A comprehensive literature search was conducted using Medical Subject Headings (MeSH) terms and related keywords, accounting for synonyms, acronyms, plurals, and spelling variations. The databases searched included PubMed, OVID, EMBASE, Web of Science, and the Cochrane Central Register of Controlled Trials, covering all records from their inception until June 2023. The search was limited to articles published in English or Spanish, with no time restrictions. An example of the search equation used in PubMed is provided below:

(“cervical vertebrae”[MeSH Terms] OR (“cervical”[All Fields] AND “vertebrae”[All Fields]) OR “cervical vertebrae”[All Fields] OR (“cervical”[All Fields] AND “spine”[All Fields]) OR “cervical spine”[All Fields]) AND (“cervical cord”[MeSH Terms] OR (“cervical”[All Fields] AND “cord”[All Fields]) OR “cervical cord”[All Fields]) AND (“magnetic resonance spectroscopy”[MeSH Terms] OR (“magnetic”[All Fields] AND “resonance”[All Fields] AND “spectroscopy”[All Fields]) OR “magnetic resonance spectroscopy”[All Fields]) AND (“spinal cord compression”[MeSH Terms] OR (“spinal”[All Fields] AND “cord”[All Fields] AND “compression”[All Fields]) OR “spinal cord compression”[All Fields])

Studies were considered eligible for inclusion if they met the following criteria:

Exclusion criteria were as follows:

All reviewers independently screened the titles and abstracts of the retrieved records. Relevant studies were then reviewed in full text to determine eligibility based on the criteria above. In cases of disagreement, a third reviewer was consulted to reach a consensus.

Data extraction was performed independently by all reviewers using separate Excel spreadsheets. Extracted data were then cross-checked among reviewers and against the original articles. The following information was collected:

Any discrepancies in data extraction were resolved through discussion, and if necessary, adjudicated by a third reviewer.

Each study was graded according to the 2011 Oxford Centre for Evidence-Based Medicine (OCEBM) guidelines. Two independent reviewers performed the grading, and any disagreements were resolved by a third reviewer.

We assessed heterogeneity across the included studies by examining clinical characteristics (e.g., patient populations, interventions), methodological aspects (e.g., study design, data collection methods), and statistical analyses.

The risk of bias for each study was evaluated using the Newcastle-Ottawa Scale. Two independent reviewers conducted the assessments, with a third reviewer resolving any discrepancies.

A narrative synthesis was performed to summarize the findings of the included studies. This synthesis described the clinical and methodological characteristics, strengths and limitations, quality of evidence, and risk of bias for each study.

The six included studies were all prospective and published between 2006 and 2017, representing Level 3 evidence according to the Oxford Centre for Evidence-Based Medicine (OCEBM) guidelines. These studies were conducted in Egypt, Turkey, and the United States, involving a total of 123 patients with an average age range of 45.8 to 63 years. Table 1 provides a summary of the main characteristics of the included studies, including patient demographics, study design, and MR spectroscopy details.

Common pre-operative symptoms among patients included neck pain, radicular upper-limb pain, paresthesia, and motor impairment. Magnetic Resonance (MR) spectroscopy was performed using a 1.5 Tesla MRI scanner in four studies and a 3 Tesla MRI scanner in two studies. The main metabolites measured were N-acetyl aspartate (NAA), choline (Cho), creatine (Cr), and lactate levels. The studies focused on metabolite ratios such as NAA/Cr, Cho/Cr, and Cho/NAA. Four studies utilized the modified Japanese Orthopedic Association (mJOA) scale to assess symptom severity and functional status.

