation Required

MRD/MCD and cfDNA screening guidance

Measurable residual disease (MRD) or minimal residual disease assays for solid tumors and multi-cancer early detection/screening tests are considered unproven and not medically necessary outside of validated research settings. cfDNA screening signals should prompt appropriate diagnostic follow-up rather than routine coverage for screening.

MRD assays (tumor-informed or tumor-naïve) and multi-cancer early detection tests (e.g., Galleri) are not covered for routine screening
Positive cfDNA screening results require diagnostic follow-up and are primarily investigational when used for population screening

Documentation Required

NGS for pancreatic cancer and limits on MCD/unproven tests

Limit use of comprehensive genomic profiling (CGP) and NGS to evidence-supported indications such as metastatic or locally advanced pancreatic cancer when guideline-supported. Use MCD tests and other unproven assays only in the context of clinical trials or when clear evidence supports clinical management change.

Consider NGS for metastatic/locally advanced pancreatic cancer when results will influence therapy or trial eligibility
MCD and other unproven tests: restrict to clinical trials or evidence-supported indications

Documentation Required

Interpretation caveats and documentation of test impact

Interpret genomic and GEP/ancillary test results with clinical context and document how results affected clinical decisions. Note potential limitations and the need for confirmatory testing when applicable. Reference guideline-supported tests and follow-up recommendations when implementing results into care plans.

Document how results influenced treatment, surveillance, or trial enrollment
Include interpretation caveats and plans for follow-up or confirmatory testing

Documentation Required

Policy use and modification notice

This policy sets payer-level expectations and may be modified. Apply federal, state, and contractual benefit terms when making coverage determinations. Operational requirements such as prior authorization, documentation, and billing changes may be enforced per payer procedures.

Policy use and modification notice: payer may change policy and operational requirements
Reference federal/state/contractual benefit terms for coverage decisions

Evidence‑supported uses and limitations (narrative criteria): Major professional guidelines (ASCO, NCCN, ESMO, NICE) recognize selective roles for some multigene and genomic assays but emphasize careful patient selection and limits on routine use. ASCO and NCCN support Oncotype DX, MammaPrint, Prosigna, EndoPredict and BCI in defined early ER+/HER2‑ breast cancer populations, with Oncotype DX having the strongest randomized trial data for predicting chemotherapy benefit (TAILORx/TAILORx updates). NICE and ESMO provide conditionally aligned recommendations with age/menopausal and nodal status caveats.

Key trial results: TAILORx and related analyses show the 21‑gene Oncotype DX Recurrence Score is prognostic for distant recurrence and overall survival and—with subgroup nuances—predictive of chemotherapy benefit in certain younger/premenopausal subgroups; MINDACT demonstrated that MammaPrint can identify clinically high‑risk but genomically low‑risk women who may safely omit chemotherapy in many cases (estimated ~46% of clinically high‑risk women might avoid chemo) though absolute DMFS differences were small and age‑dependent.

Overarching evidence conclusions: Oncotype DX has the strongest prospective evidence for predicting chemotherapy benefit, while MammaPrint and other GEAs offer high‑quality prognostic information and can change management decisions; many assays are prognostic but their prospective predictive utility for therapy decisions (or impact on hard clinical outcomes) is limited or inconsistent. For prostate cancer, genomic classifiers (GPS, Decipher, Prolaris) show prognostic associations and may inform active surveillance and post‑prostatectomy discussions (Decipher has supportive registry, meta‑analytic and ancillary RCT analyses), but routine use is recommended only when results would meaningfully change management. For thyroid indeterminate nodules, ThyroSeq v3 and Afirma GSC/ XA have high sensitivity/NPV in many studies but evidence and test‑specific limitations remain; for uveal and cutaneous melanoma, GEPs provide prognostic stratification but prospective evidence that routine use improves outcomes is limited and guidance recommends restrained use. Broad emerging areas—ctDNA/MRD, multi‑cancer early detection, whole exome/genome/transcriptome profiling—have promising analytic/associative data but insufficient prospective evidence of clinical utility; these are generally considered investigational or recommended only in clinical trials or selective contexts.

Actionable thresholds for providers

MyProstateScore actionable threshold10 (reported sensitivity 97%, NPV 98% to rule out GG ≥2 cancer)

EPI (ExoDx) predefined cutpoints15.6 (avoid ~26% biopsies, NPV ~89%, miss ~7% ≥GG2) and 20 (avoid ~40% biopsies, NPV ~89%, miss ~11% ≥GG2)

DCISionRT Decision Score (DS) categorizationDS 0–10 with 0–3 = low risk and 3–10 = elevated risk (used to stratify RT benefit)

cfDNA (ECLIPSE/Shield) CRC sensitivity/specificityOverall CRC sensitivity 83.1% (stage I-III 87.5%); specificity for advanced neoplasia ~89.6%; invalid rate 3.7%

cfDNA MCED performance (CCGA/validation)Stage I-III sensitivity ~67.3% for selected cancers; validation specificity ~99.3%; tissue-origin accuracy ~93% when cancer-like signal detected

Uveal melanoma tumor thickness + GEP thresholdsClass 2 tumors thickness ≥7.0 mm associate with higher metastasis risk; Class 1 tumors thickness ≥9.0 mm associated with increased risk