Imaging Poster Presentation

P0659 - White matter tracts that overlap with hubs are preferentially protected against multiple sclerosis pathology (ID 926)

  • M. Clarke
  • M. Clarke
  • D. Archer
  • K. Yoon
  • R. Dortch
  • S. Smith
  • J. Xu
  • G. Cutter
  • F. Bagnato
Presentation Number
Presentation Topic



The thalamus and the putamen, highly-connected brain areas (hubs), are vulnerable to MS-induced atrophy. Here we investigate if white matter (WM) tracts that pass through hubs have different susceptibility to MS pathology, due to anterograde and retrograde neurodegeneration, compared to tracts do not pass through these hubs. We use the apparent intra-axonal volume fraction (Vax) derived from the multi-compartment diffusion MRI spherical mean technique (SMT) as well as the neurite density index (NDI) and isotropic volume fraction (IVF) derived from the neurite orientation dispersion and density imaging (NODDI) coupled with high-resolution tractography to assess the degree of tract-specific axonal integrity.


(1) To compare tract-specific lesion burden, Vax, NDI and IVF between WM tracts that overlap with either the thalamus or the putamen (hub+ tracts) and those that do not (hub- tracts); and (2) to assess the relationship between these MRI metrics and those of physical impairment, as measured by the Expanded Disability Status Scale (EDSS) and the Timed 25-Foot Walk test (T25FW).


EIghteen patients underwent 3T MRI inclusive of T1- and T2-weighted sequences, SMT and NODDI. Using probabilistic tractography, we reconstructed 49 WM tracts, 12 hub+ tracts and 37 hub- tracts. Using T-tests, we compared the following MRI metrics between hub+ and hub- tracts: T1 and T2 lesion burden (calculated as percentage of tracts occupied by T1 and T2 lesions) as well as whole tract and normal-appearing WM (NAWM) average NDI, IVF and Vax values. Spearman correlation assessed the relationship between these MRI metrics and those of clinical impairment.


Hub+ tracts had a lower mean %T1 lesions (0.86 vs. 1.99) and %T2 lesions (2.90 vs. 5.42), as well as higher whole tract NDI (0.61 vs. 0.57), NAWM NDI (0.61 vs. 0.58), whole tract Vax (0.49 vs 0.44) and NAWM Vax (0.50 vs. 0.44), all significant at the p<0.001 level. Stronger correlations were seen between clinical measures and hub+ tracts compared to hub- tracts, with the strongest correlation for %T1 lesions and the T25FW (r= -0.59, p<0.0001).


WM tracts that overlap with the thalamus and the putamen have a higher degree of axonal integrity and lower lesional burden, suggesting a protective role of hubs. However, given the ramifications of disease present at the level of hub+ tracts, disease here retains a major impact on disability. If confirmed, our data suggest the role of disease location in relation to hubs as guidance for treatment personalization, considering more aggressive approaches for patients presenting with MRI changes in hub+ fiber tracks.