Florian Deisenhammer

Brain volume changes and infratentorial involvement are key predictors of disability in multiple sclerosis (MS) and can be assessed using magnetic resonance imaging (MRI) planimetry. Although MRI planimetry is less susceptible to methodological and patient‐related confounders than volumetry, it currently depends on manual measurements by unblinded experts, an approach that is time‐consuming and vulnerable to bias. In this study, we present a fully automated deep learning framework for deriving brainstem planimetric measurements from MRI. The pipeline integrates an automated midsagittal plane (MSP) detection algorithm with a convolutional neural network trained to perform the segmentations required for planimetry. The automated method shows strong agreement with manual measurements and remains robust across scanners and acquisition protocols. These findings suggest that the proposed framework enables reliable, reproducible, and scalable MRI planimetry, supporting objective assessment of disease progression and treatment response in patients with MS.

Background and objectives Biologically informative markers like glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) may help predict confirmed disability worsening (CDW) in multiple sclerosis (MS). However, data on the prognostic value of their blood concentrations in progressive MS (PMS) are limited, and there are substantial discrepancies in the published literature. This international collaboration uses individual participant data to define the prognostic value of serum GFAP and NfL in people with PMS (pwPMS). Methods Data were collected from BioMS-eu network centres and collaborating cohorts. pwPMS with primary progressive MS (PPMS) or secondary progressive MS (SPMS) with at least one GFAP value and at least three follow-up expanded disability status scale (EDSS) scores were included. The prognostic value of serum GFAP and NfL age- and sex-adjusted Z-scores for future CDW was evaluated using Cox regression models, accounting for sex, age, baseline disease duration and EDSS, and dominant treatment during follow-up. Results 1058 participants and 7530 encounters were included (median age 53 years (IQR: 44 to 59), 57% female, follow-up 4.6 years (2.9 to 8.4)) with median baseline GFAP of 0.74 (−0.10 to 1.55) and NfL of 0.64 (−0.36 to 1.51). 723 CDW events were recorded. Each GFAP Z-score increase was associated with ~10% higher CDW risk (adjusted HR (aHR) 1.107 (1.001 to 1.225), p=0.049). Results were mainly driven by SPMS participants (n=613, aHR 1.242 (1.073 to 1.438), p=0.004). Higher NfL Z-scores predicted CDW only in PPMS participants (1.236 (1.092 to 1.399), p=0.001). Conclusions GFAP was a prognostic indicator for future CDW in pwPMS, especially in pwSPMS. On the other hand, NfL was predictive of CDW only in pwPPMS.

Background: Reduced bone mass and increased osteoporosis risk are common in people with multiple sclerosis (pwMS). The aim of the study was to identify risk factors for short‐term bone loss in MS.Methods: This prospective study included 139 pwMS (ages 18–65). Baseline data included demographics, body‐mass index, physical activity, smoking, menopause status, 25‐hydroxy vitamin D levels, and history of glucocorticoid use. Bone mineral density (BMD) was measured at baseline and after 2 years using dual‐energy X‐ray absorptiometry (DXA) for the lumbar spine and hip. Disability worsening was assessed by the Expanded Disability Status Scale (EDSS). Additionally, a literature review was conducted on longitudinal data regarding BMD in MS.Results: Over the 2‐year follow‐up period, significant BMD loss was observed in the hip (baseline g/cm²: median 0.898; IQR 0.808–1.014; 2‐year follow‐up: 0.882; 0.784–1.01; p < 0.001), but not in the lumbar spine. Overall, 101 (73.1%) experienced hip BMD loss, with a median decrease of 3.5%. Patients with disability worsening had an approximately 7‐times higher risk of bone loss compared to those without disability worsening (p = 0.013). PwMS with fractures during the follow‐up period had significantly lower hip BMD (0.760, 0.546–0.890 vs. 0.909, 0.828–1.015; p = 0.024), a higher EDSS score (4.4, 2.8–5.8 vs. 2.0, 1.0–4.0 vs. p = 0.026), and were older (59, 46–62 vs. 47, 37–54; p = 0.030) compared to those without fractures.Conclusion: Disability worsening was identified as a risk factor for BMD loss. These findings underscore the need for active monitoring of pwMS with disability worsening to prevent bone loss and, thus, to reduce fracture risk.

