ObjectiveTo evaluate real-world Radiologically isolated syndrome (RIS) management practices across Germany, Austria, and Switzerland and identify key gaps for future guideline development.MethodsBetween October 1 and October 16, 2024, neurologists across university hospitals, teaching hospitals, municipal clinics, and private practices in Germany, Austria, and Switzerland (DACH) participated in an anonymous online survey on RIS diagnosis, treatment, and monitoring practices.ResultsAmong 127 physicians from diverse care settings, RIS management showed both marked heterogeneity and notable convergence. While the diagnostic work-up was relatively uniform (>85% performed comprehensive magnetic resonance imaging (MRI) and cerebrospinal fluid (CSF) analysis, as well as evoked potentials), awareness and implementation of the 2023 RIS criteria varied substantially (34% consistent use, 42% partial implementation, 16% non-adoption, and 8% unaware). Treatment practices showed the greatest heterogeneity: although 54% considered therapy initiation and clinicians agreed on treatment triggers (high lesion burden, new MRI activity, inflammatory CSF), 69% treated less than 25% of their RIS cases. Off-label treatment strategies, medication choices, and escalation approaches varied widely across settings, with 72% lacking standardized algorithms. Structural barriers were consistent across all countries and care levels, with reimbursement challenges (61%), patient communication difficulties (42%), and limited access to specialists outside university centers identified as key obstacles. Recent treatment trials increased therapeutic confidence, yet clinicians emphasized the urgent need for national recommendations to harmonize diagnostic workflows, treatment pathways, and reimbursement structures.ConclusionReal-world RIS care in the DACH region is characterized by substantial practice heterogeneity, coexisting with shared clinical approaches, which reflects the absence of formal guidelines and approved therapies. While the 2024 McDonald criteria now enable immunotherapy for asymptomatic individuals at high-risk for MS, clinical uncertainty persists regarding risk stratification, treatment decisions, and communication frameworks, challenges applicable to both RIS and preclinical MS. These findings reveal critical unmet needs and demonstrate the urgent necessity for harmonized DACH-wide recommendations addressing diagnostic standards, risk-adapted treatment pathways, patient communication, reimbursement integration, and cross-sector collaboration.

BackgroundThe absence of standardization in the documentation of routine clinical data complicates research usage of retrospective data on a large-scale basis. Medically trained personnel is required for interpretation and conversion into a structured format making it time and cost intensive and creating a potential bias of such data. To address these challenges, we have developed a semi-automated approach for evaluating Multiple Sclerosis (MS) outpatients reports that utilizes different large-language models (LLM) and their consensus in comparison to manual evaluation.MethodsWe used several commercially available LLMs by OpenAI, Anthropic and Google to create a structured output of several variables with differing complexity of 30 anonymized outpatient reports with zero-shot-learning. We added a consensus output by combining the results of three different LLMs. Over several runs, we adapted the prompt, compared the results with a reference and assessed the error rate. Any deviation from the reference was considered an error. A true-error rate was determined for the LLM consensus output and the neurology specialist output, where only content deviations are counted as errors.ResultsThrough 9 iterations of improving the structure and content of the prompt, we have seen a clear reduction in the error rate of the various LLMs. By creating an LLM consensus with the final prompt design, we were able to overcome a ceiling effect in reducing the error rate. With a true-error rate of 1.48%, the LLM consensus shows a similar error rate as neurologists (around 2%) in the creation of structured data.DiscussionOur method enables fast and reliable LLM-based analysis of large clinical routine data sets of varying complexity with a low technical barrier to entry. By generating an LLM consensus, we were able to considerably improve the quality of the output making it comparable to data created by neurology specialists. This approach allows large amounts of unstructured data to be analyzed in a time and cost-efficient manner. Nevertheless, the evaluation of errors in results produced by LLM remains difficult. Scientific work using such methods must continue to be subject to strict testing of the validity of the method in the future.

