BACKGROUND/OBJECTIVES:Characterizing spinal cord multiple sclerosis (MS) lesions in MRI is critical for diagnosis, monitoring, and treatment evaluation. However, current automated approaches for lesion detection and segmentation are typically designed for specific MRI contrasts or acquisition sites, limiting their generalizability in real-world clinical settings where imaging protocols vary widely. This work proposes a robust multi-site, multi-contrast segmentation framework for spinal cord lesions. METHODS:The segmentation model was trained and evaluated on a large-scale dataset comprising 4428 annotated images from 1849 persons with MS across 23 imaging centers, encompassing six MRI contrasts (T1w, T2w, T2*w, PSIR, STIR, and UNIT1) acquired at 1.5 tesla (T), 3 T, and 7 T. RESULTS:Likert-type assessment performed by neuroradiologist ratings demonstrated superior generalization of the model compared to existing contrast-specific pipelines (p < 0.01). Additional experiments evaluated robustness across spinal levels, acquisition resolutions, binarization thresholds, and quantitative evaluation on external labeled datasets. CONCLUSIONS:The proposed model can achieve accurate and reliable spinal cord MS lesion segmentation across heterogeneous MRI data, addressing a key barrier to clinical translation. The model is available in the Spinal Cord Toolbox v7.2 and higher.Code repository: https://github.com/ivadomed/seg-sc-ms-lesion-multicontrast.
Background Interpretation of cognitive impairment (CI) in persons with multiple sclerosis (PwMS) is limited by discrepancies between objective and subjective evaluation. Objectives Evaluate relationships between subjective and objective cognitive measures, accounting for contributors. Methods Multiple Sclerosis Neuropsychological Questionnaire (MSNQ), patient-reported outcomes, Processing Speed Test (PST), Rey Auditory Verbal Learning (RAVLT), Visual Memory Test (VMT), Face Emotion Recognition (FER), Corsi, and Flanker test scores were collected. Spearman correlations, diagnostic metrics, and latent profile analysis assessed associations and cognitive phenotypes. Results In 83 PwMS, only 4.8–15.7% showed objective domain-specific CI, yet 42.2% reported significant CI. However, 51.8% had objective impairment in ≥1 domain. Compared with PwMS without objective or subjective CI, those with subjective but no objective CI had higher Hospital Anxiety Depression Scale-A (9.92 vs. 4.83) and D (8.23 vs. 2.88), and Fatigue Severity Scale (5.56 vs. 3.80) scores, p < 0.05. Latent profile analysis identified a performant cluster (C1, n = 36) and a less performant cluster (C2, n = 46) with higher mean z-scores on PST (0.46 vs. −0.70), VMT (0.32 vs. −0.54), RAVLT (0.99 vs. −0.06), and FER (0.57 vs. −0.80), p < 0.005, in C1. The MSNQ scores were similar between clusters. Conclusion A discrepancy between subjective and objective CI is observed. Subjective CI is associated with anxiety, depression, and fatigue. Multidomain testing and assessment of contributors may help reconcile this discrepancy.
Genome-wide association studies performed in COVID-19 patients have uncovered various loci significantly associated with susceptibility to SARS-CoV-2 infection and disease severity. However, the underlying cis-regulatory genetic factors contributing to heterogeneity in the response to SARS-CoV-2 infection and their impact on clinical phenotypes remain enigmatic. Here, we use single-cell RNA-sequencing to quantify genetic contributions to cis-regulatory variation in 361,119 peripheral blood mononuclear cells of 63 acute COVID-19 patients, 39 convalescent samples, and 106 healthy controls. Expression quantitative trait loci mapping across cell types within each disease state group reveals thousands of cis-associated variants, of which hundreds are detected exclusively in immune cells derived from acute patients. Patient-specific genetic effects dissipate as infection resolves, suggesting that distinct gene regulatory networks are at play in the active infection state. Further, 20.3% of tested loci demonstrate significant cell state interactions with genotype, with pathways related to interferon responses and oxidative phosphorylation showing pronounced cell state-dependent variation, predominantly in CD14+ monocytes. Overall, we estimate that 16.8% of tested genes exhibit gene-environment interaction effects, highlighting the importance of environmental modifiers in the transcriptional regulation of the immune response to SARS-CoV-2. Our findings argue for the existence of extensive gene-environment effects among patients responding to an infection.