Introduction: Prognostication after subarachnoid hemorrhage (SAH) is essential to guide clinical management and improve patient care.Objective: To investigate whether decay rates of cerebrospinal fluid (CSF) red blood cells (RBC) and total protein (TP) after SAH predict functional outcome at 3 months.Methods: Patients with SAH treated at the Neurological Intensive Care Unit Innsbruck with a first CSF sample (CSFfirst) within 72 h after admission and at least one subsequent sample were eligible for inclusion. Decay rates of RBC and TP were measured between CSFfirst and each subsequent measurement (Weeks 1–3). Modified Rankin Scale scores at 3 months ≤ 2 were defined as good functional outcomes.Results: A total of 97 patients with a median age of 61 years [25th; 75th percentile: 52;71] and a median Hunt and Hess score of 4 [3;5] were included. Daily RBC decay rates decreased over time, while daily TP decay rates were highest in Week 1, showed a nadir in Week 2, and increased again in Week 3. In multivariable analysis, higher RBC (adjusted odds ratio (adjOR): 1.13, 95% confidence interval (95% CI): 1.02–1.26, p = 0.025) and TP (adjOR: 1.01, 95% CI: 1.00–1.01, p = 0.031) decay rates at Week 3 predicted a good functional outcome at Month 3. RBC and TP decreasing below 1180 cells/μL and 127.5 mg/dL, respectively, at Week 3 were associated with good functional outcome at Month 3 (adjOR: 11.04, 95% CI: 3.21–38.02, p < 0.001 and 6.03, 1.68–21.67, p = 0.006).Conclusions: Decay rates of CSF RBC and TP after SAH are associated with functional outcomes.

Background: Recognizing familiar faces and identifying emotions through facial expressions are essential for social functioning. This study aimed to examine whether people with relapsing-remitting multiple sclerosis (PwMS) differ from healthy control individuals (HC) in their performance on different tasks related to facial emotion processing. Methods: In a cross-sectional controlled study, 30 PwMS and 35 HC completed a baseline neuropsychological evaluation and experimental tasks assessing visual exploration of facial stimuli through eye tracking, facial emotion recognition, and facial memory recognition. The facial stimuli displayed either a neutral expression or an emotion (happiness, fear, or disgust). Results: PwMS and HC performed comparably in facial emotion recognition. In facial memory recognition, HC were significantly more accurate in recognizing previously seen fearful faces compared to neutral faces (Wilcoxon test, Z = -2.26, P = 0.024), demonstrating emotional enhancement of memory. In contrast, PwMS did not exhibit a memory advantage for fearful faces over neutral faces (P > 0.05). Groups also differed in the eye-tracking task. In all but one condition (disgust), PwMS showed a significantly greater tendency to explore the eye area rather than the mouth area compared to HC. Conclusions: Changes in visual exploration and a lack of emotional enhancement of memory are observed in PwMS, who otherwise demonstrate intact facial emotion recognition. These results suggest altered emotion-cognition interactions in PwMS. Early detection of subtle changes and targeted interventions may help prevent future debilitating impairments in social functioning.

