Matthias Grothe

Introduction: Patient digital twins (DT) are emerging as a promising paradigm for personalised and predictive healthcare. In neurology, and particularly in multiple sclerosis (MS), the availability of longitudinal, multimodal data provides a strong foundation for developing patient‐specific digital models, including simulation‐based and mechanistic approaches. However, it remains unclear to what extent current approaches represent integrated, clinically embedded DT systems. Methods: We conducted a narrative review to synthesise current concepts, modelling approaches and translational challenges of DT in neurology, using MS as an illustrative use case. Existing approaches were mapped to core DT components, including data integration, predictive modelling, dynamic updating and clinical integration. Results: Current DT‐related approaches in MS demonstrate substantial progress across complementary development tracks, including multimodal data integration, predictive modelling, digital biomarkers, mechanistic simulations, virtual‐patient frameworks and in silico trials. These capabilities represent important building blocks of DT development but have not yet been fully integrated into continuously operating clinical systems. Core DT functionalities, including continuous bidirectional data exchange, real‐time model updating and closed‐loop interaction with clinical workflows, remain limited. Key challenges include data fragmentation, limited interoperability, lack of standardised validation, and regulatory, ethical and economic constraints. Conclusion: MS represents a highly suitable model domain for DT development. Future progress will depend on further maturing and validating individual modelling and data‐integration capabilities, as well as on their integration into continuously updating, clinically embedded systems. Robust validation, scalable infrastructures and appropriate governance frameworks will be essential to establish safe, effective and clinically actionable use in practice.

Disability trajectories in aquaporin-4 immunoglobulin G-seropositive neuromyelitis optica spectrum disorder (AQP4-IgG+ NMOSD) and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) are primarily driven by attack-related damage. Confirmed disability worsening (CDW) independent of attacks has been described but occurs infrequently in AQP4-IgG+ NMOSD and MOGAD. Confirmed disability improvement (CDI) has not been evaluated in large cohorts. We determined the frequency of CDI and CDW independent of attacks and identified clinical factors associated with these outcomes in AQP4-IgG+ NMOSD and MOGAD. This retrospective, multicenter cohort study analyzed data from the German Neuromyelitis Optica Study Group (NEMOS) registry. Adult patients with AQP4-IgG+ NMOSD or MOGAD and longitudinal Expanded Disability Status Scale (EDSS) assessments were included. EDSS episodes were defined as periods with ≥3 EDSS assessments without attacks, obtained ≥90 days after attack. CDW and CDI were defined as sustained EDSS increase or decrease (≥1.5 for baseline EDSS 0; ≥1.0 for EDSS 1.0-5.5; ≥0.5 for EDSS ≥6.0) confirmed after at least 6 months. The primary outcomes were annualized CDI and CDW rates. Risk factors were assessed using multivariable Anderson-Gill regression models. A total of 338 EDSS episodes of 307 patients (n: 202/105, median age at EDSS change: 56/41 years, 88/49% female, both p < 0.001; AQP4-IgG+ NMOSD/MOGAD) were included. Adjusted annualized CDI and CDW rates did not differ between AQP4-IgG+ NMOSD (CDI: 0.083, 95% CI 0.029-0.233; CDW: 0.025, 95% CI 0.007-0.092) and MOGAD (CDI: 0.057, 95% CI 0.012-0.277; CDW: 0.036, 95% CI 0.002-0.513). In AQP4-IgG+ NMOSD, a lower number of prior attacks was associated with higher CDI rates (hazard ratio [HR] 0.89, 95% CI 0.82-0.97). Younger age was associated with increased CDI rates in both AQP4-IgG+ NMOSD and MOGAD (HR 0.96, 95% CI 0.94-0.99, for both). CDI and CDW independent of attacks, although rare, occur in AQP4-IgG+ NMOSD and MOGAD. The association between fewer prior attacks and higher CDI rates in AQP4-IgG+ NMOSD underscores the importance of early attack prevention. Limitations include the retrospective design, and the limited number of CDI and CDW events.