Dimensionality-reduction-based visualization is essential for interpreting complex biological data. Yet, unsupervised methods such as t-distributed stochastic neighbor embedding, Uniform Manifold Approximation and Projection, and Isomap reflect only the dominant data structure, which may not align with the goals of downstream analysis or expert-provided annotations. Existing supervised variants only partially address this mismatch and introduce new limitations. Here we present RF-PHATE, a supervised visualization approach that incorporates expert knowledge to reveal label-relevant structure while suppressing extraneous variation. RF-PHATE uses random forests to learn relationships between features and labels and translates this information into low-dimensional embeddings. RF-PHATE handles large datasets and is suitable for both classification and regression tasks. We demonstrate its use across four case studies, including longitudinal multiple sclerosis data, Raman spectral measurements of antioxidant effects, outcomes of patients with COVID-19, and RNA sequencing data with simulated dropout. These applications highlight RF-PHATE's ability to enhance interpretability, manage noise and expose meaningful biological structure, suggesting broad potential for improving data exploration and discovery.
Blood-brain barrier (BBB) integrity naturally declines with age. Brain endothelial cells (ECs) and pericytes (PCs) form the BBB, and aging impairs tight junctions, likely via altered PC-to-EC signaling. However, the molecular mechanisms underlying this impairment remain unclear. Using single-cell RNA sequencing, we profiled 68,316 brain ECs expressing 15,564 genes from young and old mice. Unsupervised clustering and annotation revealed five distinct EC subtypes-Capillary EC1, Capillary EC2, Arterial EC, Venous EC1, and Venous EC2-defined by marker genes Mfsd2a, Plvap, Bmx, Nr2f2, and Vcam1, respectively. Aging shifted EC subtype distribution, with reduced Capillary EC1 (45% vs. 57%) and increased Arterial (33% vs. 16%) and Venous ECs (12% vs. 2%) compared with young mice. Mio analysis further showed that Capillary EC1 and Venous EC2 neighborhoods were less abundant in aged brains. Biotin metabolism was decreased in old vs. young mice, particularly within Capillary EC1, Capillary EC2, and Arterial EC. Although widespread gene downregulation was observed across EC subsets, overall expression trends were largely consistent among clusters. Key genes-Ramp2, Hbb-bs, Ly6c1, Calm1-were less abundant, whereas Rasgrf2 was uniquely enriched in aged mice. Immunohistochemistry confirmed reduced LY6C and RAMP2 and elevated RASGRF2 in aged mouse and human brains. Cell-cell interaction analyses revealed age-associated remodeling of ligand-receptor signaling. Enrichment analyses implicated pathways involved in neurovascular integrity, inflammation, amyloid processing, and vascular remodeling. Collectively, these findings show that aging reprograms EC subtype composition, gene expression, and metabolism, thereby contributing to BBB disruption and neurovascular dysfunction.
Patients aged ≥ 35 years at multiple sclerosis (MS) symptom onset with an Expanded Disability Status Scale (EDSS) score ≥ 3 within the first year are at highest risk of developing aggressive MS (EDSS ≥ 6 within 10 years). Patients without these features are at lowest risk. This study aimed to evaluate whether high-efficacy disease-modifying therapy (HE-DMT) reduced the risk of relapse and disability accumulation in individuals at high risk of aggressive MS, and whether treatment benefit varied by MS severity. This observational cohort study used longitudinal data from two registries: MSBase (international) and OFSEP (France). Adults with relapse-onset MS and an EDSS score recorded within 12 months of symptom onset were included. Patients were classified into high-risk or low-risk groups for aggressive MS based on the above strata; those at intermediate risk were excluded. A pseudo-cohort framework compared periods of continuous HE-DMT (fingolimod, cladribine, monoclonal antibodies) with periods of non-HE-DMT states (on lower-efficacy DMTs or untreated) within each aggressive MS risk stratum. Marginal structural models with repeated adjustment for time-varying confounders of treatment and censoring were used to estimate counterfactual cumulative hazards of relapses and 6-month confirmed disability worsening and improvement. An interaction between MS risk stratum and treatment strategy was tested. A secondary analysis evaluated patients who received an HE-DMT during the study period. In total, 10,405 people (2021 high risk, 8384 low risk) were included. Continuous HE-DMT reduced the risk of relapse in both high-risk and low-risk groups. There was no evidence of a difference in disability outcomes between treatment approaches. There was no evidence of an interaction between aggressive MS risk and treatment effect. In stratified analyses, lowest relapse risk was observed in the low-risk group treated with HE-DMT (hazard ratio [HR] 0.75, 95