Multiple sclerosis (MS) is an immune-mediated chronic neuroinflammatory and neurodegenerative disorder. Inflammation in MS disrupts the barriers between blood and central nervous system and affects transport and diffusion of metabolites between blood and cerebrospinal fluid (CSF). In this exploratory retrospective case–control study, we used targeted metabolomics to evaluate differences in serum and CSF amino acid and neurotransmitter levels between patients with MS (n = 73) and non-neuroinflammatory controls (n = 78). The influence of patient characteristics, including sex, age, disease duration, severity and treatment status, was also analzyed. Although no significant differences in serum and CSF metabolite levels were found between MS and control patients, a stratification by sex uncovered significantly reduced metabolites in male MS patients compared to male controls in CSF but not in serum. While in male MS patients CSF histidine levels were decreased, female MS patients showed increased levels. Further, sex-specific associations of amino acids and neurotransmitters with disease duration and disability were observed. MS patients exhibited enhanced positive correlations between CSF and serum analyte levels. In serum, only a few amino acids, along with serotonin and glutathione, were associated with MS disease duration. Overall, this study suggests that targeted metabolomics of selected analytes in matched CSF and serum samples is a valuable approach for assessing alterations in CSF–serum metabolite associations in MS, as well as sex-specific imbalances between excitatory and inhibitory neurotransmitters across disease duration. Our findings further highlight the importance of considering sex as a key biological factor in MS.

Background: Lifestyle factors, such as smoking cigarettes and alcohol consumption, can influence disease progression in Multiple Sclerosis (MS). Objective: This study aimed to investigate the associations of smoking cigarettes and alcohol consumption with clinical parameters in people with MS (pwMS). Design: This is a prospective cohort study, analysing data from the German NationMS cohort. Methods: We analysed data of 1374 patients recruited between 2010 and 2017. Main outcomes were Expanded Disability Status Scale (EDSS), Timed 25-Foot Walk (T25FW) and Paced Auditory Serial Addition Test (PASAT). Secondary outcomes were number of T2 and GD+ lesions. We conducted logistic and linear mixed model regression, adjusted for age, sex, educational level, MS subtype. Results: Smoking cigarettes was associated with increased odds of reaching pre-defined cut-off values for EDSS and T25FW, as well as with increased mean differences across follow-ups. Alcohol consumption was associated with decreased odds of reaching the cut-off for EDSS and T25FW. Findings for PASAT were less consistent. We found no significant associations for T2 and GD+ lesions. Conclusion: This study provides further evidence that smoking cigarettes accelerates MS progression and that pwMS should be advised to quit smoking. However, this negative impact could not be demonstrated for alcohol consumption. Reverse causation cannot be ruled out and further research in this area is warranted. Trial registration: Instead of registration in a primary trial registry, a comprehensive study protocol was published a priori and is available online: https://osf.io/3xscy .

BACKGROUND:Understanding sex differences in relapsing-remitting multiple sclerosis (RRMS) and initiation of disease-modifying treatments (DMTs) is crucial for tailored approaches. OBJECTIVE:The objective of this study is to analyze sex-specific differences in early RRMS. METHODS:We analyzed data of therapy-naïve adults from the German NationMS cohort to describe first symptoms, onset relapse treatment, disability evolution, and DMT exposure separated by sex to investigate previously described sex differences. RESULTS:Relapse presentation and treatment were similar (p = n.s.). Time to Expanded Disability Status Scale (EDSS) ⩾3.0 was comparable between sexes (adjusted hazard ratio, 95% confidence interval (95% CI): 1.32 (0.95-1.81)). DMT exposure did not differ (p = 0.60). Around 5.0% of both sexes received initial high-efficacy (HE) DMT. Younger age (odds ratio (OR) (95% CI): 0.95 (0.92-0.98); p = 0.000847), higher baseline EDSS (1.79 (1.40-2.27); p = 0.00000218), and RRMS diagnosis (2.26 (1.28-4.17), p = 0.006703) were associated with initial HE-DMT, but not sex (0.99 (0.57-1.77), p = 0.943166). CONCLUSION:We did not observe major sex differences in early MS as described earlier regarding initial presentation and disability evolution suggesting a change of MS course. The decision for initial HE-DMT was influenced by younger age and higher EDSS, but not sex suggesting a lower sex bias regarding the initial treatment decision, yet only investigated in specialized academic MS centers.