Background: Neuromyelitis optica spectrum disorders (NMOSD) are rare autoimmune diseases of the central nervous system. Global prevalence estimates vary widely, and data for Germany are limited. Objectives: To estimate NMOSD prevalence, incidence, diagnostic settings, comorbid coding with multiple sclerosis (MS), and treatment patterns in Germany. Methods: We analyzed German BARMER health insurance claims data (approximately 9 million persons; 2015–2022). NMOSD cases were identified by ICD-10-GM code G36.0. Frequencies, diagnostic settings, comorbid MS coding (G35.-), and treatment patterns (Anatomical Therapeutic Chemical codes) were assessed. Results: We identified 527 incident NMOSD cases (0.82 per 100,000 insured persons per year); 69% were female, and 42% aged 40–59 years. Most diagnoses occurred in outpatient care (63%), primarily by neurologists (48%) or general practitioners (28%). Prevalence increased from 2.9 to 4.7 per 100,000 between 2015 and 2022, while NMOSD-specific medication use rose from 16% to 31%. NMOSD and MS codes co-occurred in 64% of cases; 49% had concurrent diagnoses within 9 months, and 42% were first diagnosed with MS. Conclusion: NMOSD prevalence and incidence in Germany are increasing, with frequent diagnostic overlap with MS. Findings highlight the need for improved diagnostic differentiation and early specialist involvement.

Background High efficacy therapies (HET) play a crucial role in multiple sclerosis (MS) management. HET discontinuation/de-escalation is a critical decision, especially in different age groups, due to potential changes in relapse rates. We aimed at evaluating the impact of HET discontinuation on annualized relapse rates (ARRs) in people with MS (pwMS) aged ≥ 50 or < 50 years. Methods We retrospectively analyzed data of 1,091 pwMS (German MS Register). ARR before and 12 months after the HET washout period were compared between older and younger patients for switching from HET to HET (H–H), HET to mild/moderate efficacy therapies (H-M) or HET to discontinuation (H–D). Reasons for therapy switches were assessed for all subgroups. Results Most treatment switches continued with another HET (H–H n = 786), while de-escalation (H-M n = 86) or discontinuation (H–D n = 219) occurred less frequently. The minority within each switching group were ≥ 50 years of age (H–H 29%, H-M 22%, H–D 32%). ARR in H–H decreased after switching in both age groups (< 50: 0.19–0.12; ≥ 50: 0.17 to 0.09), increased in H-M < 50 (0.13–0.63) and remained stable in ≥ 50 (0.11–0.08), and increased in H–D < 50 (0.05–0.13) and remained stable in ≥ 50 (0.14–0.11). Main reason for therapy switch was lack of efficacy in H–H, adverse events in H-M regardless of age, patient’s choice (23%) in H–D < 50, and lack of efficacy (26%) in H–D ≥ 50. Conclusions Discontinuation strategies should be individualized, considering age-related changes in disease activity, adverse events and patient’s choice.