The systematic review included six prospective studies that analyzed a total of 123 patients with CSM. These studies provided valuable insights into the metabolic changes detected by MRS and their correlations with clinical severity and outcomes. Symptomatic patients demonstrated consistent reductions in the NAA/Cr ratio, a marker of neuronal loss or dysfunction. For instance, Kendi et al. (19) reported a mean NAA/Cr ratio of 1.46 ± 0.29 in symptomatic patients, which was significantly lower than the 2.01 ± 0.30 observed in healthy controls (p < 0.001; Table 2). Similarly, elevated Cho/Cr and Cho/NAA ratios were observed across studies, reflecting increased membrane turnover or gliosis. Salamon et al. (20) highlighted this pattern, noting higher Cho/NAA ratios in symptomatic patients with T2-weighted imaging changes compared to healthy controls (mean ± SD: 1.51 ± 0.21 versus 0.89 ± 0.23, respectively; p < 0.001; Table 2).

In addition to these metabolic alterations, several studies reported the presence of lactate peaks in symptomatic patients, which were absent in controls. These findings are indicative of hypoxic or inflammatory injury. For example, Taha Ali et al. (21) observed lactate peaks in 33% of symptomatic cases, suggesting that MRS can detect pathological changes that might be missed by conventional imaging techniques.

The correlation between MRS findings and clinical severity, as measured by the mJOA scores, was a recurring theme in the included studies. Lower NAA/Cr and higher Cho/NAA ratios were strongly associated with reduced mJOA scores, reflecting more severe functional impairment. Ellingson et al. (22) demonstrated a statistically significant inverse relationship between Cho/NAA ratios and mJOA scores (correlation coefficient r = −0.65, p < 0.01; Table 2). Similarly, Holly et al. (15) found that pre-operative NAA/Cr ratios were predictive of post-operative recovery, with patients showing higher NAA/Cr ratios experiencing greater improvements in mJOA scores following surgical decompression (p < 0.05; Table 3).

Post-operative changes in metabolite ratios further underscored the potential utility of MRS in monitoring recovery. Across the studies, surgical decompression was associated with normalization of metabolite ratios, particularly an increase in NAA/Cr and a decrease in Cho/NAA ratios. Holly et al. (15) reported that patients who exhibited significant post-operative increases in NAA/Cr ratios experienced better functional outcomes, reinforcing the prognostic value of these metabolic markers (Table 3).

Notably, these metabolite changes were consistently observed in symptomatic patients regardless of T2 imaging findings, highlighting the additional value of MRS in detecting biochemical alterations not evident on conventional imaging. The patterns of reduced NAA/Cr, elevated Cho/Cr, and the presence of lactate peaks in symptomatic patients were robust across the included studies, emphasizing the reliability of these findings in characterizing the pathophysiology of CSM.

The risk of bias was assessed using the Newcastle-Ottawa Scale. Among the four case–control studies, three were rated as having a low risk of bias, while one was considered to have a moderate risk due to limitations in sample size and participant selection. The cohort study and case series were also rated as having a low risk of bias. Detailed risk of bias evaluations are presented in Tables 4, 5.

Our systematic review highlights the potential of magnetic resonance (MR) spectroscopy as a non-invasive imaging modality that provides detailed metabolic information about the spinal cord, which is not obtainable through conventional MR imaging. The studies included in this review consistently demonstrated that symptomatic CSM patients exhibit reduced N-acetyl aspartate to creatine (NAA/Cr) ratios and elevated choline to creatine (Cho/Cr) and choline to NAA (Cho/NAA) ratios compared to healthy controls (Table 2) (19–23). Lower NAA/Cr ratios suggest neuronal loss or dysfunction, while elevated Cho levels indicate increased membrane turnover or gliosis—both of which reflect underlying spinal cord pathology (13, 14).

A significant finding across the studies is the correlation between metabolite ratios obtained from MR spectroscopy and clinical severity as measured by the modified Japanese Orthopaedic Association (mJOA) scores. Ellingson et al. (2015) reported that higher Cho/NAA ratios were associated with lower mJOA scores, indicating more severe myelopathy (22). Similarly, Holly et al. (15) found that patients with higher pre-operative NAA/Cr ratios experienced greater improvement in mJOA scores following surgical decompression, and post-operative increases in NAA/Cr ratios correlated with better neurological recovery (Table 3). These findings suggest that MR spectroscopy not only reflects the current state of spinal cord injury but may also have prognostic value in predicting surgical outcomes.

