Abstracts from The 16 th Symposium of the International Neurotrauma Society September 2–5,2024 Cambridge,United Kingdom

O001 Transparency, Rigor, and Reproducibility in Neurotrauma Research: Perspective from the Journal of Neurotrauma Editor-in-Chief

Dr. David Brody ¹

¹Journal Of Neurotrauma, Bethesda, United States

Introduction: There has been a ‘crisis of reproducibility’ in the biomedical sciences, with relatively low rates of successful translation of research findings into improved clinical outcomes. Major funding bodies around the world have been making transparency, rigor and reproducibility a priority. Journal of Neurotrauma has recently implemented a requirement for transparency, rigor, and reproducibility statements accompanying each article.

Methodology: Review and discussion of the neurotrauma research community in this domain.

Results: Working definitions include the following: Transparency: Sharing research methods and findings in an open and clear fashion, both within the scientific community and with the wider public. The readers should understand exactly what you did, and exactly what you found. The methods should allow another group to reproduce the data collection and analyses. Rigor: The strict application of the scientific method to ensure unbiased and well-controlled experimental design, methodology, analysis, interpretation and reporting of results. Methods to minimize both conscious and unconscious bias are included. Reproducibility: If another research group performs the same or a very similar study, will they get the same or similar results?

Discussion points will include the following: 1) Why does transparency, rigor, and reproducibility matter? 2) What are the barriers to optimal transparency, rigor, and reproducibility? 3) What are some “Do’s” and “Don’ts”? 4) Examples of excellence in neurotrauma, 5) Other examples and lessons to be learned, 6) How can researchers, editors, and funding agencies do better? 7) The way forward.

Conclusions: Transparency, rigor, and reproducibility is important. Transparency, rigor, and reproducibility is not easy. Perfection is not possible, but each day we can try to do better than the day before. We owe it to our patients to do the best we can.

O002 Biomarker evidence of neurodegeneration in mid-life former elite rugby players in the UK ABHC cohort study

Dr Neil Graham ,^1,2 Dr Karl Zimmerman,^1,2 Ms Jessica Hain,^1,2 Ms Erin Rooney,^1,2 Dr Ying Lee,^1,2 Ms Martina Del Giovane,^1,2 Dr Thomas Parker,^1,2 Dr Maneesh Patel,³ Dr Elena Veleva,⁴ Dr Amanda Heslegrave,^4,5 Prof Henrik Zetterberg, Mr Daniel Friedland,¹ Dr Richard Sylvester,⁴ Prof David Sharp^1,2

¹Imperial College London, London, United Kingdom, ²UK DRI Centre for Care Research and Technology, London, United Kingdom, ³Imperial College Healthcare NHS Trust, London, United Kingdom, ⁴UCL Queen Square Institute of Neurology, London, United Kingdom, ⁵UK DRI at UCL, London, United Kingdom

Introduction: Traumatic brain injury (TBI) is a dementia risk factor. Repetitive head impacts are associated with neurodegenerative pathologies including chronic traumatic encephalopathy (CTE) and Alzheimer’s disease. Biomarker advances make the detection of many degenerative pathologies possible in midlife. In the Advanced Brain Health Clinic (ABHC) cohort we related trauma exposure and clinical symptoms to neurodegeneration biomarkers.

Methodology: Ex-professional rugby players underwent quantification of plasma neurofilament light (NfL), glial fibrillary acid protein (GFAP), amyloid β-42/40 ratio and phospho-tau 217 (p-tau217). Neuropsychology and MRI were performed (volumetric/diffusion tensor imaging, DTI). Unexposed controls were assessed. Biomarker concentrations/DTI metrics were defined as abnormal/normal (95th centile cut-offs) and regression models used for group comparisons/outcome associations.

Results: 200 ex-players were assessed (age 44.3, SD 7.4 years, 90.5% male). Mean career duration was 10.5 years (SD 4.1). 24 (21%) had possible traumatic encephalopathy syndrome (TES) but no players had dementia. P-tau217 was 17.6% higher in ex-players (95%CI 3.7–33.3, Pcorr=0.047), being abnormally raised in 46 (23%) and associated with increased odds of TES (95%CI 1.1–7.1, P=0.027). 18 (9.0%) had raised NfL. These players had more depressive (ß=1.4, 95%CI 0.56–2.25, Pcorr=0.003) and anxiety symptoms (ß=1.68, 95%CI 0.81–2.57, Pcorr<0.001). There was reduced frontal/anterior cingulate volume in ex-players, and hippocampal volume reduced with greater years of play. Ex-players with TES had reduced frontal/cingulate VBM region volumes (95%CI 10.0–207.3, P=0.031). There were no group-level diffusion abnormalities but individual-level fractional anisotropy reductions were present in the callosum in 9 (4.6%), whole white matter in 5 (2.5%) and corticospinal tract in 3 (1.5%) ex-players.

Conclusions: Elite rugby participation is associated with raised p-tau217 and reduced frontal/cingulate volume in mid-life ex-players. Hippocampal volume reductions related to longer careers and elevated NfL to psychiatric symptoms. The results suggest neurodegeneration and possible amyloid pathology in some ex-players, although the nature of any neuropathology requires further validation.

O003 Current use of PRx and derived cerebral perfusion pressure targets in traumatic brain injury: shared clinical protocols

Dr Erta Beqiri ,¹ Dr Marcel Aries,^2,3 Dr Soojin Park,⁴ Dr Andrea Lavinio,⁵ Dr Celeste Dias,⁶ Dr Cornelia Hoedemaekers,⁷ Dr Bart Depreitere,⁸ Dr Raimund Helbok,⁹ Dr Miriam Weiss,^10,11 Dr Gerrit Schubert,^10,11 Dr Matthew Kirschen,¹² Dr Brian Appavu,¹³ Dr Peter Smielewski,¹ CLINICCA collaborators

