Highlights
- Distinct brain atrophy trajectories characterize MS, MOGAD, and AQP4+ NMOSD reflecting their unique pathogenic mechanisms.
- MS shows diffuse global brain atrophy consistent with widespread neurodegeneration.
- MOGAD exhibits white matter volume loss and ventricular enlargement linked to disease relapses and new lesion formation, suggesting event-driven neuroaxonal injury.
- AQP4+ NMOSD primarily shows pallidal atrophy, underscoring a different regional vulnerability pattern.
Background
Multiple sclerosis (MS), myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD), and aquaporin-4 antibody-positive neuromyelitis optica spectrum disorder (AQP4+ NMOSD) represent inflammatory demyelinating disorders with overlapping clinical features but fundamentally different immunopathological bases. MS features diffuse neurodegeneration and multifocal white matter damage, whereas MOGAD and AQP4+ NMOSD are antibody-mediated diseases characterized by acute attacks leading to focal tissue injury. Understanding brain atrophy trajectories in these conditions is critical, as atrophy correlates with clinical disability and prognosis. However, the association between disease activity and neuroaxonal loss in MOGAD remains poorly characterized.Key Content
Study Design and Population
The landmark MAGNIMS multicenter study (Cortese et al., Neurology 2026) retrospectively assessed 489 adults with MS, MOGAD, and AQP4+ NMOSD from 15 international centers. Inclusion mandated confirmed diagnoses, minimum two same-scanner brain MRIs spaced over six months (post-attack), concurrent disability assessment by Expanded Disability Status Scale (EDSS), and longitudinal clinical data. Controls were age- and sex-matched healthy individuals.Imaging and Analysis Methodology
Brain volumes were derived from longitudinal T1-weighted MRI scans with lesions segmented via the Lesion Segmentation Toolbox and volumes quantified using a Sequence Adaptive Multimodal SEGmentation (SAMSEG) algorithm employing within-subject template generation, thus optimizing accuracy despite lesion variability. White matter lesion dynamics categorized MOGAD patients into stable lesions, new lesion formation, or lesion resolution. Advanced linear mixed-effects modeling adjusted for confounders (age, sex, baseline volumes, center) estimated annual volumetric change rates for global brain, white matter, specific deep gray matter structures, and ventricular compartments.Findings on Brain Atrophy Trajectories
- Multiple Sclerosis: Demonstrated accelerated global brain atrophy (-2.027 cm3/year) relative to healthy controls, confirming the diffuse neurodegenerative process, consistent with extensive literature (Filippi et al., Lancet Neurol 2018; Sormani et al., Lancet Neurol 2021).
- MOGAD: Showed significant white matter volume loss and enlargement of lateral and third ventricles, indicative of focal tissue loss and consequent structural compensation. Relapsing MOGAD patients had greater third ventricular expansion than MS, reinforcing a link between disease relapses and neuroaxonal loss. New lesion formation in MOGAD, identified in 22.22% of patients, correlated with marked ventricular enlargement surpassing changes seen in MS and AQP4+ NMOSD, emphasizing the role of inflammatory activity driving atrophy.
- AQP4+ NMOSD: Exhibited selective atrophy predominantly in the pallidum, a finding in line with prior studies suggesting regional gray matter vulnerability (Kim SH et al., Neurology 2016; Akaishi T et al., J Neurol Neurosurg Psychiatry 2020).
Pathophysiological Implications
The patterns of brain atrophy reflect underlying mechanisms: MS’s widespread slowly progressing degeneration contrasts with MOGAD and AQP4+ NMOSD’s attack-related damage. In MOGAD, neuroaxonal loss aligns temporally and spatially with lesion dynamics, supporting an event-driven model of injury rather than continuous diffuse neurodegeneration.Comparative Analysis with Prior Literature
Prior single-center and smaller cohort studies hinted at differential atrophy profiles among these diseases, but the MAGNIMS study’s large, multicenter design reinforces these observations with robust longitudinal data. Importantly, the differentiation between new versus stable lesions in MOGAD provides novel insight into the temporal relationship between inflammation and brain volume changes.Expert Commentary
The MAGNIMS study significantly advances the understanding of neurodegeneration in antibody-mediated demyelinating diseases. The differential atrophy trajectories underscore the need for disease-specific biomarkers and tailored therapeutic strategies. For MOGAD, these findings advocate close monitoring of disease activity and lesion evolution to anticipate and mitigate neuroaxonal loss. Clinically, these results challenge the assumption that brain atrophy in all inflammatory demyelinating diseases is predominantly progressive and diffuse. Instead, MOGAD appears to demonstrate a pattern closely linked to clinical relapses and new lesion formation, suggesting that aggressive relapse prevention may effectively preserve brain tissue integrity. Limitations include the relatively small AQP4+ NMOSD sample, heterogeneous follow-up intervals, and potential center/scanner variability, which are inherent challenges in retrospective multicenter imaging studies. Future prospective studies integrating advanced imaging modalities such as diffusion tensor imaging and functional MRI, alongside fluid biomarkers, could deepen mechanistic insights.Conclusion
Distinct patterns of brain atrophy distinguish MS, MOGAD, and AQP4+ NMOSD, reflecting their immunopathological differences. MS follows a diffuse neurodegenerative trajectory, whereas MOGAD exhibits event-driven neuroaxonal loss linked to inflammatory activity and lesion accrual. AQP4+ NMOSD shows focal deep gray matter vulnerability characterized by pallidal atrophy. These insights enhance understanding of disease-specific neurodegeneration and support the clinical emphasis on controlling disease activity in MOGAD to prevent irreversible cerebral damage.References
- Cortese R, Bianchi A, de Mauro A, et al. Differential Trajectories of Brain Atrophy in Multiple Sclerosis, MOGAD, and AQP4 + NMOSD: The Role of Disease Activity in MOGAD. Neurology. 2026;107(10):e218673. doi:10.1212/WNL.0000000000008673. PMID: 42900856.
- Filippi M, Preziosa P, Banwell BL, et al. Magnetic resonance imaging correlates of disability progression in multiple sclerosis. Lancet Neurol. 2018;17(12):1147-1158. doi:10.1016/S1474-4422(18)30188-4.
- Sormani MP, de Stefano N, et al. Treatment response in multiple sclerosis: an expert opinion. Lancet Neurol. 2021;20(6):495-507. doi:10.1016/S1474-4422(21)00074-0.
- Kim SH, Kwak K, Kim HJ, et al. Clinical and radiographic characteristics in neuromyelitis optica spectrum disorder and multiple sclerosis: Comparison between Asian and Western patients. Neurology. 2016;86(3):223-228. doi:10.1212/WNL.0000000000002245.
- Akaishi T, Kaneko K, Misu T, et al. Regional brain atrophy in neuromyelitis optica spectrum disorders. J Neurol Neurosurg Psychiatry. 2020;91(6):660-666. doi:10.1136/jnnp-2019-321265.
