This policy applies to molecular oncology tests for solid tumors that are not FDA-cleared or -approved as companion diagnostics. Tests that meet the FDA companion diagnostic designation fall outside the scope of this policy and should be evaluated under the separate FDA Cleared or Approved Companion Diagnostic Testing policy. Tests identified in the policy as “Unproven Molecular Tests” are considered not supported by sufficient evidence and are subject to noncoverage or denial consistent with the policy position on unproven testing.

For reference and adjudication purposes the policy lists examples of tests designated as “Unproven Molecular Tests.” Examples include, but are not limited to: Tempus Immune Profile Score, whole-genome methylation profiling, EpiSwitch Checkpoint Inhibitor Response Test, CancerTYPE ID, Bladder EpiCheck, Cxbladder Triage/Monitor, Decipher Bladder Genomic Classifier, DecisionDx-Melanoma, DecisionDx-SCC (40-GEP), DermTech PLA, Merlin, MyPath Melanoma, Confirm mdx, ExoDx Prostate, MyProstateScore 2.0, Select mdx, Guardant Reveal, Signatera, Galleri, and other named assays enumerated in the Unproven Molecular Tests list.

Cancer screening tests that are covered under a member’s preventive care benefit are addressed separately in the plan’s Preventive Care Services policy. Where a molecular oncology assay is intended for routine population screening and that indication is included in the Preventive Care Services policy, follow that policy for coverage determination rather than this molecular oncology testing policy.

Evidence limitations create gaps in applicability of some gene expression assays (GEAs) across patient subpopulations. Several GEAs have strong prognostic data overall but remain unvalidated or of uncertain predictive value in defined subgroups (for example, by menopausal status or nodal category). Published analyses and assessments note limited prospective data, enrichment for higher-risk trial cohorts, and insufficient follow-up for late recurrences that constrain generalizability of findings to all patient populations.

Key study limitations cited across several GEA evaluations include cohort enrichment for higher-risk participants, frequent concomitant chemotherapy use that complicates assessment of assay-predicted treatment benefit, and relatively short or incomplete follow-up for detection of late distant recurrences. These design limitations and potential sponsor conflicts of interest were repeatedly noted as reasons why prospective, randomized validation is still needed for certain clinical claims.

Gene expression profiling assays developed for ductal carcinoma in situ (DCIS) and other breast cancer assays beyond the principal, well-studied GEAs currently lack sufficient evidence of clinical utility. Specifically, assessments (Hayes and other systematic reviews) conclude that DCISionRT and the Oncotype DX Breast DCIS Score have low-quality or insufficient evidence to support routine clinical implementation for guiding radiotherapy or other definitive management decisions.

Clinical guideline panels do not recommend use of GEAs to guide management in patients with HER2-positive or triple-negative breast cancer; GEAs are primarily relevant to prognostication and adjuvant systemic therapy selection in hormone receptor–positive, HER2-negative disease. NCCN specifically notes GEAs are not required for staging and does not recommend them for HER2+ or TNBC populations.

Plasma circulating tumor DNA (ctDNA) testing has a defined role in the advanced/metastatic setting or when tissue biopsy is unsafe or insufficient. The NCCN and the policy make clear that ctDNA should not be used in lieu of a tissue diagnosis in non‑metastatic disease; when ctDNA is performed because tissue is unavailable, plans should document the clinical rationale and pursue tissue testing or follow-up as clinically indicated if no actionable driver is identified on plasma testing.

Several commercially marketed assays for prostate cancer detection or risk stratification (for example, MyProstateScore 2.0, OncoAssure, Confirm mdx) have been judged to lack sufficient peer‑reviewed evidence of clinical validity or utility. Hayes assessments and other evidence reviews concluded that current data are inadequate to support routine clinical use of these tests outside specified evidence‑based indications.

In prostate cancer evaluation, guideline panels (NCCN) state biomarker tests should not replace a multiparametric MRI when MRI is available and of adequate quality. Biomarker testing may be considered in select circumstances when MRI is unavailable, contraindicated, or of low quality, but should not be used as a routine substitute for mpMRI.

Molecular testing is not recommended routinely to guide postoperative radioiodine therapy in differentiated thyroid cancer. For indeterminate thyroid cytology and unresectable or anaplastic thyroid cancers, molecular testing may inform management, but routine molecular testing to determine postoperative radioiodine decisions lacks supporting outcome evidence.

Tests supported only by low- or very-low–quality evidence, such as MyPath Melanoma per Hayes assessment, are not supported for routine clinical use as diagnostic adjuncts because available studies are insufficient to demonstrate improved patient outcomes.