¹Brain Physics Laboratory, Department of Clinical Neuroscience, Division of Neurosurgery, University of Cambridge, Cambridge, United Kingdom, ²Department of Intensive Care Medicine, University Maastricht, Maastricht, The Netherlands, ³School for Mental Health and Neuroscience (MHeNS), University Maastricht, Maastricht, The Netherlands, ⁴Departments of Neurology and Biomedical Informatics, Columbia University Vagelos College of Physicians and Surgeons, NewYork-Presbyterian Hospital, New York, USA, ⁵Cambridge University Hospitals NHS Foundation Trust, Cambridge, United Kingdom, ⁶Centro Hospitalar Sao Joao, Faculty of Medicine, Porto, Portugal, ⁷Department of Intensive Care, Radboud University Medical Center, Nijmegen, The Netherlands, ⁸Department of Neurosurgery, University Hospitals Leuven, Leuven, Belgium, ⁹Department of Neurology, Kepler University Hospital, Johannes Kepler University Linz,, Linz, Austria, ¹⁰Department of Neurosurgery, Cantonal Hospital Aarau, Aarau, Switzerland, ¹¹Dept of Neurosurgery, University Hospital Aachen, RWTH Aachen University, Germany, ¹²Departments of Anesthesiology and Critical Care Medicine, Neurology, and Pediatrics, Children’s Hospital of Philadelphia, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, United States of America, ¹³Department of Child Health and Neurology, University of Arizona College of Medicine Phoenix

Introduction: To characterize variability in the use of continuously monitored cerebral autoregulation status (CCA) in management of critically ill patients, we surveyed 44 clinicians working in the critical care community worldwide[1]. Half of the respondents indicated to use a CCA index in clinical decision making, with 39% using it as part of a local clinical protocol. We aim to share these protocols with the community.

Methodology: We focused on protocols that describe applying the intracranial pressure (ICP) derived pressure reactivity index (PRx) and cerebral perfusion pressure (CPP) targets in adult and paediatric traumatic brain injury (TBI) patients. We drafted a protocol based on the results of the survey, covering 1) use of PRx as a trigger for clinical review, 2) use of knowledge of cerebrovascular reactivity to manipulate ICP in case of intracranial hypertension, 3) use of derived CPP targets to ‘optimise’ brain physiology, 4) use of derived CPP targets to decrease the burden of therapy, 5) use of PRx and derived CPP targets in a stepwise approach within a multimodality protocol. Subsequently, we asked feedback from the respondents that use the index in clinical decision making.

Results: We received feedback from 8 centres with details regarding practical actions and examples of implementation, advantages, and reservations. Not all respondents implement all five protocol points. The agreement between protocols was higher in the use of knowledge of vascular reactivity for triggering a clinical review or for manipulating ICP. The higher variability was related to the use of derived CPP targets and type of other monitoring modalities considered for the patient assessment.

Conclusions: Sharing currently used protocols for CCA implementation in the clinical practice can enable an open discussion on advantages and disadvantages of currently used CCA protocols, facilitating autoregulation-guided therapy at a larger scale, and aiding future clinical trial design.

O004 Improving the precision of the Glasgow Outcome Scale-Extended: Evaluating a novel scoring approach from 2 weeks through 12 months post-injury in the TRACK-TBI sample

Lindsay Nelson ,¹ Brooke Magnus,² Steve Balsis,³ Nancy Temkin,⁴ Geoffrey Manley⁵

¹Medical College Of Wisconsin, Brookfield, United States, ²Boston College, Boston, USA, ³University of Massachusetts Lowell, Lowell, USA, ⁴University of Washington, Seattle, USA, ⁵University of California San Francisco, San Francisco, USA

Introduction: The limited precision of the Glasgow Outcome Scale Extended (GOSE)—an 8-point ordinal scale—may hinder the success of traumatic brain injury (TBI) clinical trials. We evaluated how a more refined approach to score the GOSE structured interview—previously developed at 3 months post-injury—extends to 2 weeks through 12 months and tested the hypothesis that novel GOSE scoring strengthens associations with other patient-reported outcomes (PROs).

Methodology: The Transforming Research and Clinical Knowledge in TBI (TRACK-TBI) study (N=1608 here) enrolled U.S. level 1 trauma center patients within 24 hours of injury and collected the following outcomes at 2 weeks and 3, 6, and 12 months: the GOSE interview and 4 PROs: Rivermead Post Concussion Symptoms Questionnaire (RPQ), Quality of Life After Brain Injury Overall Scale (QOLIBRI-OS), 18-item Brief Symptom Inventory (BSI-18), and Satisfaction with Life Scale (SWLS).

Analyses used item response theory (IRT) to model the dimension of TBI-related disability from the GOSE interview questions, examined the degree to which our GOSE-IRT model established at 3 months post-injury was displayed measurement invariance across other timepoints, estimated individuals’ GOSE-IRT disability scores, and evaluated the difference in correlations between traditional GOSE-Ordinal scores and GOSE-IRT scores with the 4 PROs.

Results: The GOSE-IRT model was not time-invariant, meaning the degree and manner in which responses about daily life (e.g., independence in the home, shopping, travel, work, social functioning, relationships) reflected the latent dimension of disability varied over time. GOSE-IRT scores derived separately at each timepoint were more granular than GOSE-Ordinal scores, were correlated in expectable ways with GOSE-Ordinal scores and PROs, and displayed stronger correlations with PROs than GOSE-Ordinal scores.

Conclusions: The findings support the dynamic, complex nature of disability ratings over the first year post-injury and the potential value of the GOSE-IRT scoring approach for TBI studies.

