Cervical spinal cord MRI in ALS individuals: a systematic review and meta-analysis

Disease tracking and prognostication of amyotrophic lateral sclerosis (ALS) can be quite challenging in people living with ALS, due to the complexity of central nervous system disease biology. This systematic review and meta-analysis aim to summarize cervical spinal cord quantitative MRI (qMRI) biomarker changes in individuals with ALS.

PubMed, Scopus, Cochrane Library, and Web of Science databases were searched up to August 2023. The terms used were "ALS", "cervical spinal cord", "MRI"," diffusion tensor imaging (DTI)", " fractional anisotropy (FA)", " mean diffusivity (MD) "," magnetization transfer ratio (MTR)", " cross-sectional area (CSA)", " radial diffusivity (RD) ", and " atrophy ". The Newcastle-Ottawa scale (NOS) was used to assess study quality. We calculated the pooled: 1) Standardized mean difference (SMD) and 95% CIs for comparative assessment of qMRI parameters in ALS individuals and the healthy population. 2) Estimate the mean of qMRI parameters for normative values in two groups by CMA software. Heterogeneity and publication bias were determined by the I-squared statistic and funnel plots.

Thirty studies, with 1817 participants (35.9 % female) were included in this review, and 29 had a NOS ≥ 5 which indicates high-quality of data overall. The SMD analysis showed (a) significant decrease in CSA along the whole length of cervical cord (C1-C7) (p value < 0.0001), with a preferential thinning of the cervical enlargement region (C4-C6 region) (p value < 0.0001) (b) significant decrease in FA (p value < 0.0001), particularly FA left lateral corticospinal tract (p value < 0.0001) and (c) a significant increase in MD (p value < 0.0001) in ALS individuals compared to controls. The pooled analysis reveals that the mean (SD) values for ALS individuals versus controls for (a) CSA (in mm²) were C1 [73.4 (0.75), 78.5 (0.67), 6.9 % difference]; C2 [70.6 (3.1), 71.5 (3.5), 1.2 % difference]; C3 [69.8(1.5), 74.9 (1.9), 7.3 % difference]; C4 [71.9 (1.8), 77.6 (2.8), 7.9 % difference]; C5 [71.8 (2.5), 79.5 (3.3), 10.7 % difference]; C6 [66.8 (2.7), 73.7 (3.7), 10.3 % difference]; C7 [56.7 (F2.2), 62.1 (2.5), 9.5 % difference]; (b) FA [0.54 (0.03), 0.56 (0.03)]; (c) MD[1.11 (0.18), 0.88(0)]; and (d) FA LLCST [ 0.65 (0.04), 0.77 (0.04)], respectively. The mean (SD) value of the MTR and RD for ALS individuals was 40.3 (2.3), and 0.70 (0.0).

qMRI metrics of spinal cord show discriminatory potential between ALS and healthy controls. The selective atrophy of the cervical enlargement (C4-C6) is replicable across multiple studies as seen in this metanalysis and hence is a potential imaging marker for quantifying and tracking lower motor neuron degeneration in ALS.