Clinical–neurological examination may underestimate internuclear ophthalmoplegia (INO). We investigated immersive virtual reality (VR)-based eye-tracking in patients with multiple sclerosis (MS), other neurological disorders (ND), and healthy controls (HC) to detect INO. In this prospective monocentric study, 471 participants (243 MS, 104 ND, 124 HC) underwent horizontal prosaccade testing (± 10°, ± 20°) using VR-based oculography. INO was determined in a blinded manner based on software-generated reports comparing adduction and abduction velocities. VR-based and clinical assessments were compared using Cohen’s kappa (κ). Associations with demographic and clinical variables were analyzed using logistic regression. VR-based assessment identified 43 cases of INO (40 ND, 3 HC). Agreement between VR-based and clinical assessment was moderate (κ = 0.42; 92.5

Neurological disorders are often difficult to diagnose and monitor, particularly in the early stages when symptoms may be subtle or nonspecific. Because the visual system engages a large portion of the cerebral cortex and relies on well-defined neural pathways, it offers a unique and accessible window into brain function. In this context, the concepts of oculomics and oculometrics have gained increasing attention. Oculomics refers to the study of systemic and neurological diseases through ocular biomarkers, while oculometrics involves the computational quantification of eye and pupil parameters. Together, these approaches provide noninvasive, objective, and reproducible methods to assess neurological function, with strong potential to improve diagnostic precision, monitor disease progression, and guide individualized care. This review synthesizes recent advances in ocular motor and pupillary biomarkers in three major neurological conditions: multiple sclerosis, Parkinson's disease, and Alzheimer's disease. In multiple sclerosis, early ocular motor disturbances such as internuclear ophthalmoplegia, saccadic dysmetria, and impaired smooth pursuit are frequently observed and may reflect brainstem and cerebellar involvement. Relative afferent pupillary defect, objectively measured with pupillometry, is a strong indicator of optic neuritis. In Parkinson's disease, impaired saccadic initiation, hypometric saccades, and convergence abnormalities reflect basal ganglia dysfunction, while pupil irregularities suggest underlying autonomic imbalance. In Alzheimer's disease, impairments in saccades, smooth pursuit, fixation instability, and the pupillary light reflex have been associated with early cortical and brainstem pathology, reflecting deficits in attention, executive control, and cholinergic function. We also discuss the integration of eye-tracking data with neuroimaging and electrophysiology biomarkers to support multimodal diagnostic frameworks with the potential to improve diagnostic accuracy and disease monitoring. In addition, we highlight how recent technological developments in virtual reality-based eye-tracking could offer immersive, standardized testing conditions to enable scalable implementation of oculometric assessments in clinical practice. As the fields of oculomics and oculometrics continue to evolve, these approaches hold promise to bridge the gap between research and clinical application. However, large-scale validation studies, standardized protocols, and normative datasets are essential for broader clinical adoption. By embedding ocular motor and pupillary biomarkers into routine neurological assessments, clinicians may be able to detect disease earlier, differentiate between overlapping syndromes, and monitor therapeutic outcomes more effectively.