Predictive gene expression profiling should not replace surgical or pathological staging discussions (for example, sentinel lymph node biopsy) in routine practice. Major guideline statements caution against using GEPs as a substitute for established surgical staging procedures outside of clinical trials; molecular test results may be used as adjunctive information but should not be the sole basis to forego recommended staging procedures.

Across many test evaluations, the evidence base is limited by retrospective or single‑center study designs, potential selection bias, and frequent conflicts of interest or manufacturer sponsorship. The policy notes these limitations and highlights the need for larger, multicenter prospective trials with independent assessment to establish clinical utility and generalizability.

Assays intended to infer the primary site in cancers of unknown primary (for example, gene expression classifiers such as CancerTYPE ID) currently lack sufficient peer‑reviewed evidence demonstrating improved clinical outcomes or reliable clinical effectiveness for guiding site‑specific therapy, and therefore are considered unsupported for routine clinical use pending higher‑quality validation.

The policy’s default position is that molecular oncology tests for solid tumors that are not specifically listed as proven and medically necessary in this document are considered unproven and therefore not medically necessary. Tests named in the Unproven Molecular Tests list should be treated as lacking coverage support unless later evidence or policy revision documents a proven indication.

Named tests included in the Unproven Molecular Tests lists are examples of assays that the policy implies are currently without coverage support pending stronger evidence of clinical validity and demonstrable clinical utility. Providers should assume these named assays are not supported for routine coverage unless an applicable medical necessity criterion elsewhere in the policy or a separate FDA companion-diagnostic designation applies.

Some chunks summarize evidence without making explicit ‘‘not medically necessary’’ statements; the document emphasizes uncertainty in predictive utility for many assays except for Oncotype DX, which has the strongest predictive data for chemotherapy benefit among the breast GEAs assessed. Policy language and evidence summaries should be interpreted with that context in mind.

Hayes assessments highlighted limitations for Oncotype DX in certain settings, for example the lack of sufficient evidence to support its use to estimate distant recurrence risk and chemotherapy benefit in patients with N2 disease (four to nine positive lymph nodes). These assessments inform restricted use in more heavily node‑positive populations.

Hayes assessments for EndoPredict and Breast Cancer Index (BCI) concluded that although each assay demonstrates prognostic signal for distant recurrence, evidence is insufficient to reliably predict benefit of extended endocrine therapy or to estimate long-term distant recurrence risk in all patient subgroups without further prospective validation. The policy therefore treats certain predictive indications for these assays as uncertain pending higher‑quality data.

Hayes reviews concluded there is inadequate or low‑quality evidence to support routine clinical use of DCISionRT and that the evidence supporting the Oncotype DX DCIS score is low quality; outcome data demonstrating benefit from score‑directed management are lacking, and the Hayes updates indicate current ratings are unlikely to change absent new high‑quality studies.

No top-level eligibility nodes are specified in the extracted eligibility chunks. Where eligibility criteria are defined elsewhere in the policy (for example, age limits for CRC screening with Shield or clinical scenario requirements for breast GEPs), those specific criteria govern coverage decisions.

This block is a placeholder. The policy’s eligibility sections contain specific, test- and indication-based requirements elsewhere (e.g., age and risk criteria for Shield CRC screening; clinical criteria for breast GEPs).

Fallback placeholder: where the inventory contains no explicit eligibility nodes for a given subsection, providers should rely on the test-specific coverage criteria and the member’s benefit document when determining coverage eligibility.

Test-specific eligibility placeholder for thyroid nodules: molecular testing is considered when fine-needle aspiration cytology is indeterminate (Bethesda III/IV) and the test result will be used to inform the decision about further surgery.

This block is a placeholder where no top-level eligibility nodes are present in the inventory excerpt; consult the policy text for detailed, indication-specific eligibility criteria.

Placeholder: when eligibility is not specified in a subsection, clinical documentation supporting intended use and concordance with policy-listed indications will be required for consideration.

Placeholder: multigene analysis is generally defined as a panel of five or more genes, and documentation should demonstrate that the assay meets the policy’s intended-use population and analytic scope.

Placeholder: for ctDNA tests used when tissue is unavailable, documentation should describe why tissue biopsy was not feasible and how ctDNA results will influence management.

Placeholder: where timing thresholds apply (for example, some breast GEPs are intended for newly diagnosed patients), providers should document recency of diagnosis per the policy criteria.

Placeholder: eligibility confirmation should reference the detailed clinical criteria in the policy sections associated with each covered test and indication.

Placeholder: when more than one molecular test is considered for a single thyroid nodule, the policy notes that performing multiple tests is unproven and not medically necessary.

