Linking Microstructural Integrity and Motor Cortex Excitability in Multiple Sclerosis

Abstract

Motor skills are frequently impaired in multiple sclerosis (MS) patients following grey and white matter damage with cortical excitability abnormalities. We applied advanced diffusion imaging with 3T magnetic resonance tomography for neurite orientation dispersion and density imaging (NODDI), as well as diffusion tensor imaging (DTI) in 50 MS patients and 49 age-matched healthy controls to quantify microstructural integrity of the motor system. To assess excitability, we determined resting motor thresholds using non-invasive transcranial magnetic stimulation. As measures of cognitive-motor performance, we conducted neuropsychological assessments including the Nine-Hole Peg Test, Trail Making Test part A and B (TMT-A and TMT-B) and the Symbol Digit Modalities Test (SDMT). Patients were evaluated clinically including assessments with the Expanded Disability Status Scale. A hierarchical regression model revealed that lower neurite density index (NDI) in primary motor cortex, suggestive for axonal loss in the grey matter, predicted higher motor thresholds, i.e. reduced excitability in MS patients (p = .009, adjusted r² = 0.117). Furthermore, lower NDI was indicative of decreased cognitive-motor performance (p = .007, adjusted r² = .142 for TMT-A; p = .009, adjusted r² = .129 for TMT-B; p = .006, adjusted r² = .142 for SDMT). Motor WM tracts of patients were characterized by overlapping clusters of lowered NDI (p <.05, Cohen's d = 0.367) and DTI-based fractional anisotropy (FA) (p <.05, Cohen's d = 0.300), with NDI exclusively detecting a higher amount of abnormally appearing voxels. Further, orientation dispersion index of motor tracts was increased in patients compared to controls, suggesting a decreased fiber coherence (p <.05, Cohen's d = 0.232). This study establishes a link between microstructural characteristics and excitability of neural tissue, as well as cognitive-motor performance in multiple sclerosis. We further demonstrate that the NODDI parameters neurite density index and orientation dispersion index detect a larger amount of abnormally appearing voxels in patients compared to healthy controls, as opposed to the classical DTI parameter FA. Our work outlines the potential for microstructure imaging using advanced biophysical models to forecast excitability alterations in neuroinflammation.

Keywords: NODDI; excitability; motor threshold; multiple sclerosis; neurite orientation dispersion and density imaging; tract-based spatial statistics.

Figure 1

Study pipeline. (1) T1-weighted images were used for reconstruction of the GM that was individually coregistered to FA, NDI, ODI and IVF maps. Motor threshold as measure of excitability was regressed on average FA, NDI, ODI and IVF within the GM of left M1 (1a). Scores of cognitive-motor performance were similarly regressed on average FA, NDI, ODI and IVF within the GM of left M1 (1b). (2) Individual FA maps of all participants were used to compute the TBSS mean skeleton that was applied to FA, NDI, ODI and IVF for comparisons of MS and HC group. Percentage of intersecting significantly different voxels with regions of the sensorimotor area tract template (SMATT) (27) were determined for each diffusion parameter. FA, fractional anisotropy, NDI, neurite density index, ODI, orientation dispersion index, IVF, isotropic volume fraction, MS, multiple sclerosis, HC, healthy control, TBSS, tract-based spatial statistics, SMATT, sensorimotor area tract template, GM, grey matter, M1, primary motor cortex.

Figure 2

Scatterplot with regression line and density estimates of the regression of motor threshold on neurite density index in (A) patients with MS and (B) HC. HC, healthy controls, MS, multiple sclerosis.

Figure 3

Scatterplot with regression line and density estimates of the regression of (A) TMT-A, (B) TMT-B and (C) SDMT z-scores on neurite density index in patients with MS. MS, multiple sclerosis, TMT-A, Trail Making Test part A, TMT-B, Trail Making Test part B, SDMT, Symbol Digit Modalities Test.

Figure 4

Results of tract-based spatial statistics analyses. Color bar indicates TFCE-corrected p-value, thresholded at p <.05. Contrast HC > MS for FA and NDI, and MS > HC for ODI. MNI coordinates: x = 90, y = 109, z = 95. FA, fractional anisotropy, NDI, neurite density index, ODI, orientation dispersion index, L, left, R, right, TFCE, threshold-free cluster enhancement.

Figure 5

Comparison of diffusion parameters contributing to observed group differences. (A) Percentage of FA, NDI and ODI separately or combined showing group differences in the tract skeleton in intersection with SMATT regions. (B) Percentage averaged over regions with any significant group difference (B). FA, fractional anisotropy, NDI, neurite density index, ODI, orientation dispersion index, SMA, supplementary motor area, M1, primary motor cortex, S1, primary somatosensory cortex, PMv, ventral premotor cortex. PMd, dorsal premotor cortex.