Cortical thickness measurements from MRI are increasingly used as biomarkers for neurodegenerative disease progression. However, variations in MRI acquisition parameters, such as inversion time (TI) and repetition time (TR), which are common in clinical settings, can compromise the reliability and sensitivity of these measurements. We fine-tuned a deep-learning-based segmentation tool (DL+DiReCT) to reduce its dependence to image contrast variations by training it on simulated MPRAGE images derived from quantitative relaxation maps. Fine-tuning markedly reduced contrast sensitivity, with the Pearson correlation coefficient decreasing from - 0.644 $$ -0.644 $$ to 0.094 $$ 0.094 $$ . Evaluation on a synthetic atrophy dataset demonstrated that our model accurately replicated atrophy trends with minimal underestimation, outperforming FreeSurfer and SynthSeg. When applied to a dataset of relapsing-remitting multiple sclerosis (RRMS) patients, the fine-tuned model showed a substantial reduction in contrast sensitivity and maintained stable performance after controlling for covariates such as age, sex, field strength, and Expanded Disability Status Scale (EDSS) score. Overall, the proposed approach achieves robust contrast invariance without sacrificing sensitivity to cortical atrophy, offering a practical improvement for longitudinal and multi-center clinical studies.

Background While data on multiple sclerosis (MS) prevalence are widely available, incidence data, crucial for understanding disease onset dynamics, are less frequently reported. This study aims to estimate MS incidence in Switzerland from 2017 to 2021 and characterize persons newly diagnosed with MS. Methods The Swiss MS Registry is an ongoing longitudinal study in Switzerland. It has established an innovative methodology to assess MS epidemiology by integrating SMSR data with administrative data on reimbursement approvals for disease-modifying therapies and by use of a modified benchmark-multiplier method. We calculated overall, sex- and age-specific MS incidence (2017-2021) and described the demographic and clinical characteristics of newly diagnosed individuals. Results The estimated annual MS incidence in Switzerland was 663 [95% simulation interval 617-711] cases, corresponding to an annual MS incidence of 7.7 [7.1-8.3] per 100,000 inhabitants. This equates to approximately two new MS diagnoses per day. Incidence remained stable throughout the study period. Peak incidence occurred in the 30 to 35-year age group for women and 35 to 40-year age group for men. Most individuals were diagnosed with relapsing-remitting MS (94.0%). Following diagnosis, 56.9% received oral therapy, 27.7% infusion therapy, and 15.4% injectable therapy. Conclusion Our study found stable MS incidence over five years, consistently higher in women than men. Most diagnoses occurred between ages 20-40, though new diagnoses among middle-aged individuals appear to be increasingly common.

N-acetyl-L-leucine (NALL), an acetylated derivative of the amino acid leucine, has been shown to reduce neuronal cell death and neuroinflammation in murine models. Its beneficial effects in patients with the lysosomal storage disease Niemann-Pick Type C have led to recent FDA and EMA approval. However, neuroprotective effects of NALL remain to be further elucidated. In this study, we investigate and characterize the neuroprotective effects of NALL. To this end, we used human induced primary neurons (hiPNs), which were generated from induced pluripotent stem cells derived from reprogrammed renal proximal tubule epithelial cells obtained from either healthy controls (HC) or individuals with relapsing-remitting multiple sclerosis (MS). We demonstrated that NALL exhibits neuroprotective properties in MS- and HC-derived hiPNs subjected to acute damage induced by the microtubule-destabilizing agent nocodazole determined by neurite length. This effect can be blocked by inhibition of transporter-specific NALL uptake. HiPNs from MS donors treated with NALL expressed higher levels of glutamate cysteine ligase (GCL), the ratelimiting enzyme in glutathione synthesis. MS-specific cells are more susceptible to stress induced by the protein kinase inhibitor staurosporine, whereas in HC-specific cells only, NALL is able to modulate this stress induction. In summary, we demonstrate differential responses to induced stress in MS- and HC-specific neurons and the capacity of NALL to modulate neurite outgrowth and stress mechanisms in this cell culture system. NALL may represent an interesting additive treatment for MS or other diseases associated with oxidative stress and neurodegeneration if further substantiated.

