MRI researchers, neuroscientists and clinicians involved in research on amyotrophic lateral sclerosis (ALS) and degenerative spinal cord (SC) MR imaging.Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease characterized by a progressive alteration, dysfunction and death of both upper and lower motor neurons. Its diagnosis is based on clinical and electrophysiological criteria, but multi-parametric MRI may also be helpful to provide information about the degree of tissue degradation. Cerebral white and gray matter (WM/GM) involvement, although not yet fully understood, have already been largely described using structural MRI[1, 2]. However, few SC descriptions have been given so far, mainly due to technical limitations related to the small size of the cord and physiological motion. In this study, we used DTI and the emerging myelin-specific inhomogeneous magnetization transfer imaging technique (ihMT)[3, 4] to assess structural changes occurring within specific SC regions of interest. Correlations with disease severity evaluated by the revised ALS functional rating scale (ALSFRS-R) were also investigated.

MR explorations: 9 patients with ALS (56±8yo, ALSFRS-R=37.0±5.8) and 9 age-matched healthy controls (HC, 55±8yo) were scanned at 3T using commercial RF coils and a multi-parametric MR protocol consisting in anatomical, diffusion, and inhomogeneous MT acquisitions (cf. Fig.1 for sequence parameters). Two patients with progressive Multiple Sclerosis (MS) (53±7yo, EDSS=5.0±1.4), for which primary demyelination is expected, were additionally included as preliminary controls to study the specificity of our MR parameters.

Post-processing relied on the SC Toolbox [5], probabilistic atlases [6-8] and a customized MATLAB pipeline allowing specific processing for ihMT SC data [9] and automatic metrics extraction within bilateral corticospinal tracts (CST), posterior sensitive tracts (SP), and anterior gray matter horns (aGM) (Fig.2). Statistical analysis was conducted to compare the metrics obtained in ALS and HC groups. MS patients were not yet considered for statistics (n=2).

Clinical scores for ALS included ALSFRS-R, grip strength, Medical Research Council (MRC) Scale and Upper Motor Neuron (UMN) score [10].

Representative multi-parametric maps collected on patients and controls are illustrated on Fig.3. When analyzing metrics extracted at each level within each ROI, significant structural differences were found in ALS compared to HC. Distribution for the main diffusivities, MT and ihMT are summarized in Fig.4. Interestingly, elevated $$$\lambda_{\bot}$$$ and MTR were observed, with moderate ihMTR decrease, suggesting destructuration and increase of the macromolecular pool, without pronounced demyelination. Statistical differences were also observed for FA (decrease, mainly in the CST) and ADC (increase, mainly in GM) (data not shown), most probably as a consequence of an increased $$$\lambda_{\bot}$$$.

Correlations of MR metrics with clinical scores were also observed (i) in the CST for $$$\lambda_{\bot}$$$, $$$\lambda_{//}$$$,FA and MTR at C2, and for $$$\lambda_{\bot}$$$ C5; (ii) in the SP for all DTI metrics and MTR; and (iii) for almost all metrics in the aGM at C5. Moreover, MTR values, initially higher than those of HCs, seemed to decrease with duration of the disease, suggesting a multi-phasic evolution with preliminary axonal impairment followed by installation of a demyelinating process (decrease of ihMTR, cf. Fig.5). However, no strong correlation could yet support this observation.

Finally, considering the preliminary results extracted from the 2 MS patients, DTI metrics were altered, although not significantly, with the same trend as for ALS patients. Most importantly, the ihMTR was lower by at least 13% (p<0.05) in the whole SC at both levels, all ROI considered, suggesting an important diffuse demyelination of WM tracts.

Altogether, the DTI, MT and ihMT metrics in ALS patients showed significant SC structural changes consistent with ALS pathophysiology, not only in the CST [11-14] and the SP tracts [15], but also in the aGM horns (where motoneuron death is expected). The variations observed in the MTR and ihMTR also seem to reinforce the literature: the alteration observed in MS patients is consistent with the primary demyelination in MS pathophysiology, and tendency of alteration in ALS patients suggestive of a secondary demyelination process.

Although not yet possible due to low statistical power, one could possibly hypothesize that a multivariate analysis combining the different and complementary parameters collected in this study will give new insights into the pathology and lead to a better characterization of the spatio-temporal SC neurodegeneration and stratification of patients (Fig.5). For that purpose, further inclusions and ideally patient follow-ups will be conducted using both multi-parametric MR and atlas-based approaches. Further methodological developments will also be conducted so as to be able to investigate multiple SC levels within an acceptable scan time, and to fully exploit the contrast mechanism provided by the ihMT technique [3, 9].