Magnetic Resonance Spectroscopy (MRS) Clinical Policy Bulletin

Coverage Summary

Scope: This policy defines Aetna's determinations of medically necessary and experimental/investigational (not medically necessary) uses of magnetic resonance spectroscopy (MRS), and lists covered and not-covered CPT and ICD-10 codes and guidance for documentation and use. Coverage stance: mixed (some indications are medically necessary; others are experimental/investigational). Background (brief): MRS is a non-invasive NMR technique used to evaluate metabolic changes in the brain and other organs; the MRS probe accessory has FDA 510(k) clearance, but the clinical role and impact on outcomes remain unestablished because of limited high-quality evidence.

Metadata: Effective 1998-03-05; Last review 2023-04-11; Next review 2024-02-08.

Key counts and evidence summary

Medically necessary indications listed9

Experimental/investigational examples listed29+

Covered CPT codes1 (76390)

CPT codes explicitly not covered4 (0609T, 0610T, 0611T, 0612T)

Evidence-quality noteNumerous small studies and meta-analyses show promising but mixed diagnostic performance; further standardized, multicenter outcomes research needed

Medical Necessity — Covered and Not Covered Criteria

Medically Necessary

Aetna considers magnetic resonance spectroscopy (MRS) medically necessary for the following indications:

ALL of the following

Assessing prognosis in hypoxic ischemic encephalopathy
Distinguishing low grade from high grade gliomas
Evaluate a brain lesion of indeterminate nature when the MRS findings will be used to determine whether biopsy/resection can be safely postponed
Distinguishing recurrent brain tumor from radiation-induced tumor necrosis
Diagnosis and monitoring of metabolic disorders (ONE of)

Coding

CPT codes covered for indications listed in the CPBCPTCovered

76390

Magnetic resonance spectroscopy

CPT codes not covered for indications listed in the CPBCPTNot Covered

0609T	Magnetic resonance spectroscopy, determination and localization of discogenic pain (cervical, thoracic, or lumbar); acquisition of single voxel data, per disc, on biomarkers (ie, lactic acid, carbohydrate, alanine, laal, propionic acid, proteoglycan, and collagen) in at least 3 discs
0610T	Transmission of biomarker data for software analysis
0611T	Postprocessing for algorithmic analysis of biomarker data for determination of relative chemical differences between discs
0612T	Interpretation and report

ICD-10 codes covered if selection criteria are metICD-10Covered

C71.0 - C71.9	Malignant neoplasm of brain [not covered for differentiating primary central nervous system lymphoma (PCNSL) from other focal brain lesions]
C78.31	Secondary malignant neoplasm of brain [not covered for differentiating primary central nervous system lymphoma (PCNSL) from other focal brain lesions]
D33.0 - D33.2	Benign neoplasm of brain [not covered for differentiating primary central nervous system lymphoma (PCNSL) from other focal brain lesions]
D43.0 - D43.2	Neoplasm of uncertain behavior of brain [not covered for differentiating primary central nervous system lymphoma (PCNSL) from other focal brain lesions]
D49.6	Neoplasm of unspecified behavior of brain [not covered for differentiating primary central nervous system lymphoma (PCNSL) from other focal brain lesions]
E71.0	Maple Syrup Urine disease
E71.2	Disorder of branched-chain amino-acid metabolism, unspecified [Creatine deficiency]
E71.520 - E71.529	X-linked adrenoleukodystrophy
E72.51	Nonketoic hyperglycinemia
E75.23	Krabbe disease

1–10 of 20

1/2

ICD-10 codes not covered for indications listed in the CPB (not all-inclusive)ICD-10Not Covered

A69.20 - A69.29	Lyme disease
C15.3 - C15.9	Malignant neoplasm of esophagus
C22.0 - C22.9	Malignant neoplasm of liver and intrahepatic bile ducts
C50.011 - C50.929	Malignant neoplasm of breast (male and female)
C61	Malignant neoplasm of prostate
C72.9	Malignant neoplasm of central nervous system, unspecified [Primary central nervous system (CNS) lymphoma]
C79.49	Secondary malignant neoplasm of spinal cord
C79.81	Secondary malignant neoplasm of breast
C79.82	Secondary malignant neoplasm of genital organs [prostate]
D05.00 - D05.92	Carcinoma in situ of breast

1–10 of 76

1/8

Provider Actions & Documentation Requirements

Documentation Required

Documentation for medically necessary indications

Documentation must support the specific diagnosis (one of the listed covered indications). For disease monitoring, recent MRI findings must be inconclusive and a change in therapy is being considered.