Fibrinogen is a blood-derived protein involved in coagulation and can make its way into the central nervous system (CNS) following breakdown of the blood-brain barrier. This molecule has been implicated in multiple sclerosis (MS), a disease marked by inflammation and demyelination in the CNS. However, the effect of this molecule has not been studied on human myelinating cells. This study examines how fibrinogen influences human oligodendrocyte (OL) lineage cells at various stages of development. Using induced pluripotent stem cell-derived (iPSC) OL precursors and human primary OLs, we examined the effects of fibrinogen on cell differentiation, viability, and myelination-related function. Here we show the differential effect of fibrinogen, based on OL-lineage stage. While fibrinogen induced aberrant differentiation of early lineage OLs, by inhibiting their maturation and inducing an astrocytic phenotype, on mature OLs fibrinogen was found to promote myelination capacity, as shown by ensheathment assays as well as on the RNA level. These effects were associated with the activation of bone morphogenetic protein (BMP) signaling, both in early and mature OLs. We further found BMP signaling enrichment in OLs to be correlated with the inflammatory activity of an MS lesion and confirmed fibrinogen deposition on OLs in situ. Unlike previous rodent studies, these findings indicate that fibrinogen has a lineage-dependent effect, where it may be inhibitory earlier in the lineage while promoting OL function in later stages. Understanding this dual role will provide insight into remyelination failure in MS and highlights the importance of timing and target in future therapeutic strategies.
BACKGROUND:The Phase IV MAGNIFY-MS Extension study evaluated the long-term efficacy and durability of cladribine tablets (CladT) in participants with highly active relapsing multiple sclerosis (RMS) during treatment-free Years (Y) 3 and 4. METHODS:Data were analysed for all participants and by subgroups (treatment-naïve vs experienced). Time to no evidence of disease activity (NEDA-3) and first confirmed Symbol Digit Modalities Test (SDMT) score change (⩾4/8 point improvement or ⩽4/8 point worsening) were assessed using Kaplan-Meier analysis. SDMT changes were confirmed if sustained across two visits ⩾166 days apart; others were classified as stable. Percentage brain volume change (PBVC) was analysed using the SIENA-XL method. RESULTS:Of 270 MAGNIFY-MS participants, 219 entered MAGNIFY-MS Extension (64.8% female; mean age ± standard deviation: 40.4 ± 9.45 years). NEDA-3 rates were 78.6% (Y3), 79.2% (Y4) and 54.2% (Y3-Y4 combined). At Y4, 83.1% had no T1 gadolinium-enhancing lesions, 67.4% had no active T2 lesions, and the annualised relapse rate was 0.09. Mean annualised PBVC was <0.4% in Y4. SDMT scores were 4/8-point stable or improved in 79.0%/88.1% of participants, and 84.5% had no 6-month confirmed disability progression. CONCLUSION:Two years of short-course CladT provided sustained clinical and cognitive benefits, supporting the potential for treatment-free remission in RMS. CLINICALTRIALS:gov Identifier: NCT04783935. Date registered: 3 March, 2021.
BACKGROUND:Ocrelizumab, a monoclonal antibody targeting CD20+ B cells, is a high-efficacy therapy for multiple sclerosis (MS). METHODS:Patients with relapse-onset MS treated with ocrelizumab, fingolimod, natalizumab or alemtuzumab for ≥6 months were identified from three registries: MSBase, OFSEP and Danish MS Registry. Pairwise comparisons were performed in the overall cohort and 14 predefined clinicodemographic subgroups based on sex, disease activity, MS duration, Expanded Disability Status Scale, prior therapy and reason for prior treatment cessation. Relapses, progression independent of relapse activity (PIRA) and relapse-associated worsening (RAW) were compared in pairwise-censored groups. RESULTS:Fingolimod was associated with a higher annualised relapse rate (ARR) (0.14 vs 0.06, p<0.001), relapse risk (HR 2.26, 95% CI 1.98 to 2.58), RAW (1.62, 1.08 to 2.43) and lower risk of disability improvement (0.78, 0.63 to 0.96) than ocrelizumab. Superiority of ocrelizumab over fingolimod on relapses was maintained in all subgroups. Natalizumab was associated with marginally higher ARR (0.10 vs 0.07, p<0.001), relapse risk (1.35, 1.16 to 1.57) and RAW (1.77, 1.07 to 2.94) than ocrelizumab. Alemtuzumab was associated with higher ARR (0.18 vs 0.12, p<0.001) and relapse risk (1.48, 1.25 to 1.76) than ocrelizumab, but there was no evidence for a difference in risk of RAW. There was no evidence for a difference on PIRA in any comparisons. Ocrelizumab was superior to natalizumab and alemtuzumab on relapses in patients who were not treatment-naïve, experienced disease activity on the prior therapy or stopped prior therapy due to lack of efficacy. CONCLUSIONS:Ocrelizumab provides superior control of relapses and RAW, especially among patients with prior on-treatment disease activity. Treatment of PIRA remains an unmet need.
