Immuno-proteomic features associated to relapse risk in myelin oligodendrocyte glycoprotein antibody-associated disease

BACKGROUND AND OBJECTIVES: Myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) is an inflammatory demyelinating disorder that overlaps clinically with multiple sclerosis (MS) but immunopathologically distinct. Although often considered an acute inflammatory disease, recurrent attacks in MOGAD can lead to demyelination, axonal injury, and secondary neurodegeneration. Reliable biomarkers associated with relapse risk and disease subphenotypes, including optic neuritis, remain limited. Here, we aimed to define molecular and cellular signatures that distinguish MOGAD from MS as a prototypical neuroinflammatory disease and from Alzheimer disease (AD) as a proxy of neurodegeneration and to identify candidate immune-proteomic features associated with relapse frequency and clinical phenotype in MOGAD. METHODS: CSF, serum, and whole-blood samples from patients with MOGAD (n = 67), MS (n = 49), and AD (n = 36) were profiled using NULISAseq™ CSF proteomics, Olink Explore 3072 CSF and serum proteomics, and high-dimensional mass cytometry for immune cell characterization. In MOGAD, longitudinal clinical data, including total attack counts from the earliest documented attack through follow-up, were integrated with immune and proteomic profiles to assess associations with disease course and clinical phenotype. RESULTS: CSF and blood proteomic profiling revealed distinct inflammatory and cardiometabolic proteomic profiles in MOGAD, differentiating it from both MS and AD. Compared with MS, MOGAD showed relative reductions in lymphocyte populations with regulatory phenotypes. Within MOGAD, relapsing disease was associated with reduced frequencies of CD8(+)CCR7(+)CD31(+)CTLA4(+) T cells and concurrent expansion of double-negative γδ T-cell subsets. IL-13 correlated positively with relapse frequency and inversely with circulating regulatory T cells, whereas IL-32 and CASP4 showed opposite associations, correlating negatively with relapse count and positively with Treg frequency. IL-13 was also inversely associated with CD31-expressing CD8(+) T cells. Phenotype-stratified analyses suggested that these immune-proteomic relationships differed according to clinical presentation, including optic neuritis vs nonoptic neuritis phenotypes. DISCUSSION: This integrative immune-proteomic analysis identifies cellular and molecular features associated with relapsing vs monophasic MOGAD, suggesting a model of impaired peripheral immune regulation in relapsing disease. While exploratory, these findings generate a concrete hypothesis for future longitudinal and functional studies aimed at refining biomarker-based monitoring and informing individualized therapeutic strategies in MOGAD.