Introduction: Lumbar puncture (LP) is a routine clinical procedure and, in some cases, is repeatedly performed for diagnostic or therapeutic reasons. The impact of repeated LP on cerebrospinal fluid (CSF) findings is not clear.Objective: To investigate whether repeated LP is associated with reactive pleocytosis and disruption of blood–CSF barrier function and to determine the role of interval between repeated LP.Methods: Patients with non‐inflammatory neurological disease (NIND) and at least two consecutive LP were included. Longitudinal changes in CSF white blood cell count (WBC), CSF total protein (TP), and CSF/serum albumin quotient (Q_alb) were assessed depending on the time interval between the LP.Results: A total of 73 patients with a median age of 35 years (25th–75th percentile: 25–45) and a female predominance of 75% had second LP after 6 (3–19) days. Twenty (27%) patients developed pleocytosis with an increase of WBC count to 8/μl (6–15) with a maximum of 30/μl. Patients with pleocytosis had the follow‐up LP significantly earlier than patients without pleocytosis, 3.5 (3–7) versus 7 (3–28) days. The majority of patients (90%) with CSF pleocytosis had the second LP within 10 days. Further repeated LP in a subgroup of patients revealed similar findings. CSF TP and Q_alb slightly increased in patients with pleocytosis.Conclusion: Mild “reactive” CSF pleocytosis occurred in approximately one‐third of patients after repeated LP, mostly when performed within 10 days.

Objective: To investigate the impact of transition interval length when switching from natalizumab (NTZ) to anti‐CD20 monoclonal antibodies (antiCD20) on recurrent disease activity and safety in relapsing multiple sclerosis (RMS).Methods: Aggregating data from 8 MS centres in Austria, Switzerland, and Germany, we included RMS patients who (i) continuously received NTZ for ≥3 months, (ii) were switched to antiCD20, and (iii) had ≥12 months follow‐up after switch. The primary endpoint was occurrence of relapse after switch, secondary endpoints included severe infections (CTCAE grade ≥3).Results: Overall, 139 RMS patients were included (70.5% females, mean age at switch 38.8 years [SD 9.7], mean disease duration at switch 11.3 years [SD 6.2], median duration on NTZ 4.4 years [range: 0.3–16.4], median transition interval 58 days [0–180]). Relapse occurred in 18 patients (12.9%) after NTZ discontinuation. Of those, 11 (61.1%) patients relapsed during the transition interval. No patient with a transition interval below 30 days experienced a relapse, compared to 11.1% and 16.1% with transition intervals of 30–44 days and ≥ 45 days, respectively. In multivariable Cox regression, a transition interval ≥ 45 days predicted a 4.73‐fold increased risk of relapse. Over approximately 4 years of follow‐up, six severe infections were reported without any noticeable effect of transition interval length. No PML occurred.Conclusions: Switching from NTZ to antiCD20 is generally both effective and safe. Keeping the transition interval below 30 days provides the optimal balance between preventing recurrent disease activity and ensuring safety.

Background: Neurological symptoms are common in acute mountain sickness (AMS); however, the extent of neuroaxonal damage remains unclear. Neurofilament light chain (NfL) is an established blood biomarker for neuroaxonal damage.Objective: To investigate whether plasma (p) NfL levels increase after simulated altitude exposure, correlate with the occurrence of AMS, and might be mitigated by preacclimatization.Methods: Healthy subjects were exposed to simulated high altitude (4500 m) by the use of a normobaric hypoxic chamber at the University of Innsbruck two times, that is, within Cycle 1 (C1) over 12 h, and within Cycle 2 (C2) for another 12 h but with a random assignment to prior acclimatization or sham acclimatization. Before each cycle (measurement [M] 1 and 3) and after each cycle (M2 and M4), clinical data (arterial oxygen saturation [SaO₂], heart rate, and Lake Louise AMS score [LLS]) and plasma samples were collected. pNfL was measured using single‐molecule array (Simoa) technique.Results: pNfL levels did not significantly change within each study cycle, but increased over the total study period (M1: 4.57 [3.34–6.39], M2: 4.58 [3.74–6.0], M3: 5.64, and M4: 6.53 [4.65–7.92] pg/mL, p < 0.001). Subjects suffering from AMS during the study procedures showed higher pNfL levels at M4 (6.80 [6.19–8.13] vs. 5.75 [4.17–7.35], p = 0.048), a higher total pNfL increase (2.88 [1.21–3.48] vs. 0.91 [0.53–1.48], p = 0.022) compared to subjects without AMS. An effect of preacclimatization on pNfL levels could not be observed.Conclusions: pNfL increases alongside exposure to simulated altitude and is associated with AMS.