BACKGROUND:Advanced magnetic resonance imaging (MRI) of the cerebellum remains underutilized to detect early microstructural abnormalities associated with multiple sclerosis (MS) clinical disability. OBJECTIVES:To examine associations between cerebellar magnetization transfer ratio (MTR) and clinical measures in people with radiologically isolated syndrome (RIS), early relapsing-remitting MS (RRMS), and primary progressive MS (PPMS). METHODS:MTR data were acquired at 3.0 T across four sites in 53 RIS, 202 RRMS, 46 PPMS, and 42 control participants, as part of the Canadian Prospective Cohort Study to Understand Progression in MS (CanProCo). Multiple linear regression analyses evaluated associations between cerebellar MTR and clinical measures. RESULTS:Across MS subtypes, lower cerebellar MTR was associated with greater motor disability, most notably with impaired manual dexterity (β = -1.04 to -0.67). After the false discovery rate correction, two associations remained statistically significant (p < 0.01): lower MTR in the inferior cerebellar peduncles was associated with worse cerebellar function in RRMS, and lower MTR in the anterior lobe was associated with worse manual dexterity in PPMS. CONCLUSION:This large, multi-center, hypothesis-generating study identified two statistically significant associations between cerebellar MTR and clinical disability, alongside several exploratory findings. These results suggest that cerebellar MTR may capture clinically relevant microstructural abnormalities in early MS.
Morphometric measures derived from spinal cord segmentations can serve as diagnostic and prognostic biomarkers in neurological diseases and injuries affecting the spinal cord. For instance, the spinal cord cross-sectional area can be used to monitor cord atrophy in multiple sclerosis and to characterize compression in degenerative cervical myelopathy. While robust, automatic segmentation methods to a wide variety of contrasts and pathologies have been developed over the past few years, whether their predictions are stable as the model is updated using new datasets has not been assessed. This is particularly important for deriving normative values from healthy participants. In this study, we present a spinal cord segmentation model trained on a multisite (n = 75 sites, 1,631 participants) dataset, including 9 different MRI contrasts and several spinal cord pathologies. We also introduce a lifelong learning framework to automatically monitor the morphometric drift as the model is updated using additional datasets. The framework is triggered by an automatic GitHub Actions workflow every time a new model is created, recording the morphometric values derived from the model’s predictions over time. As a real-world application of the proposed framework, we employed the spinal cord segmentation model to update a recently introduced normative database of healthy participants containing commonly used measures of spinal cord morphometry. Results showed that (i) our model performs well compared with its previous versions and existing pathology-specific models on the lumbar spinal cord, images with severe compression, and in the presence of intramedullary lesions and/or atrophy achieving an average Dice score of 0.95 ± 0.03; (ii) the automatic workflow for monitoring morphometric drift provides a quick feedback loop for developing future segmentation models; and (iii) the scaling factor required to update the database of morphometric measures is nearly constant among slices across the given vertebral levels, showing minimum drift between the current and previous versions of the model monitored by the framework. The code and model are open source and accessible via Spinal Cord Toolbox v7.0.