O005 Incremental prognostic performance of pupillary reactivity with GCS; validation of the GCS-Pupil scale in the prospective observational cohorts CENTER-TBI

MSc Rick Vreeburg ,^1,2 MSc Florian van Leeuwen,¹ Prof. dr. Geoffrey Manley,³ Prof. dr. Ewout Steyerberg,¹ dr. Paul Brennan,^4,5 Dr. Thomas van Essen,^1,2,6 Prof. dr. Andrew Maas,^7,8 Prof. dr. Wilco Peul,^1,2 Prof. dr. David Menon⁹

¹Leiden University Medical Center, The Hague, Netherlands, ²University Neurosurgical Center Holland, Leiden University Medical Center, Leiden & the Hague, Netherlands, ³University of California, San Francisco, United States, ⁴Royal Infirmary of Edinburgh, Edinburgh, United Kingdom, ⁵Centre for Clinical Brain Sciences, Edinburgh, United Kingdom, ⁶QEII Health Sciences Center and Dalhousie University, Halifax, Canada, ⁷Antwerp University Hospital, Edegem, Belgium, ⁸University of Antwerp, Antwerp, Belgium, ⁹University of Cambridge and Addenbrooke's Hospital, Cambridge, United Kingdom

Background: The Glasgow Coma Score (GCS) and pupillary reactivity are important prognostic factors in traumatic brain injury (TBI). Combining the GCS and pupil reactivity in the so-called GCS-P score has shown potential to increase information yield compared to using GCS alone while maintaining ease of use. The aim of this study is to analyze the comparative performance of GCS and GCS-P for predicting outcome after TBI in a representative, contemporaneous TBI cohort, and to compare their performance to a model including GCS and pupil reactivity as separate prognostic factors.

Methods: We analyzed patients enrolled between 2014 and 2018 in the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI). We used logistic regression to quantify the prognostic performances of GCS-P versus GCS according to Nagelkerke’s R2. Endpoints were mortality and unfavorable outcome (score < 5 on the Glasgow Outcome Scale-Extended score) at 6 months after injury. We estimated 95% confidence intervals with bootstrap resampling of pooled estimates in improvement of prognostic capability.

Results: Among the 4509 participants in CENTER-TBI, 3521 (78%) met the inclusion criteria. GCS as a linear score had a R2 of 23.9% for mortality and 29.3% for unfavorable outcome in the CENTER-TBI cohort. Pupils as a separate variable improved the CENTER-TBI R2 with 6.3% and 2.5% for mortality and unfavorable outcome, with around half the improvement captured in the GCS-P score (2.7% and 1.4%, respectively).

Conclusions: GCS-P has a stronger association with outcome after TBI than the GCS alone, and may merit evaluation as a clinically useful summary of injury severity. However, for prognostic models, inclusion of GCS and Pupils as separate scores is preferable to the use of a GCS-P summary score.

O006 The Impact of Barbiturate Therapy in Children with Severe TBI

Dr Andrew Appiah-Baiden ,¹ Dr Devon van Eck,¹ Dr Nqobile Thango,¹ Dr Thandani Mlambo,¹ Professor Nico Enslin,¹ Professor Ursula Rohlwink,¹ Professor Anthony Figaji¹

¹Red Cross War Memorial Children's Hospital, Cape Town, Cape Town, South Africa

Introduction: There are no clear guidelines for how to use sedation and second-tier therapies for increased intracranial pressure (ICP) in children with severe traumatic brain injury (TBI). Data are limited on the use of barbiturate therapy as a second-tier option for ICP control in children. We evaluated the impact of thiopentone on physiology in children with severe TBI and describe the patient outcomes.

Methodology: Retrospective cohort study of children <13 years of age with severe TBI who received thiopentone. Data were abstracted from a prospectively maintained database and where needed from patient clinical records. High frequency physiological data (ICM+, Cambridge University) were analyzed for the 12-hour period before the drug was initiated, and compared these to recordings for the duration that the drug was used.

Results: Data were analyzed from 70 children: most were male (67%), and most were road traffic accident victims (73%). The average time from admission to initiation of thiopentone therapy was 48 hours; the average treatment duration was 4.8 days. Thiopentone use was associated with a reduction in median ICP and median mean arterial pressure (MAP), and no change in CPP. Brain tissue oxygenation was slightly higher on thiopentone, but not significantly. On average, patients were extubated 5.9 days after cessation of thiopentone infusion; 21.4% received tracheostomies, and average duration of ICU stay was 15 days. Decompressive craniectomy (DC) was used in 12.6% of patients. The mortality rate was 15.6%.

Conclusion: This is the largest study to analyze barbiturate therapy in children with TBI. Thiopentone was useful in decreasing ICP, but with an associated decrease in MAP, although CPP remained similar. Despite this being a selected group of patients on second-tier therapies, the mortality rate was acceptable. Thiopentone use may avoid the surgical morbidity of DC, but at the expense of longer stays in ICU.

O007 Astrocyte/blood vessel unit injury following blast exposure: Histologic and ultrastructural alterations

Professor Daniel Perl ,¹ Doctoral Student Nicholas Breehl,¹ Sharon Juliano PhD, Michael Cranston PhD, Dara Diskstein PhD

¹Uniformed Services University, Bethesda, United States

Background: Astrocytes provide many vital brain functions, including water/ion homeostasis (particularly via Aquaporin-4 [AQP4]), and in the blood-brain barrier. Astrocyte endfeet (AEs), which envelope blood vessels (BVs), are critical in these processes. Dysmorphic astrocytes, with fragmented, beaded processes, have been observed at cortical grey-white interfaces in blast-exposed human brains.

Methods: Electron microscopy (EM) was carried out on AEs enveloping BVs at gray-white interfaces of frontal lobe specimens derived from five blast-exposed Service Members with astrocyte dysmorphology, and paired control specimens. BVs enveloped by AEs were identified, and BV and AE surface areas and diameters were compared.

Further, we performed fluorescent immunohistochemistry for astrocyte proteins GFAP, AQP4, and connexin-43 on brains from similarly exposed Service Members and controls.

Results: EM analysis revealed increased AE:BV surface area ratios (t(4)=2.796, p=0.049), reduced frequency of full ensheathment of BVs by AEs (t(4)=3.556, p=0.024), and a general increase in BV thickness in blast cases. AEs also showed frequent fragmentation and irregularities in thickness/contour.

Astrocyte immunoreactivity at the grey-white interface of blast cases was altered. Controls manifested two dominant populations, labeling as either GFAP+ or AQP4+ only, however the blast samples showed an additional astrocyte population that co-labeled of GFAP and AQP4. The GFAP+ beaded processes of the dysmorphic astrocytes demonstrated additional AQP4 staining, and co-labeling with phosphorylated connexin-43, indicating an inflammatory phenotype.

Conclusions: We report histologic and ultrastructural abnormalities in astrocytes in the frontal lobes of human blast-exposed brains. These findings suggest that blast may directly damage the astrocyte/blood vessel complex and thus have implications regarding its function.

Disclaimer: The information/content and/or conclusions do not necessarily represent the position or policy of, nor should any endorsement be inferred on the part of, USU, the DoD, the U.S. Government, or the Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc.

O008 Potentiating glymphatic drainage minimizes post-traumatic cerebral oedema

Dr. Rashad Hussain ,¹ Dr. Jeffrey Tithof,² Dr. Pia Weikop,³ Dr. Maiken Nedergaard^1,3

¹Department of Neurology/Center for Translational Neuromedicine, University of Rochester, Rochester New York, USA, ²Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN, USA, ³Center for Translational Neuroscience, University of Copenhagen, Copenhagen Denmark,

Traumatic brain injury (TBI) is a leading cause of death and disability among young adults, with its management hindered by various complex factors, particularly cerebral edema, which significantly increases mortality risks. Our research utilizing the “Hit and Run” mouse model of brain injury reveals that cerebral edema results from fluid entrapment within the brain, coinciding with/caused by the noradrenergic surges following TBI. We observed a marked reduction in the glymphatic/perivascular flow of cerebral spinal fluid (CSF), indicating disrupted waste clearance mechanisms. The administration of pan-adrenergic receptor antagonists, which target α1, α2, and β adrenergic receptors, proved effective in significantly reducing edema and enhancing neurological functions. Moreover, our studies demonstrate that TBI impairs CSF drainage through the cervical lymphatic system, characterized by decreased vessel contractility and reduced volume flow rate. Our findings advocate a novel therapeutic strategy whereby cerebral edema can be effectively managed through pan-adrenergic inhibition. This approach not only mitigates edema but the collection/analysis of cervical lymphatic fluid may also facilitate a better understanding of the severity of injury/biomarkers which is typically masked by the entrapment of fluid inside the brain.

O009 Drug identified with new astroglial wound healing assay promotes locomotor function after spinal cord injury

Cedric Geoffroy ¹

¹Texas A&m University, Bryan, United States

After spinal cord injury (SCI), glial cells become activated as part of a plethora of ongoing inflammatory processes, a cascade of events that includes the formation of a glial scar. The glial scar has been shown to perturb axon regeneration and release neurotoxic cytokines into the local environment, however, it can also play a beneficial role. Astrocytes are involved in the formation of the astroglial border which sequesters the inflammatory lesion, prevent its detrimental expansion, and reduce secondary tissue loss. Astrocytes can also provide a scaffold for regenerating axons. Here, we wanted to determine if increasing astrogliosis acutely and locally could reduce would reduce the lesion size and enhance locomotor recovery after SCI. We show that promoting astrogliosis by deletion of SOCS3, a negative regulator of STAT3, in primary astrocytes in vitro improves ‘wound’ closure in an artificial wound assay. This was consistent in vivo with acute SOCS3 deletion in spinal astrocytes at the site of a T8 SCI reducing the spread of the SCI lesion and improving functional recovery. This demonstrated there is a therapeutic interest to find drugs candidates that can promote this complex wound healing process acutely. To this end, we developed a scratch assay that incorporates primary adult mouse astrocytes and uses a novel scratch tool reducing variance and enhancing high-throughput capabilities. Our assay led us to identifying positive hits reducing the artificial wound in vitro. Several hits were tested in a preclinical SCI model and were found to reduce the size of lesion and promotes locomotor function . We demonstrate here the utility of our adult astrocytic scratch assay for the identification of new therapeutic options for acute SCI treatment.

O010 Imaging Findings in Acute TBI: a NINDS Common Data Element-Based Appraisal and Analysis of over 4,000 Admission Brain CT Scans from the CENTER-TBI Study

Dr. Dana Pisică,^3,4 Dr. Thijs Vande Vyvere ,¹⁰ Paul M. Parizel,^5,6,7 Andrew I.R Maas^8,9

¹Department of Radiology, Antwerp University Hospital, Antwerp, Belgium, ²Department of Molecular Imaging and Radiology (MIRA), Faculty of Medicine and Health Science, University of Antwerp, Antwerp, Belgium, ³Department of Neurosurgery, Erasmus MC - University Medical Center Rotterdam, Rotterdam, Netherlands, ⁴Department of Public Health, Erasmus MC - University Medical Center Rotterdam, Rotterdam, Rotterdam, Netherlands, ⁵Department of Radiology, Royal Perth Hospital, Perth, Australia, ⁶University of Western Australia, Perth, Australia, ⁷Western Australia National Imaging Facility (WA NIF) node, Australia, ⁸Department of Neurosurgery, Antwerp University Hospital, Antwerp, Belgium, ⁹Department of Translational Neuroscience, Faculty of Medicine and Health Science, University of Antwerp, Antwerp, Belgium, ¹⁰Antwerp University Hospital, Antwerp, Belgium

Background: In 2010, the National Institute of Neurological Disorders and Stroke (NINDS) created a set of common data elements (CDEs) to help standardize neuroimaging assessment and reporting in patients with traumatic brain injury (TBI). This represents an outside-in approach of inspecting and reporting individual findings, in tiers of increasing detail. Using over 4000 admission CTs from the CENTER-TBI study, reported using CDE methodology, we aim to: 1) provide a comprehensive overview of the neuroimaging case-mix of contemporary TBI patients in Europe; 2) offer a critical appraisal of the CDEs.

Methods: We report observed frequencies of 22 distinct lesion types in the CDE template, along with detailed information on multiplicity, locations, and volumetrics, and compare findings between mild (mTBI; GCS 13–15) and moderate-severe TBI (msTBI; GCS 3–12). Additionally, we investigate lesion co-occurrence, clustering, and the distribution of 6 CT scores.

Results: We included 4087 patients, 2744 with mTBI and 1193 with msTBI; median age was 50 years and 37% had a negative admission CT. Traumatic subarachnoid hemorrhage (45%), skull fractures (37%), contusions (31%), and acute subdural hematomas (29%) were the most frequent findings, ranking similarly in patients with mTBI and msTBI, but with up to three times higher frequencies in msTBI. In most patients with findings, distinct lesion types co-occurred. Lesion clusters were more complex in msTBI, with more co-existing lesions and signs of mass effect, higher and more heterogeneous CT score distributions. Our critical appraisal of CDEs was positive, but revealed that assessment can be time-consuming, some CDEs occur infrequently, and we identified a few redundancies and definition ambiguities.

Conclusions: Whilst primarily developed for research, implementation of CDE templates in clinical practice is advocated, potentially in an abbreviated version. Our findings and critical appraisal can contribute towards improving future CDE updates. The high prevalence of lesion co-occurrences enables further research into complex, heterogeneous pathology.

O011 Predicting recovery in patients with mild traumatic brain injury and a normal CT using serum biomarkers and diffusion tensor imaging (CENTER-TBI): an observational cohort study

Dr Sophie Richter,¹ Dr Stefan Winzek,¹ Dr Marta Correia,¹ Dr Endre Czeitzer,² Dr Daniel Whitehouse,¹ Dr Evgenios Kornaropoulos,¹ Professor Guy Williams,¹ Mr Jan Verheyden,³ Dr Tilak Das,⁴ Professor Olli Tenovuo,⁵ Professor Jussi Posti,⁵ Professor Anne Vik,⁶ Dr Kent Moen,⁶ Professor Asta Haberg,⁶ Professor Kevin Wang,⁷ Professor Andras Buki,⁸ Professor Andrew Maas,⁹ Professor Ewout Steyerberg,¹⁰ Professor David Menon,¹ Dr Virginia Newcombe ¹

¹University of Cambridge, Cambridge, United Kingdom, ²University of Pécs, Pécs, Hungary, ³Icometrix, Antwerp, Belgium, ⁴Addenbrooke's Hospital, Cambridge, United Kingdom, ⁵University of Turku, Turku, Finland, ⁶Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ⁷Morehouse School of Medicine, Atlanta, USA, ⁸Örebro University, Örebro, Sweden, ⁹Antwerp University Hospital, Antwerp, Belgium, ¹⁰University of Leiden, Leiden, Australia

Background: Even patients with normal computed tomography (CT) head imaging may experience persistent symptoms for months to years after mild traumatic brain injury (mTBI). There is currently no good way to predict recovery and triage patients who may benefit from early intervention. We aimed to assess if existing prognostic models can be improved by serum biomarkers or diffusion tensor imaging from MRI (DTI), and if serum biomarkers can identify patients for DTI.

Methods: We included 1025 patients aged >18 years with a Glasgow Coma Score >12 and normal CT from the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI) study. Biomarkers (GFAP, NFL, S100B) were obtained at a median of 8·8h (Q1-Q3 4·2–16·7) and DTI at 13 days (3–19) after injury. DTI metrics were available in 153 patients for 48 white matter tracts (ICBM-DTI-81 atlas). Incomplete recovery at three months was defined as an extended Glasgow Outcome Scale score <8. Existing prognostic models were fitted with and without biomarkers, or with and without DTI, and internally validated using bootstrapping.

Findings: 385 (38%) patients had incomplete recovery. Adding biomarkers did not improve performance beyond the best existing clinical prognostic model [optimism-corrected AUC 0·69 (95% CI 0·65–0·72) and R2 17% (11–22)]. Adding DTI significantly enhanced all models [best optimism-corrected AUC 0·82 (0·79–0·85) and R2 75% (39–100)]. The top three prognostic tracts were the left posterior thalamic radiation, left superior cerebellar peduncle and right uncinate fasciculus. Serum biomarkers could have avoided 1 in 5 DTI scans, with GFAP <12h and NFL 12–24h from injury performing best.

Interpretation: DTI substantially improved existing prognostic models for functional outcome in patients with mTBI and a normal CT, and biomarkers could help select patients for MRI. If validated, DTI could allow for targeted follow-up and enrichment of clinical trials of early interventions to improve outcome.

O012 External Validation of Admission CT Classification Scores in CENTER-TBI

Dr. Dana Pisică ,^1,2 Ana Mikolić,^3,4 Thijs Vande Vyvere,^5,6 David van Klaveren,² Andrew I.R. Maas,^7,8 Ewout W. Steyerberg,^2,9 Hester F. Lingsma²

¹Department of Neurosurgery, Erasmus MC - University Medical Center Rotterdam, Rotterdam, Netherlands, ²Department of Public Health, Erasmus MC - University Medical Center Rotterdam, Rotterdam, Netherlands, ³Department of Psychology, University of British Columbia, Vancouver, Canada, ⁴Rehabilitation Research Program, Centre for Aging SMART at Vancouver Coastal Health, Vancouver, Canada, ⁵Department of Radiology, Antwerp University Hospital, Antwerp, Belgium, ⁶Department of Molecular Imaging and Radiology (MIRA), Faculty of Medicine and Health Science, University of Antwerp, Antwerp, Belgium, ⁷Department of Neurosurgery, Antwerp University Hospital, Antwerp, Belgium, ⁸Department of Translational Neuroscience, Faculty of Medicine and Health Science, University of Antwerp, Antwerp, Belgium, ⁹Department of Biomedical Data Sciences, Leiden University Medical Center and Haaglanden Medical Center, Leiden and The Hague, Netherlands

Introduction: Traumatic brain injury (TBI) admission CT classification scores can be used for individual outcome prediction, trial enrollment/stratification, and baseline risk adjustment in observational studies. We aimed to evaluate the performance of the Marshall, Rotterdam and Helsinki scores for outcome prediction in the contemporary CENTER-TBI study, and explore model updating.

Methodology: We assessed score performance in terms of calibration (intercept, slope) and discrimination (C-statistic), for the outcomes specified in each score’s development study, 1) in the original target population, selected from the CENTER-TBI cohort according to inclusion/exclusion criteria in the development studies, and 2) in the entire CENTER-TBI cohort. We explored the most appropriate method for score updating with a closed-testing procedure.