While conventional MR imaging remains the gold standard for diagnosing CSM, it has limitations in predicting clinical outcomes and correlating imaging findings with symptom severity (4–12). Advanced imaging techniques such as diffusion tensor imaging (DTI) and functional MRI (fMRI) have been explored for their potential to provide additional insights. DTI assesses the integrity of white matter tracts and has shown promise in detecting microstructural changes in the spinal cord (24–28). However, DTI parameters like fractional anisotropy can be influenced by various factors, including edema and inflammation, and may not fully capture the metabolic state of the tissue.

In contrast, MR spectroscopy directly measures biochemical changes within the spinal cord. Recent studies have suggested that combining MR spectroscopy with other imaging modalities may enhance diagnostic accuracy. Wang et al. demonstrated that integrating MR spectroscopy with DTI improved the prediction of surgical outcomes in CSM patients (29). This multimodal approach could offer a more comprehensive assessment of spinal cord pathology.

The metabolic changes observed through MRS, such as reduced NAA/Cr and elevated Cho/Cr and Cho/NAA ratios, provide insights into the underlying mechanisms of neuronal loss, gliosis, and hypoxia in CSM. Understanding these processes informs the potential development of targeted therapies aimed at mitigating neuronal damage. For instance, interventions that restore neuronal energy metabolism, enhance neuroprotection, or reduce gliosis could complement surgical decompression and improve functional outcomes.

These findings align with prior research emphasizing the role of metabolic modulation in preserving neuronal function and preventing progressive spinal cord damage (15, 22). Future studies should evaluate whether pharmacological agents, rehabilitation strategies, or other targeted interventions addressing these metabolic pathways could improve long-term recovery and outcomes in CSM patients.

The metabolic alterations observed in MR spectroscopy reflect underlying pathophysiological mechanisms in CSM. Decreased NAA levels indicate neuronal loss or dysfunction, which may result from chronic compression and ischemia of the spinal cord. Elevated Cho levels suggest increased cell membrane turnover due to demyelination or gliosis (13). The detection of lactate peaks in symptomatic patients further indicates anaerobic metabolism secondary to hypoxia or inflammatory processes (21, 23).

Despite the promising findings, several limitations must be acknowledged. First, the number of studies and sample sizes are relatively small, with the included studies involving a total of 123 patients. This small sample size limits the generalizability of the findings and increases the risk of type II errors. Second, there is heterogeneity among the studies in terms of MR spectroscopy protocols, including differences in magnetic field strength, voxel placement, acquisition parameters, and metabolite quantification methods. Such variability makes it challenging to standardize metabolite ratio thresholds and compare results across studies.

Moreover, potential publication bias cannot be excluded. Studies with positive findings are more likely to be published, which may overestimate the effectiveness of MR spectroscopy in CSM. Additionally, the included studies were conducted in specific geographic locations (Egypt, Turkey, and the United States), which may limit the applicability of the results to other populations.

Technical challenges also exist in performing MR spectroscopy of the spinal cord due to its small size, susceptibility to motion artifacts, and proximity to cerebrospinal fluid and vertebral bone. Advanced techniques and specialized equipment are required, which may not be widely available. The interpretation of MR spectroscopy data requires expertise that may not be readily accessible in all clinical settings.

These limitations may affect the interpretation of our results and the conclusions drawn. The heterogeneity among studies and small sample sizes necessitate caution when generalizing these findings to broader clinical practice. Larger, multicenter studies with standardized MR spectroscopy protocols are needed to validate these preliminary findings.

The cost and availability of MR spectroscopy are practical considerations that could limit its widespread adoption. High-field MR scanners and specialized coils are expensive, and the procedure is time-consuming compared to standard MR imaging. Cost-effectiveness studies are needed to determine whether the benefits of incorporating MR spectroscopy into routine clinical practice justify the additional resources required.

To fully realize the potential of MR spectroscopy in CSM, future research should focus on: