Normal appearing white matter and disability in multiple sclerosis
Stepanov, V. N.; Coelho, S.; Ades-aron, B.; Chen, J.; Shepherd, T. M.; Novikov, D. S.; Kister, I.; Fieremans, E.
Show abstract
Background and ObjectivesMultiple sclerosis (MS) is characterized by lesions and atrophy on conventional MRI, yet these often fail to explain disability. Diffusion MRI (dMRI) detects microstructural injury with diffusion tensor (DTI) and kurtosis imaging (DKI) offering sensitivity, and Standard Model Imaging (SMI) providing biologically interpretable parameters. We evaluated whether these clinically feasible dMRI metrics are associated with disability beyond volumetric and lesion measures, and whether effects arise from normal-appearing white matter (NAWM). MethodsThis cross-sectional study included MS patients who underwent 3T MRI including T1- and T2-weighted and a [~]7-minute multi-shell dMRI protocol. Brain volumes (gray matter, white matter, thalamus) and lesion load were derived using FreeSurfer and Icobrain. Diffusion metrics included radial diffusivity (RD) from DTI, radial kurtosis (RK) from DKI, intra-axonal water fraction (f) and fiber dispersion (p2) from SMI. Clinical outcomes were the Expanded Disability Status Scale (EDSS), Multiple Sclerosis Functional Composite (MSFC), 9-Hole Peg Test (9HPT), Symbol Digit Modalities Test (SDMT), and disease duration. Voxelwise and tract-based regions of interest analyses were adjusted for sex and age at onset and repeated after excluding lesions. ResultsNinety-two patients (68 women; mean age 48, range 24-73; median disease duration 14 years; EDSS 3.0, range 0-8.5) were included. dMRI revealed widespread associations with all clinical measures that persisted after lesion exclusion, implicating NAWM. Functional outcomes were tract-specific: 9HPT correlated with corticospinal tract and optic radiations (RD {rho}=0.45; RK {rho}=-0.44; f {rho}=-0.42); MSFC with brainstem and optic radiations (RD {rho}=-0.52; RK {rho}=0.40; f {rho}=0.39). SDMT showed widespread correlation with diffusion and atrophy (white matter {rho}=0.49; thalamus {rho}=0.47). EDSS showed weaker diffusion highlighting commissural disorganization (forceps major/minor {rho} {approx}- 0.30 to -0.32) and was most strongly associated to infratentorial lesion load ({rho}=0.42). Disease duration was dominated by gray-matter atrophy ({rho}=-0.54) with commissural p2 reductions ({approx}-0.45). DiscussiondMRI detects NAWM injury underlying functional impairment beyond atrophy and lesions. SMI adds specificity (f for axonal loss/demyelination; p2 for inflammation). Structural measures capture the effect of cumulative burden in terms of disease duration and EDSS. Together, diffusion, volumetric, and lesion metrics offer complementary insights, supporting multimodal imaging for MS monitoring and stratification.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Magnetisation transfer, diffusion and g-ratio measures of demyelination and neurodegeneration in early relapsing-remitting multiple sclerosis: a longitudinal microstructural MRI study 98%
- Brain disconnectome mapping and serum neurofilament light levels in multiple sclerosis 97%
- In vivo assessment of neurodegeneration in Spinocerebellar Ataxia type 7 96%
Similar papers in this journal
- The sequence of regional structural disconnectivity due to multiple sclerosis lesions 97%
- Spinal cord MRI and MRS Detect Early-stage Alterations and Disease Progression in Friedreich Ataxia 96%
- In vivo myelin imaging and tissue microstructure in white matter hyperintensities and perilesional white matter 96%
Similar papers in this journal
- Cortical morphology predicts long-term placebo response in multiple sclerosis patients 95%
- Frontoparietal connectivity correlates with working memory performance in multiple sclerosis 93%
- Opposite white matter abnormalities in post-infectious vs. gradual onset chronic fatigue syndrome revealed by diffusion MRI 92%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.