Information processing speed is a core cognitive process, highly relevant in everyday life and the most frequent and disabling cognitive symptom in patients with relapsing multiple sclerosis. Correlational evidence from brain imaging suggests involvement of the superior parietal lobe in the speed component of information processing, thereby providing a neurobiological foundation for neuromodulatory interventions. By using regionally specific, focalized transcranial direct current stimulation (tDCS) in healthy individuals and patients with relapsing multiple sclerosis, we provide causal evidence for superior parietal lobe involvement in information processing speed and identified a clinically relevant predictor of tDCS response in patients with relapsing multiple sclerosis. The study employed a registered, randomized, sham tDCS-controlled, three-way-blinded, cross-over trial and a mixed-factors design with eight arms [between-subjects: group (patients/healthy controls; N = 32/group); tDCS polarity (excitatory/inhibitory); within-subjects: stimulation (active/sham tDCS)]. Concurrently with tDCS (1.5 mA; active: 20 min; sham: 40 s), participants completed a computerized version of the Symbol Digit Modalities Test, the current gold standard for quantifying information processing speed impairment in patients with relapsing multiple sclerosis. Data were analysed in a Bayesian framework with generalized linear mixed models. Bayesian modelling provided strong causal evidence of bilateral superior parietal lobe involvement in information processing speed and a double dissociation of stimulation response in patients and controls (i.e. a significant three-way interaction of group × stimulation × polarity). Healthy individuals showed the expected canonical pattern of significantly reduced and increased response latency during anodal or cathodal tDCS, respectively. Across the patient groups, a reversed pattern was found and tDCS response was predicted by baseline Symbol Digit Modalities Test performance. More impaired patients benefited from cathodal tDCS, while less impaired patients benefited from anodal tDCS. For standardized Symbol Digit Modalities Test scores, the transition from beneficial to non-beneficial effects (anodal: < −0.58; cathodal: > −0.70) was consistent across the patient groups. tDCS was well tolerated, with no evidence for differences in mild adverse effects across groups and tDCS conditions. Blinding integrity was confirmed and behavioural outcomes were not explained by factors unrelated to tDCS. Our results provide direct causal evidence for superior parietal lobe involvement in information processing speed in health and disease and suggest that the degree of information processing speed impairment in the patients reflects compensatory or dysfunctional neuroplastic processes that can be counteracted by tDCS in a polarity-specific way. Identified standardized transition scores for the effectiveness of excitatory or inhibitory tDCS will inform future individually tailored stimulation protocols in patients with relapsing multiple sclerosis (trial registration: NCT04667221).

Background: Gait impairments and fatigue are the most common and disabling symptoms in people with multiple sclerosis (PwMS). Objective 6-min walk test (6MWT) gait testing can be improved through body-worn accelerometers, but its association to subjective fatigue and objective fatigability is contradictory. This study aims to validate an algorithm using smartphone sensor data for spatial–temporal gait parameters in PwMS and healthy controls, and evaluate its accuracy in detecting fatigability, and quantify its association with fatigue in PwMS. Methods: We recruited PwMS with mild to moderate disability (EDSS 0.0–6.5) and healthy controls in a supervised, lab-based cohort study. All participants performed the 6MWT while wearing a smartphone at the hip, which collected acceleration data of step count, cadence and walking speed. Algorithm validation included the mean absolute percentage error (MAPE) and Bland–Altman analysis. Fatigability and fatigue were measured in PwMS, with fatigability defined as a 10% decline in gait performance, and fatigue using the fatigue scale for motor and cognitive functions (FSMC). Further, correlations between gait parameters and FSMC were assessed. Results: A total of 38 PwMS and 24 healthy controls were included. The algorithm demonstrated high validity for step count (MAPE < 3%) and cadence (MAPE < 10%). Gait analyses revealed fatigability in between 2.6 and 15.8% of PwMS, with large differences between the gait parameter assessed. Significant correlations were found especially between FSMC motor fatigue scores and step count (r = − 0.50), cadence (r = 0.51) and walking speed (r = 0.50). Conclusion: Smartphone-based gait analysis provides an accessible and valid method for detecting steps and cadence. There are major differences in the assessment of fatigability, but an allover association to subjective motor fatigue.

Background Recurrent attacks in neuromyelitis optica spectrum disorders (NMOSDs) or myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) can lead to severe disability. We aimed to analyse the real-world use of immunotherapies in patients with NMOSD and MOGAD, focusing on changes in treatment strategies, effects on attack rates (ARR) and risk factors for attacks. Methods This longitudinal registry-based cohort study included 493 patients (320 with aquaporin-4 immunoglobulin G (AQP4-IgG) seropositive NMOSD (65%), 44 with AQP4-IgG seronegative NMOSD (9%) and 129 MOGAD (26%)) with 1247 treatments from 19 German and one Austrian centre from the registry of the neuromyelitis optica study group (NEMOS). We analysed unadjusted ARR and implemented survival analyses and Cox proportional hazard regression to assess efficiency and risk factors for subsequent attacks over time. Results Rituximab and azathioprine are the most widely used immunotherapies in NMOSD as well as in MOGAD, with changes in distribution over the last decade. Immunotherapy demonstrated significant therapeutic effects in NMOSD but less pronounced effects in MOGAD. Risk factors for attacks included younger age and prior attacks under the same therapy. Efficacy varied among the different immunotherapies, with azathioprine, rituximab and eculizumab showing significant risk reductions in AQP4-IgG seropositive NMOSD. Conclusions This study provides insights into the evolving treatment landscape and effectiveness of immunotherapies in NMOSD and MOGAD. Established off-label therapies continue to play an important role, especially for patients with stable disease, with emerging evidence supporting newly approved therapies. Future studies are needed to refine treatment algorithms and address the ongoing uncertainties in MOGAD management.