Results: Discrimination in the original target population was good, with c-statistics over 0.70: Marshall 0.70 (discharge mortality; n=795), Rotterdam 0.75 (6-months mortality; n=701), and Helsinki 0.76/0.74 (6-months mortality and unfavorable outcome respectively; n=1367). Discrimination was better in the entire CENTER-TBI cohort (c-statistics 0.79–0.84; n=3481). In terms of calibration, Marshall and Rotterdam scores overestimated the risk of mortality at discharge and 6 months, respectively (observed over expected ratio 0.81 for both), and showed somewhat different predictor effects compared to their development studies, particularly in the entire cohort (slope 1.35, 1.49 respectively). Due to different estimates for individual score components compared to development studies, model revision was considered indicated for all three scores, in both the target subgroups and entire cohort. However, this did not lead to increased discrimination.

Conclusions: All three CT scores had good discrimination for mortality (at discharge – Marshall; at 6 months - Rotterdam and Helsinki) and unfavorable outcome (at 6 months - Helsinki). The overestimation of mortality is likely driven by overall increased survival. Model updating might be indicated to reflect increased survival and different predictor effects, particularly if models are to be applied across the entire TBI spectrum.

O013 Bridging nodes between layers of mental health, cognition, and brain morphometry in patients with mild traumatic brain injury: A multilayer network analysis of the TRACK TBI study

Professor Karen Caeyenberghs ,¹ Dr Mervyn Singh,¹ Juan Dominguez,¹ Lyndon Firman-Sadler,¹ Dr Phoebe Imms,² Prof Andrei Irimia,² TRACK TBI Investigators

¹Deakin University, Melbourne, Australia, ²University of Southern California, Los Angeles, United States

Introduction: Ample studies have demonstrated that mild TBI (mTBI) patients suffer from several mental health deficits (e.g., anxiety), compared to controls (e.g., Howlett et al., 2022). These studies have investigated mTBI symptoms using a univariate approach whereby each symptom is investigated in isolation, ignoring the interactions among the large-scale brain networks and the symptoms. In the present study, we constructed a multilayer network, to examine the direct relationships between networks of cognition, mental health, and structural brain measures in mTBI patients.

Methods: Secondary de-identified data from the subacute phase (6-month follow-up) of 457 mTBI patients (17–83y, M=38.3y, 150 females) was extracted from the TRACK TBI dataset (Sibilia et al., 2023). We selected 4 variables (i.e., “nodes”) from self-report questionnaires of mental health (mental health layer), 8 cognitive tests scores from the NIH toolbox (cognitive layer), and grey matter (GM) density values within 8 brain regions of the central executive network and salience network of the anatomical MRI scans (brain layer). Relationships between nodes (i.e., “edges'') across layers were modelled as undirected, weighted partial correlations by applying graphical LASSO regularisation. Finally, we used the bridge strength metric to identify nodes that ‘bridge’ the behavioral and brain layers.

Results: Results from our bilayer network analysis revealed that insomnia severity, immediate verbal memory, and processing speed were important nodes that bridged significant relationships across mental health and cognitive layers in the mTBI group. Results from our trilayer network analysis of cognitive, mental health and GM layer data showed the existence of 4 bridging nodes (including insomnia severity, immediate verbal memory, somatisation, and processing speed) across layers for the mTBI group.

Conclusion: The bridging nodes identified in our multilayer network analyses can be used to develop more efficient treatment programs that target key nodes involved in the mental health symptoms of mTBI patients.

O014 Trajectories of cytokines, chemokines, and acute phase proteins after TBI: Results from the 18-Center TRACK-TBI Study

Dr. Firas Kobaissy ,¹ Dr. John Yue,² Dr. Sonia Jain,³ Dr Ava Puccio,⁴ Dr. Eman Elbayoumi,¹ Dr Rawad Arja,¹ Mr. Xiaoying Sun,³ Dr Thomas van Essen,⁵ Dr Romit Samanta,⁶ Dr Patrick Belton,² Dr. Esther Yuh,² Dr Lindsay Nelson,⁷ Dr. Mahmoud . Elguindy,² Miss Joye Tracey,² Dr Shawn Eagle,⁴ Dr Frederick Korley,⁸ Dr Andrea Schneider,⁹ Dr. Pratik Mukherjee,² Dr Raquel Gardner,² Dr. Amy Markowitz,⁴ Dr David Okonkwo,⁴ Dr. Ramon Diaz-Arrastia,⁹ Dr Geoffrey Manley,² Dr Kevin Wang¹

¹Morehouse School of Medicine, Atlanta, United States, ²University of California, San Francisco, San Francisco, USA, ³University of California, San Diego, USA, ⁴University of Pittsburgh Medical Center, Pittsburgh, USA, ⁵University Neurosurgical Center Holland, Leiden University Medical Center, Haaglanden Medical Center, HAGA, Leiden and The Hague, t, Leiden, Netherlands, ⁶University of Cambridge, Cambridge, United Kingdom, ⁷Medical College of Wisconsin, Milwaukee, USA, ⁸University of Michigan, Ann Arbor, USA, ⁹University of Pennsylvania Perelman School of Medicine, Philadelphia, USA

Introduction: Neuroinflammation is a significant contributing factor to TBI. We sought to examine cytokine, chemokine and acute phase proteins trajectory in serum post-TBI.

Methodology: 18-center prospective TRACK-TBI Study (2014–2020) enrolled patients receiving head CT within 24-hours of TBI. 16 cytokines /nine chemokines and two acute-phase markers were measured using MesoScale-Discovery immunoassays in serum at day 1(D1), week 2(W2) and month 6(M6) post-TBI of different severity (GCS3-12 (n=194), GCS13–15 (n=200)) and healthy controls (HC, n=69). Linear mixed-effects models were used to estimate biomarker trajectories (log-transformed, standardized based on D1 levels from HC) over 6-months for both groups.

