Neurite orientation dispersion and density imaging in myelin oligodendrocyte glycoprotein antibody-associated disease and neuromyelitis optica spectrum disorders

Abstract

Background

Aquaporin-4 antibody positive (AQP4+) neuromyelitis optica spectrum disorders (NMOSD) and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) are two distinct antibody-mediated neuroinflammatory diseases. Diffusion Tensor Imaging (DTI) and Neurite Orientation Dispersion and Density Imaging (NODDI) are advanced diffusion-weighted MRI models providing quantitative metrics sensitive to cerebral microstructural changes. This study aims to differentiate brain tissue damage in NMOSD and MOGAD from controls and investigate its association with clinical disability, using NODDI and DTI-derived measures, including fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD) and radial diffusivity.

Methods

This study included 31 AQP4+ NMOSD, 21 MOGAD patients and 45 healthy controls. Clinical information included disease duration, Expanded Disability Status Scale (EDSS), Timed 25 Foot Walk test (T25FW), Nine-Hole Peg Test (9HPT), Symbol Digit Modalities Test (SDMT) and monocular 100 % high contrast visual acuity (HCVA). All participants underwent MRI scanning with multi-shell diffusion-weighted imaging, T2w fluid-attenuated inversion recovery and T1w magnetization prepared–rapid acquisition gradient echo sequences to obtain manually segmented T2-hyperintense white matter lesions (WML) and normal-appearing brain tissue (NABT) masks, including white matter (NAWM), cortical and deep gray matter (NACGM, NADGM). DTI and NODDI metrics were compared between groups using region-of-interest (ROI) analysis and tract-based spatial statistics. Tissue-weighted means were obtained for the NODDI metrics (weighted neurite density index, wNDI; weighted orientation dispersion index, wODI). Group differences in ROI analyses were assessed using age and sex adjusted linear regression models, followed by post-hoc comparisons with estimated marginal means. Stepwise multivariable linear regression models were used to evaluate the association between MRI biomarkers and clinical outcomes.

Results

NMOSD patients had higher T2 lesion volume (1120.5 mm³ vs. 374.6 mm³, p<.001) and number (median 22 vs. 6, p<.001) than MOGAD patients. Both NMOSD and MOGAD lesions displayed lower wNDI and higher isotropic volume fraction (ISOVF) compared to microvascular lesions in controls (p<.05). In NACGM, NMOSD patients showed higher wODI but lower ISOVF compared to HC (p=.029). MOGAD patients had lower wNDI in NACGM compared to NMOSD (p=.012). Tract-based spatial statistics revealed damage to specific white matter abnormalities in NMOSD, with higher AD, ODI and ISOVF compared to controls, particularly in the corpus callosum and corticospinal tract. Clinical associations in NMOSD included higher EDSS with higher NAWM ISOVF (R²=0.46, p=.006), higher 9HPT with lower intralesional FA and higher NAWM MD (R²=0.54, p=.022), lower SDMT with lower intralesional FA and higher NACGM ISOVF (R²=0.54, p=.013), worse visual acuity with higher NAWM wODI. In MOGAD, higher EDSS was associated with lower NAWM FA (R²=0.29, p=.022), slower T25FW with higher NADGM ISOVF (R²=0.48, p<.001), lower SDMT with higher NAWM ISOVF (R²=0.62, p=.005) and worse visual acuity with higher NADGM MD.

Conclusion

NODDI and DTI measures are sensitive to pathological alterations in myelin and axon integrity, as water diffusion is less restricted in demyelinated tissue. Compared to MOGAD, patients with NMOSD tend to exhibit more extensive chronic white matter damage, demyelination or axonal injury. NODDI demonstrates greater sensitivity and specificity to alterations in NACGM compared to DTI. Given their association with clinical disability, NODDI metrics appear to be valuable neuroimaging biomarkers for assessing microstructural damage in clinical research.