The choroid plexus (CP), a key regulator of cerebrospinal fluid production and immune cell trafficking, is increasingly recognized as a potential magnetic resonance imaging (MRI) biomarker in multiple sclerosis (MS). Serum neurofilament light chain (sNfL) serves as a sensitive blood-based indicator of neuroaxonal injury. We investigate the prognostic value of CP volume and its longitudinal change for neurodegeneration, defined by sNfL levels and disability progression. In a prospective, multicenter, longitudinal study, 891 people with MS undergo high-resolution 3T MRI, sNfL measurement, and clinical assessment over 6 years; 434 meet criteria for inclusion. CP volume correlates with sNfL after adjustment for demographic, clinical, treatment, and imaging variables. In a 2-year MRI subcohort (n = 209), CP enlargement likewise associates with sNfL. High CP volume confers a 1.8-fold increased risk of disability worsening and a 2.7-fold increased risk of progression independent of relapse activity. These findings identify CP imaging as a promising non-invasive biomarker of neuroaxonal loss and relapse-free progression.

Introduction Following the coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), post-COVID-19 syndrome (PCS) has emerged as a major health concern, affecting approximately 3.0-11.7% of infected individuals. While neuropsychiatric symptoms such as fatigue, pain, and sleep disturbances dominate the clinical picture, recent evidence suggests a pivotal role for immune dysregulation in the pathophysiology of PCS. This study aimed to analyze the clinical course and laboratory features of patients with PCS. Methods In this retrospective single-center study, 74 patients with PCS were evaluated between November 2020 and June 2022 at the Department of Neurology, University Hospital Bern. Clinical assessments included standardized questionnaires [Fatigue Severity Scale (FSS), Fatigue Scale of Motor and Cognitive Function (FSMC), Beck Depression Index II (BDI-II), Epworth Sleepiness Scale (ESS)] and laboratory investigations at two timepoints. Particular focus was placed on immune-related biomarkers, including ferritin and antinuclear antibodies (ANA). Results Fatigue, sleep disturbances, and pain were the most frequently reported symptoms. At follow-up (a mean duration of 14.4 weeks after the first consultation), both symptom burden and severity scores decreased significantly. Nevertheless, 82.4% of patients continued to report fatigue. Elevated ferritin levels were found in 78.1% of tested patients (mean: 306.9 µg/l), and ANA titres ≥1:160 were observed in 40% of patients without known autoimmune disease. At follow-up, 80% of patients with initially elevated ferritin and 68% with elevated ANA remained above normal thresholds, suggesting ongoing immune activation. Conclusions While symptom burden declined over time, most PCS patients continued to experience clinically relevant symptoms. The persistence of elevated ferritin and ANA levels underscores possible immune dysregulation and highlights the potential of these biomarkers in characterizing PCS. These findings underline the need for further studies investigating their diagnostic and therapeutic relevance.

BACKGROUND:Glatiramer acetate (GA) is a well-tolerated treatment for multiple sclerosis (MS) and comparable in its efficacy to high-dose interferon beta (IFN). As a lack of validated treatment response biomarkers for MS hampers progress in personalised treatment, the study goal was to search for biomarkers of a successful treatment response utilising the known observation of T-cell expansions after GA treatment. METHODS:T-cell receptor beta chain (TRB) sequencing was performed in 3021 patients with MS: a discovery cohort of 1627 patients with MS, 204 of whom had previously been treated with GA, and then validated in 1394 patients with MS, 424 of whom had previously been treated with GA. Clinical data from 1987 patients with MS treated with GA or IFN and available HLA information from the NationMS, ACP, EPIC, BIONAT, and CombiRx trial cohorts were used for a subsequent analysis. FINDINGS:Common GA-associated TRB expansions were exclusively detected in HLA-A∗03:01 or in HLA-DRB1∗15:01 backgrounds, within CD8+ effector- or CD4+ central-memory T cells. Both sets of common sequences clonally expanded after GA treatment in a first validation cohort and predicted GA exposure in two further validation cohorts. To evaluate whether restriction of public TRBs to only two HLA alleles is also associated with GA's clinical efficacy, we analysed five cohorts of patients with MS for a potential benefit of the two HLAs concerning the GA response compared to IFN. We consistently found positive interactions with HLA-A∗03:01. This included a relative reduction in relapse risk compared to IFN in HLA-A∗03:01 carriers of 33% (CombiRx: GA + IFN arm: HR 0.67 [95% CI: 0.47-0.96], p = 0.0269) and 34% (CombiRx: GA arm: HR 0.66 [95% CI: 0.45-0.98], p = 0.0377), and in risk to first relapse of 63% (NationMS: HR 0.37 [95% CI: 0.16-0.88], p = 0.0246), but no positive association with DRB1∗15:01. INTERPRETATION:HLA-A∗03:01 carrying patients with MS specifically benefit from GA treatment and GA significantly outperforms IFN in these patients. Therefore, determining HLA-A∗03:01 status before choosing a platform treatment for MS, would allow for a personalised treatment decision between GA and IFN. FUNDING:German Research Foundation, National Institutes of Health, National Multiple Sclerosis Society, Valhalla Foundation, Westridge Foundation, Mayer Foundation, German Federal Ministry of Education and Research.