Results: Trajectories of 23 biomarkers differed by TBI severity. In GCS3–12, IL-6, IL-2, IL-10, IL-4, IL-15, IL-12p70, TNFα, IL-1β, MCP1, SAA. IL-16, MIP1β, and CRP showed a persistent decline with IL-6 and TNFα decline the most by −4.93 and −3.084 standardized log units (SLU) respectively from D1 to M6. GCS13–15 showed a persistent decline in IL-6, TNFα, CRP, IL-2, IL-7, TARC, IL10, IL-12p70, IL-4, IL-1β, and SAA. IL-6 showed the greatest decline by −3.359 from D1 to M6. For GCS 3–12, some markers demonstrated an inverted U-shaped trajectory; an increase from D1 to W2, while D1 to M6 showed a lesser increase. Peaking from D1 to W2 IL-5 up by 0.748, and IL-7 increased by 0.946 then less by M6. Eotaxin-3, IP10, MCP-4, MIP-α, IFNγ and IL-17A also included. For GCS13–15 group, inverted U-shape markers include MDC, MCP-4, IFNγ and IL-12/23p40. IFNγ demonstrated the greatest increase from D1 to W2 for both GCS groups. Lastly, for GCS3–12, Eotaxin, TNFβ, TARC, IL-12/23-p40 and MDC showed increasing trajectories. In GCS13–15, the persistent incline trajectory included IP10, IL-5, TNF-β, Eotaxin and Eotaxin-3. IP10 has the greatest increase from D1 to M6.

Conclusion: Different subsets of inflammatory markers demonstrated distinct temporal profiles where TBI severity impacted biomarker trajectories.

O015 Evaluation of changes in diffusion tensor image analysis along the perivascular space (DTI-ALPS) after traumatic brain injury

Miss Lydia Tridgett ,¹ Dr Alexa Walter,¹ Mr Trevor Nelson,¹ Dr Junghoon Kim,¹ Dr Ramon Diaz-Arrastia,¹ Dr Jeffrey Ware¹

¹University of Pennsylvania, Philadelphia, United States

Introduction: Glymphatic dysfunction is implicated in many neurodegenerative diseases however, its association with traumatic brain injury (TBI) remains unclear. Diffusion tensor image analysis along the perivascular space (DTI-ALPS) can be used to investigate glymphatic function as a biomarker of TBI¹.

Methodology: Two cohorts were included. Cohort 1 included 98 mild TBI (mTBI) patients and 45 healthy controls (HC). TBI participants underwent 3T MRI at 2-weeks and 6-months post-injury. Cohort 2 included 44 moderate to severe TBI (msTBI) patients and 35 HC. TBI participants underwent 3T MRI at 3-, 6- and 12-months post-injury. In both cohorts, HC underwent a single 3T MRI. DTI-ALPS was calculated by placing regions-of-interest manually in major fibre bundles at the level of the lateral ventricles orthogonal to perivascular spaces. DTI-ALPS index was computed by normalising diffusivity along these perivascular spaces to diffusivity perpendicular to these.

Results: In Cohort 1 (median age 33(25), males (72%), black (45%)), there was no difference in mTBI DTI-ALPS index compared with controls or over time from 2-weeks to 6-months. Cohort 2 (age 35(22), males (70%), black (44%)) also showed no difference in DTI-ALPS index over time from, 3- to 6- to 12-months. However, msTBI had significantly lower DTI-ALPS index (2.74, (2.53–3.14)) at 3 months compared to HC (3.34 (2.94–3.83)).

When examining demographic factors using linear regression, there was a significant effect of age. In Cohort 1 HC DTI-ALPS index was found to reduce with age but not for mTBI. In Cohort 2, no significance was found between DTI-ALPS index with age for either HC or msTBI. There was no significant effect of sex or race in either cohort.

Conclusion: Differences in DTI-ALPS index seems to be related to severity of TBI and remain fairly stable over time post-injury. These effects may be prominent in younger populations.

O016 Volumetric assessment of traumatic intracranial hematomas: is ABC/2 reliable?

Dr. Alexander Fletcher-Sandersjöö ,^1,2 Dr. Anders Lewén,^3,4 Dr. Anders Hånell,^3,4 Dr. David Nelsson,^1,2 Professor Marc Maegele,^5,6 Professor Michael Svensson,^1,2 Dr. Bo-Michael Bellander,^1,2 Professor Per Enblad,^3,4 Dr. Eric Thelin,^1,2 Dr. Teodor Svedung Wettervik^3,4

¹Karolinska Institute, Stockholm, Sweden, ²Karolinska University Hospital, Stockholm, Sweden, ³Uppsala University, Uppsala, Sweden, ⁴Uppsala University Hospital, Uppsala, Sweden, ⁵Cologne-Merheim Medical Centre, Cologne, Germany, ⁶University Witten/Herdecke, Cologne, Germany

Introduction: Precise measurement of traumatic intracranial hematoma volume is important for assessing disease progression and prognosis, as well as for selecting participants in clinical trials where hematoma volume is an important endpoint. This study aimed to compare the accuracy of the ellipsoid volumetric formula (ABC/2) with computer-assisted volumetric analysis (CAVA) in estimating volumes of traumatic intracranial hematomas.

Methodology: This was a multicenter observational study that included adult patients with moderate-to-severe traumatic brain injury (TBI). Hematoma volumes, calculated using ABC/2 and CAVA from admission CT scans, were compared using the Wilcoxon signed-rank test, Spearman's rank correlation, Bland–Altman plots, and Lin’s concordance correlation coefficient (CCC). Regression models were also used to evaluate the prognostic utility of these methods for clinical outcomes.

Results: The study included 1,179 patients with 1,543 hematomas: 655 intracerebral hematomas (ICH), 694 subdural hematomas (SDH), and 194 epidural hematomas (EDH). A robust correlation (Spearman coefficients between 0.95 and 0.98) and excellent concordance (Lin’s CCC from 0.89 to 0.96) were observed between ABC/2 and CAVA across all hematoma types. However, ABC/2 consistently showed a tendency to overestimate volumes, with mean differences of +1.58 ml, +6.25 ml, and +3.75 ml for ICH, SDH, and EDH, respectively, and some variances exceeding 50 ml. Moreover, while both methods demonstrated comparable accuracy in predicting functional outcomes, CAVA showed a slight superiority in forecasting surgical interventions and midline shift.

Conclusions: While ABC/2 provides a generally reliable volumetric assessment for traumatic intracranial hematomas, its consistent overestimation relative to CAVA limits its applicability in research necessitating precise volume measurements.