Placeholder: where the inventory is empty for an eligibility subsection, standard documentation and justification aligning with the policy’s definitions and covered indications are expected.

Placeholder: providers should attach clinical notes showing how test results would influence management when eligibility is not explicitly enumerated for a given assay in the inventory excerpt.

Placeholder: where no explicit eligibility nodes exist, clinical justification consistent with guideline-supported indications should be provided for review.

Placeholder: documentation of the multigene panel composition (number of genes, gene classes evaluated) may be requested when eligibility is uncertain.

Placeholder: if an assay or indication is not listed, the test will be evaluated against the policy’s evidence framework and definitions for multigene analyses and intended clinical use.

All molecular oncology tests for solid tumors not specifically listed as proven/covered in the policy are considered unproven and not medically necessary due to insufficient evidence of clinical efficacy.

The policy explicitly identifies a set of tests labeled “Unproven Molecular Tests”. Examples listed as not covered include many named assays across tumor types (see the policy’s unproven tests list), such as multicancer detection assays, specific urine or plasma biomarker tests, and certain GEPs that lack sufficient validation.

Predictive evidence is insufficient to support some assays for predictive use in node-positive disease (for example, Prosigna and EndoPredict predictive data are limited in node-positive populations). These insufficiencies may result in noncoverage for the predictive indication absent stronger, prospective evidence.

Hayes-based evaluations cited in the policy indicate that Oncotype DX is not supported for N2 disease (four to nine positive lymph nodes) because available evidence does not sufficiently establish benefit for this population.

Hayes assessments found limited prospective utility data for EndoPredict and identified gaps in evidence for BCI’s predictive claims, which may justify classifying certain uses of these assays as noncovered where utility is unproven.

Other gene expression profiling assays for breast cancer or DCIS beyond those described in the coverage sections (for example, DCISionRT and Oncotype DX DCIS Score) lack sufficient evidence of clinical utility and may be excluded from coverage.

Hayes concluded that the evidence supporting DCISionRT is insufficient and that the Oncotype DX DCIS evidence body is low quality; these findings support noncoverage for routine clinical application for DCIS management pending stronger evidence.

Use of ctDNA in place of tissue diagnosis in non-advanced or non-metastatic disease is not recommended and may be considered noncovered when used as a substitute for tissue-based diagnosis rather than as an adjunct or when tissue is unobtainable.

For prostate-related biomarkers, tests that lack peer-reviewed evidence of clinical utility—such as MyProstateScore 2.0, OncoAssure, Confirm mdx—are not supported for routine clinical implementation and may be considered noncovered for those indications.

Select mdx was judged by Hayes to have a very low-quality evidence base and inadequate data to conclude effectiveness for predicting clinically significant prostate cancer; such findings support noncoverage for routine use pending stronger evidence.

Earlier assessments for thyroid panels (ThyGeNEXT/ThyraMIR) were considered inadequate to reclassify indeterminate nodules conclusively in some contexts; as a result, certain uses may be noncovered where evidence is insufficient to support management changes.

Routine molecular profiling to stratify progression risk for low-risk papillary thyroid carcinoma under active surveillance is not supported by the evidence and is not recommended for guiding enrollment or routine management decisions.

The Hayes assessment found the evidence for MyPath Melanoma insufficient to support diagnostic adjunct use; the test is therefore not supported for routine clinical use to change management.

Professional guideline groups and technology assessments discourage routine prognostic GEP testing for cutaneous melanoma to guide sentinel lymph node decisions or routine management outside of clinical trials; such routine use is not supported and may be noncovered.

Tests lacking demonstrated clinical utility or with very low-quality supporting evidence (per Hayes and other assessments) when used to change management beyond investigational or guideline-supported contexts may be excluded from coverage.

Molecular tests intended solely to infer tissue of origin for CUP (for example, CancerTYPE ID and similar GEP assays) have insufficient evidence of clinical effectiveness and are noted as likely noncovered absent convincing outcome data.

Molecular oncology testing encompasses a wide range of technologies used to evaluate DNA, RNA, methylation, and other tumor-derived analytes. The policy covers molecular profiling intended to inform diagnosis, prognosis, and treatment selection across solid tumors, but applies only to tests that are not FDA-cleared/approved as companion diagnostics.

Clinical evidence summaries and scope statements in the policy underscore that several gene expression profiling assays (Oncotype DX, Prosigna, EndoPredict, MammaPrint, BCI) have prognostic value in early HR-positive/HER2-negative breast cancer, but predictive evidence for chemotherapy benefit is strongest for Oncotype DX and remains limited for other assays.