BACKGROUND:Ofatumumab and ocrelizumab are widely used high-efficacy anti-CD20 therapies for relapsing-remitting multiple sclerosis (RRMS), but direct comparative evidence remains limited. We aimed to compare their effectiveness in routine clinical practice. METHODS:We conducted an observational cohort study emulating a target trial using data from the MSBase and Observatoire Français de la Sclérose en Plaques registries (January 2021 to December 2024). Adults with RRMS initiating ofatumumab or ocrelizumab were included. Patients were matched 1:1 using propensity scores. Primary outcomes were annualised relapse rate (ARR) and time to first relapse. Secondary outcomes included time to confirmed disability progression (CDP), progression independent of relapse activity (PIRA), confirmed disability improvement (CDI), MRI activity and treatment discontinuation. Negative binomial and Cox regression models were applied. RESULTS:A total of 5288 patients were matched with a median follow-up of 1.2 years for ofatumumab and 1.4 years for ocrelizumab. ARR was 0.07 (95% CI 0.05 to 0.08) with ofatumumab and 0.04 (0.03 to 0.05) with ocrelizumab, corresponding to an ARR ratio of 1.75 (1.43 to 2.13). Ofatumumab was associated with a lower risk of CDP (HR 0.66; 0.49 to 0.87) and PIRA (0.53; 0.39 to 0.73), but a lower probability of CDI (0.76; 0.60 to 0.96). No significant differences were observed in MRI activity or treatment discontinuation. CONCLUSIONS:Both therapies were highly effective in a large cohort of patients with RRMS, with very low relapse or CDP rates. Ofatumumab was associated with slightly greater disability control, while ocrelizumab more effectively suppressed relapses. These differences were modest, and their clinical relevance requires further evidence and should be interpreted with caution.
Depletion of B cells in multiple sclerosis (MS) is beneficial yet there is no defined pathogenic B cell subset in MS. Here, we demonstrate that T-bet+ memory B cells are rare in control human brain specimens but are found in MS lesions. Transfer of murine T-bet+ memory B cells into mice with ongoing experimental autoimmune encephalomyelitis (EAE), a model of MS, worsened their disability and demyelination. Microglia/macrophage density increased in central nervous system parenchyma even though B cells mostly remained in barriers. In culture, secreted factors from T-bet+ memory B cells promoted macrophage migration. IFN-γ produced by T-bet+ memory B cells activated microglia to secrete chemokines that further enabled macrophage migration. Consistent with their age-associated elevation, conditional deletion of T-bet in B cells lowered EAE severity in aged but not young mice. We define T-bet+ memory B cells as pathogenic in EAE and MS through their IFN-γ-facilitated interactions with microglia/macrophages.
Clinically silent MRI lesions occur frequently in people with relapsing-remitting multiple sclerosis (RRMS) despite disease modifying therapy (DMT). Guidelines only routinely recommend DMT escalation after multiple silent lesions, and adherence is variable. We explored outcomes and the effect of treatment escalation following single and multiple on-treatment silent lesions. This cohort study and emulated target trial used MSBase registry data from 99 clinics in 26 countries between 2007 and 2025. Clinically stable participants receiving any DMT for RRMS with silent lesions versus without silent lesions were compared. Among participants with silent lesions while taking platform or moderate-efficacy DMTs, outcomes following treatment escalation within 6 months versus no treatment escalation (unless a post-MRI clinical event occurred) were compared. The primary outcome was an MS relapse, and the secondary outcome was 6-month confirmed disability worsening. A total of 10,232 participants met inclusion criteria (71.7% female, mean age 41 [SD 11]). The 2-year cumulative incidence of relapse was 27.8% (95% CI: 25.7%-29.9%) in participants with silent lesions versus 14.3% (95% CI: 13.5%-15.2%) without (adjusted hazard ratio [aHR] 1.76 [95% CI: 1.57-1.97]). The 2-year cumulative incidence of disability worsening was 13.8% (95% CI: 12.2%-15.5%) in participants with silent lesions versus 11.4% (95% CI: 10.7%-12.2%) without (aHR 1.38 [95% CI: 1.18-1.62]). Rates of relapse and disability worsening were higher following single and multiple silent lesions versus no silent lesions. The emulated trial included 2,264 participants with ≥1 silent lesion on platform or moderate efficacy DMTs, 286 of whom escalated DMT within 6 months following silent lesions. The 4-year cumulative incidence of relapse was lower following treatment escalation (16.8% [95% CI: 12.4%-23.4%]) versus continuation (38.9% [95% CI: 35.8%-42.1%]), aHR 0.34 (95% CI: 0.23-0.47), with similar aHRs following single and multiple silent lesions. The 4-year cumulative incidence of disability worsening was similar following treatment escalation (16.0% [95% CI: 10.8%-22.2%]) versus continuation (17.7% [95% CI: 15.3%-20.1%]), aHR 0.89 (95% CI: 0.56-1.33). People with RRMS with single or multiple on-treatment silent MRI lesions have higher subsequent risks of relapse and disability worsening than people without silent lesions. DMT escalation mitigates the relapse risk, though disability worsening continues at a similar rate over 4 years. Contrary to guidelines, DMT escalation should be considered after single or multiple silent lesions.