Objective: B‐cell–depletion via CD20 antibodies is a safe and effective treatment for active relapsing multiple sclerosis (RMS). Both ocrelizumab (OCR) and ofatumumab (OFA) have demonstrated efficacy in randomized controlled trials and are approved for treatment of RMS, yet nothing is known on their comparative effectiveness, especially in the real‐world setting.Methods: This prospective cohort study includes patients that were started on either OCR or OFA between September 2021 and December 2023. Patients were followed until June 2024 and recruited at 3 large tertiary centers in Germany (Duesseldorf, Essen, and Giessen). Propensity‐score‐matching was used to address baseline imbalances among patients. Clinical relapses, presence of new or enlarging MRI lesions and 6‐month confirmed disability worsening were evaluated. Non‐inferiority of OFA compared to OCR was evaluated through comparison of Kaplan–Meier‐estimates.Results: A total of 1,138 patients were initially enrolled in the cohort. Following patient selection and propensity‐score‐matching, 544 OCR and 417 OFA patients were included in the final analysis. In our primary analysis, OFA was non‐inferior to OCR in terms of relapses, disability progression, and accrual of MRI lesions. Subgroup analyses confirmed findings in previously naïve and platform‐treated patients. Potential differences between OFA and OCR were seen in patients switching from S1P receptor modulators or natalizumab.Conclusion: We here provide comparative data on the effectiveness of OCR and OFA in patients with active RMS. OFA was non‐inferior to OCR in the overall cohort. Potential differences observed in patients switching from S1P receptor modulators or natalizumab require further validation. ANN NEUROL 2025;97:583–595

Information processing speed (IPS) is a core cognitive process, highly relevant in eve-ryday-life and the most frequent and disabling cognitive symptom in patients with Relapsing Multiple Sclerosis (pwRMS). By using regionally-specific focalized transcranial direct current stimulation (tDCS) in healthy individuala and pwRMS, we provide causal evidence for superior parietal lobe (SPL) involvement in IPS and identfied a clinical predictor of tDCS-response in pwRMS. The study employed a registered, randomized, sham-tDCS controlled, three-way blind-ed, cross-over trial and a mixed-factors design with eight arms [between-subjects: group (pwRMS, controls); tDCS-polarity (excitatory, inhibitory); within-subjects: stimulation (active-, sham-tDCS)]. Concurrently with tDCS, participants completed a computerized version of the Symbol-Digit-Modalities-Test (SDMT), the current gold standard for quantifying IPS impairment in pwRMS. Bayesian modeling with generalized linear mixed models provided strong evidence for polarity specific modulation of IPS by SPL-tDCS and a double-dissociation of stimulation response in pwRMS and controls. Healthy individuals showed the canonical pattern of signifi-cantly enhanced and reduced processing speed during excitatory or inhibitory tDCS. In pwRMS, a reversed pattern was found and tDCS-response was predicted by baseline SDMT performance; i.e., more or less impaired patients benefited from inhibitory or excitatory tDCS, respectively. Our results provide direct causal evidence for SPL involvement in IPS in health and disease and suggest that the degree of IPS impairment in pwRMS reflects compensatory or dysfunctional neuroplastic processes that can be counteracted in a polarity specific way. Identified standardized transition scores for the effectiveness of excitatory or inhibitory tDCS will inform future individ-ually-tailored stimulation protocols in pwRMS. One Sentence Summary Polarity-specific modulation of information processing speed in mul-tiple sclerosis.