Internuclear ophthalmoplegia (INO) is an eye movement disorder that occurs in approximately one-third of people with multiple sclerosis (MS). We used a novel, head-mounted, virtual reality (VR)-based oculography device (PRET (TM), machineMD AG, Switzerland) to objectively measure oculomotor symptoms reported by a patient with MS to aid in the clinical diagnosis confirmation of INO. The patient's symptoms, primarily diplopia, were managed successfully with 4-aminopyridine. The VR-based eye-tracking tool documented the subjective improvement in oculomotor function tracking treatment effects.

Background: Different definitions of disability progression by Expanded Disability Status Scale (EDSS) may influence frequency and/or time to event. Methods: In this multicenter cohort study, we included PPMS patients with follow-up >= 24 months and >= 3 available EDSS scores overall (>= 1 per year). We applied 672 definitions of disability progression including different minimal EDSS increase, required confirmation and fixed/roving-baseline score. Results: We analyzed follow-up periods from 131 PPMS patients (median age at baseline 53.0 years [45.0 - 63.0], 51.9 % female, median follow-up 3.9 years [2.6 - 6.0], median baseline EDSS 4.0 [2.5 - 6.0]). The most sensitive definition of a progression event was an unconfirmed increase of >= 0.5 points with a roving baseline (81.8 % event rate). The least sensitive definition was an increase of >= 1.0 points with a fixed baseline, minimal distance to reference 48 weeks, and confirmed at >= 48 weeks (28.4 % event rate). Comparing roving vs. fixed baseline over all cutoffs and confirmation definitions, average time to progression was 227 days shorter applying the roving baseline (1405 days [550 - 2653] vs. 1632 days [760 - 2653]). Conclusions: Different definitions of disability progression result in significantly differing rates of disability progression, which may influence study results and create confusion in clinical practice.

Background Persistence to B cell depleting therapies (BCDT) like ocrelizumab and rituximab may be higher compared with other disease-modifying therapies (DMT) in multiple sclerosis (MS). Clinical trials directly comparing these treatments are lacking. Objective To compare the risk of treatment discontinuation, relapse, and confirmed disability worsening in patients starting BCDT vs other DMT within real-world settings. Methods In a longitudinal cohort study, patients with relapsing MS starting BCDT (ocrelizumab/rituximab, n = 269) after enrolment into the Swiss MS Cohort (SMSC) were evaluated for treatment discontinuation, occurrence of relapses, and disability worsening in comparison with platform ( n = 57) and oral ( n = 454) DMT, or natalizumab ( n = 73) using Cox regression with double robust adjustment for baseline covariates. Results Patients starting BCDT were less likely to discontinue treatment than all other DMT combined (HR = 0.26, 95% CI = 0.18–0.36, p < .01), oral DMT (HR = 0.28, 95% CI = 0.20–0.39, p < .01) and natalizumab (HR = 0.35, 95% CI = 0.21–0.58, p < .01). BCDT were associated with lower risk of relapses as compared to oral DMT HR = 0.59, 95% CI = 0.39–0.88, p < .01), but not to natalizumab (HR = 0.90, 95% CI = 0.45–1.82, p = .778). Disability worsening was not significantly different between treatment groups. Conclusion We provide real-world evidence for patients being more persistent to BCDT than to other treatments, and better clinical outcomes may partly explain this.