Background Cognitive impairment (CI) is a common feature of multiple sclerosis (MS). Radiologically isolated syndrome (RIS) can demonstrate CI patterns similar to MS, providing an opportunity to explore the associations between CI and early stage disease pathology. Objectives To investigate myelin damage in RIS normal appearing white matter (NAWM) using the myelin heterogeneity index (MHI) and to determine the relationship between MHI and processing speed. Methods A total of 28 people with RIS and 22 controls completed an MRI, including multi-component T 2 mapping to calculate MHI for whole brain (WB), corpus callosum (CC), superior longitudinal fasciculus (SLF), and cingulum (CING) NAWM. The RIS cohort additionally completed the Processing Speed Test (PST), a measure of processing speed. Results CC NAWM MHI was higher in RIS versus controls (indicative of myelin damage, p = 0.0140), with similar trends seen in WB, SLF, and CING ( p > 0.05). A moderate correlation was found between CING NAWM MHI and PST scores ( ρ = –0.39, p = 0.040), with similar trends seen in WB, CC, and SLF ( p > 0.05). Conclusion Diffuse myelin damage was detected in RIS NAWM, along with associations between increased damage and slower processing speed, suggesting a potential pathological mechanism for RIS-related CI.
Multiple sclerosis (MS) is a disorder of the CNS in which autoreactive immune cells migrate through a damaged blood-brain barrier, resulting in focal demyelinating lesions. Beyond focal lesions, there are also diffuse 'surface-in' gradients of pathology in MS, wherein damage is most severe directly adjacent to CSF-contacting surfaces, such as the subpial and periventricular areas. This observation suggests that toxic factors within MS CSF contribute to the emergence and/or evolution of surface-in gradients. Directly separating the CSF from the periventricular parenchyma are ependymal cells-a glial epithelium-that are equipped with tufts of motile cilia, which are critical for circulating CSF solutes and regulating local fluid flow. While damage to ependymal cilia has the potential to drastically modify CSF homeostasis and thus contribute to the damage of CSF exposed regions, these motile cellular structures have yet to be investigated in the context of MS. We first conducted single-cell RNA sequencing of fresh human periventricular brain tissue containing ependymal cells from patients with MS and non-MS disease controls. We subsequently collected CSF from patients with MS and exposed cultured rodent ependymal cells to this CSF to evaluate the impact on ependymal ciliary function. To complement our direct evaluation of cilia in the context of MS, we also confirmed whether cilia were altered in an animal model of MS, experimental autoimmune encephalomyelitis (EAE), and designed a novel transgenic animal model to evaluate the cellular and behavioural effect(s) of adult ependymal ciliary disruption. Single-cell RNA sequencing analysis of human ependymal cells in MS demonstrated large-scale dysregulation of ciliary genes, and in situ stains of MS brain tissue confirmed a loss of ependymal cilia. Exposure of ependymal cells to MS CSF led to transcriptional modification of ciliary gene and protein expression and reduced ciliary beating frequency. Likewise, analysis of ependymal cells in EAE demonstrated altered cilia gene and protein expression. We showed that IFNγ, which is elevated in MS CSF, could alter cilia protein expression and motility. Lastly, conditional knockout of Ccdc39 in ependymal cells of adult mice led to transient ventricular enlargement, increased periventricular microglial density and alterations in nesting behaviour. These data suggest that motile cilia in ependymal cells are dysregulated in CNS autoimmunity. More importantly, they suggest that ependymal cilia disruption could play a role in periventricular pathology formation in MS and be associated with behavioural deficits underlying non-motor symptomatology.
The 2024 revision of the McDonald diagnostic criteria is an important step toward earlier and more inclusive diagnosis of multiple sclerosis. Realizing its full potential will require continued refinement of biomarkers, disease stratification and clinical trial approaches.