Background and Evidence Summary

MRS (also called NMR spectroscopy) is a non-invasive technique that produces spectra of metabolites rather than images and can be performed on commercially available MRI instruments; the probe accessory for MRS received FDA 510(k) clearance. The technique has been applied to brain tumors, stroke, seizure disorders, neurodegenerative and psychiatric conditions, and to study other organs. Although many studies demonstrate technical feasibility and suggest potential utility (for example, meta-analyses showing moderate diagnostic performance for certain brain tumor distinctions), there is a paucity of high-quality, standardized clinical trials demonstrating that MRS measurably improves diagnostic thinking, therapeutic decision-making, or patient outcomes. Multiple systematic reviews and assessments (AHRQ/Tufts, BCBSA TEC, CMS) have concluded that evidence is insufficient or limited by methodological shortcomings, heterogeneity, small sample sizes, and lack of standardization, and therefore MRS remains investigational for many indications.

Study / Source	Key finding
Zhang et al (2014) meta-analysis
Cho/Cr pooled sensitivity 0.83 and specificity 0.83 (AUC 0.9001); Cho/NAA sensitivity 0.88 and specificity 0.86 (AUC 0.9185) for differentiating recurrent glioma vs radiation necrosis
Chuang et al (2016) meta-analysis
Cho/Cr and Cho/NAA ratios and rCBV higher in tumor recurrence vs radiation injury; pooled differences significant though heterogeneity observed across studies
CMS (2004) decision memo / national non-coverage determination
Determined insufficient evidence to deem MRS reasonable and necessary for brain tumor diagnosis (national non-coverage at that time) due to methodological shortcomings
Gornet et al (2019) lumbar discs vs provocative discography
MRS total accuracy 85%, sensitivity 82%, specificity 88% vs discography; non‑herniated discs accuracy up to 93%; surgical success higher when treating MRS-positive discs (97% vs 57%)
Baltzer & Dietzel (2013) breast MRS meta-analysis
Pooled sensitivity 73%, specificity 88%, AUC 0.88; notable publication bias and between-study heterogeneity
Umbehr et al (2009) prostate MRI/MRS meta-analysis
Combined MRI/MRS pooled sensitivity ~68% and specificity ~85% (subpart level); may help rule-in in selected low-risk patients but limited small studies
Wang et al (2014) brain tumor meta-analysis
Pooled sensitivity 80.05%, specificity 78.46%, AUC 0.78; authors noted moderate study quality and technique-dependent performance
Overall evidence limitations / heterogeneity
Numerous meta-analyses and systematic reviews show moderate diagnostic performance for select indications but note small samples, methodological heterogeneity, publication bias, and need for standardized, multicenter outcome studies

Evidence synthesis: MRS shows moderate diagnostic performance for some specific indications — for example, meta-analyses report pooled sensitivities and specificities for metabolite ratios (Cho/Cr and Cho/NAA) in differentiating recurrent glioma from radiation necrosis (pooled SEN/SPE ~0.83/0.83 for Cho/Cr and ~0.88/0.86 for Cho/NAA; AUCs ~0.90–0.92), and perfusion/spectroscopy combinations increase diagnostic accuracy in some studies. However, the overall body of evidence is limited by small, heterogeneous studies, methodological variability, possible publication bias, and few prospective multicenter trials; therefore further well-designed studies are needed to define MRS' clinical role and effect on outcomes.

Definitions

Glossary:

MRS — Magnetic resonance spectroscopy: Magnetic resonance spectroscopy, also known as nuclear magnetic resonance (NMR) spectroscopy; a non‑invasive technique producing spectra of metabolites rather than images.

MRS — Magnetic Resonance Spectroscopy: Noninvasive MR technique measuring biochemical/metabolic markers in tissue (eg, NAA, choline, creatine, myo‑inositol, lactate).

Pfirrmann grade — MRI-based grading system for intervertebral disc degeneration.

NAA — N‑acetylaspartate, a neuronal marker measured by MRS.

Cho — Choline-containing compounds, marker of membrane turnover.

Cr — Creatine, often used as an internal reference in MRS.

Glx — Glutamine plus glutamate peak.

Lac — Lactate.

1H-MRS — Proton magnetic resonance spectroscopy.

Medicare Determinations

Effective	Decision / Name
2004
CMS national non-coverage determination for MRS (brain tumor diagnosis) (NCD) — determined insufficient evidence to deem MRS reasonable and necessary for brain tumor diagnosis
2004-01-29
Decision memo for magnetic resonance spectroscopy for brain tumors — CAG-00141N (CMS)

Revision History

1998-03-05effective

Policy effective

2023-04-11last_review

Last review

2024-02-08next_reviewLatest

Next review scheduled