Background:With the development of highly effective disease-modifying treatments, vaccinations are becoming increasingly important in people with neurological autoimmune diseases. However, questions regarding the safety and efficacy of vaccinations under immunotherapy remain. Objective:To provide recommendations on types and timing of vaccinations for people with neuroimmunological diseases under different immunotherapies. Design:Our study presents a German evidence-based expert consensus on vaccination under immunotherapies in neurological autoimmune diseases. Methods:Based on literature research, a consortium of experts evaluated the quality of evidence, integrated clinical experience, and responded to a questionnaire determining an agreement (>75%) on statements concerning vaccination upon immune therapies in neuroimmunological diseases. Results:The specific humoral and cellular response to vaccination can be compromised under alemtuzumab, azathioprine, cladribine, cyclophosphamide, CD19/CD20 antibodies (inebilizumab, ocrelizumab, ofatumumab, rituximab, ublituximab), dimethyl fumarate/diroximel fumarate, FcRn inhibitors (efgartigimod, rozanolixizumab), complement C5 inhibitors (eculizumab, ravulizumab, zilucoplan), interleukin-6 receptor antibodies (tocilizumab, satralizumab), intravenous immunoglobulins, long-term steroid administration, methotrexate, mitoxantrone, mycophenolate mofetil, tacrolimus, teriflunomide, tumor necrosis factor-α blockers, and sphingosine-1-phosphate receptor modulators (fingolimod, ozanimod, ponesimod, siponimod), as well as after autologous stem cell transplantation. The lymphocyte count can have an influence here. Overall, it is generally advisable to complete vaccination before starting immunotherapy. However, in the case of an active inflammatory disease course with possible irreversible neurological deficits, a delay in therapy initiation until immunization has been completed cannot be justified. The application of live vaccines is contraindicated for most therapies and is only recommended after a strict risk-benefit assessment. Conclusion:Vaccinations are necessary for individuals on immunotherapy to reduce the risk of infections and the associated risk of worsening neurological autoimmune diseases. However, the humoral and cellular vaccination response may be impaired under immunotherapy necessitating close monitoring. Here, we provide applicable recommendations to optimize immunization for individuals receiving immunotherapy due to a neurological autoimmune disease.

BACKGROUND:Serum neurofilament light (sNfL) chain levels, a sensitive measure of disease activity in multiple sclerosis (MS), are increasingly considered for individual therapy optimization yet without consensus on their use for clinical application. OBJECTIVE:We here propose treatment decision algorithms incorporating sNfL levels to adapt disease-modifying therapies (DMTs). METHODS:We conducted a modified Delphi study to reach consensus on algorithms using sNfL within typical clinical scenarios. sNfL levels were defined as "high" (>90th percentile) vs "normal" (<80th percentile), based on normative values of control persons. In three rounds, 10 international and 18 Swiss MS experts, and 3 patient consultants rated their agreement on treatment algorithms. Consensus thresholds were defined as moderate (50%-79%), broad (80%-94%), strong (≥95%), and full (100%). RESULTS:The Delphi provided 9 escalation algorithms (e.g. initiating treatment based on high sNfL), 11 horizontal switch (e.g. switching natalizumab to another high-efficacy DMT based on high sNfL), and 3 de-escalation (e.g. stopping DMT or extending intervals in B-cell depleting therapies). CONCLUSION:The consensus reached on typical clinical scenarios provides the basis for using sNfL to inform treatment decisions in a randomized pragmatic trial, an important step to gather robust evidence for using sNfL to inform personalized treatment decisions in clinical practice.