O017 Multimodal brain network connectivity as a concussion biomarker

Dr. Reut Raizman,¹ Mr. Tim Buchbinder,¹ Mrs Anat Lebovici,¹ Dr. Galia Tsarfaty, Dr. Zion Zibli,¹ Dr. Anton Peled,¹ Dr. Aleksandra Plavsic,¹ Dr. Abigail Livny ¹

¹Sheba Medical Center, Ramat-gan, Israel

Introduction: Existing diagnosis methods for psychiatric and neurological disorders, including concussion and mild traumatic brain injury, lack objective assessment tools. Concussion is characterized by alterations in brain connectivity, which to-date are not identifiable with conventional CT or MR imaging, yet these patients often suffer from long-term cognitive deficits. This study aims to combine multimodal brain connectivity with artificial intelligence (AI) techniques to identify biomarkers associated with concussion with cognitive outcome post-concussion.

Methods: Forty-nine concussed patients and 42 healthy controls in the acute phase (2 weeks post injury), underwent structural (Diffusion Tensor Imaging) and functional (resting-state fMRI) imaging, along with cognitive assessments. Personalized structural and functional multimodal connectomes were constructed, and graph theory network measures were used as features for classifying AI models to patients and controls. In addition, the combined structural and functional AI model was further used to examine association of the brain network measures and working memory scores using ordinary least squares linear regression.

Results: The multimodal connectivity AI model yielded an accuracy of 84.21% in distinguishing concussed/mild TBI patients from healthy controls (sensitivity=0.7, specificity=1, AUC=0.85). Feature importance analysis highlighted that the degree graph theory measure from both structural and functional modalities of well-established TBI-related regions such as frontal regions, hippocampus and insula, contributed the most to the classification. An association between graph theory network measures and working memory (R = 0.44, p = 0.0005) was found.

Conclusion: Our study demonstrates the value of integrating structural and functional connectomics with AI modelling for improving concussion diagnosis and outcome prediction. Well-known TBI-related brain regions played a crucial role in the diagnosis process. We conclude that the connectome holds central information linked to clinical and cognitive outcome, suggesting its potential to evolve into an objective diagnostic and prognostic tool.

O018 Abnormal large-scale network dynamics following TBI correlate with tau abnormalities

Dr Emma-Jane Mallas ,^1,2 Dr Nádia Moreira da Silva,^1,3 Dr Karl A. Zimmerman,^1,2 Dr Neil S.N. Graham,^1,2 Dr Gregory P.T. Scott,^1,2 Dr Marc A. Busche,⁴ Dr Peter N. Taylor,³ Prof David J. Sharp^1,2

¹Department of Brain Sciences, Imperial College London, London, United Kingdom, ²UKDRI Care Research &Technology Centre, United Kingdom, ³CNNP Lab, Interdisciplinary Complex Systems Group, School of Computing, Newcastle University, Newcastle Upon Tyne, United Kingdom, ⁴UKDRI at UCL, London, United Kingdom

Introduction: Tau pathology can impair neural circuitry by suppression of neuronal activity (Busche et al., 2019) and this may disrupt functional connectivity of large-scale brain networks. Tau abnormalities can be measured after TBI using blood biomarkers and levels are associated with injury severity and long-term outcomes. Here we test whether blood biomarkers of TBI, including tau, correlate with changes in brain network dynamics after TBI using dynamic functional connectivity measures that relate to cognitive function.

Methods: 76 moderate-severe TBI patients (longitudinal: subacute (<6-weeks post-injury); n=65; 6-months n=42), and 48 healthy-controls (cross-sectional) underwent blood-sampling and resting-state functional magnetic resonance imaging (fMRI). Neurofilament-light chain (Nf-L), T-Tau, glial fibrillary acidic protein (GFAP) and ubiquitin C-terminal hydrolase-L1 (UCH-L1) concentrations were sampled from plasma and S100 calcium binding protein B (S100B) from serum. fMRI time-series from 19 regions of interest representing key large-scale networks were extracted. Connectivity matrices were computed using sliding window correlation analysis and subjected to k-means clustering. Temporal characteristics of resulting states were assessed.

Results: All blood biomarkers were elevated in TBI patients compared with controls at baseline (p<0.05). Four distinct brain states were identified, including two with default mode and cognitive-control network anti-correlations. Patients spent less time in anti-correlation states and made fewer transitions compared to controls. TBI patients showed reduced entropy of state transitions (more time in fewer states) which was significantly (p<0.001) associated with GFAP (R=-0.59), UCHL1 (R=-0.58), total-Tau (R=-0.42) and Nf-L (R=-0.42) at baseline. Changes in total-Tau concentration at 6-months were negatively correlated with changes in entropy (R=-0.62, p=0.005).

Conclusion: Temporal dynamics of large-scale brain networks are altered following TBI and their normalisation correlates with the normalisation of tau levels, measured in the blood. High concentrations of tau may disrupt functional connectivity following TBI, contributing to disturbances of large-scale network function and associated cognitive deficits.

O019 Sulcal Morphological Alterations in Former American Football Players

Dr. Hector Arciniega ,^1,2 Mr. Leonard Jung,^3,4 Dr. Katherine Breedlove,⁵ Mr. Tim L.T. Wiegand,^3,4 Mr. Omar John,^1,2 Mr. Nicholas Kim,³ Dr. Zachary H. Baucom,⁶ Ms. Fatima Tuz-Zahra,⁶ Dr. Yorghos Tripodis,⁶ Mr. Tashrif Billah,³ Dr. Ofer Pasternak,^3,7,8 Mr. Michael J. Coleman,³ Dr. Charles H. Adler,⁹ Dr. Charles Bernick,^10,11 Dr. Laura J. Balcer,^12,13,14 Dr. Michael L. Alosco,¹⁵ Dr. Inga K. Koerte,^3,4,8,16 Dr. Alexander P. Lin,^5,7 Dr. Jeffrey L. Cummings,¹⁷ Dr. Eric M. Reiman,^18,19,20,21 Dr. Robert A. Stern,^15,22,23 Dr. Martha E. Shenton,^3,7,8 Dr. Sylvain Bouix²⁴

¹Arciniega Lab, Department of Rehabilitation Medicine, NYU Grossman School of Medicine, New York, United States, ²NYU Concussion Center, New York, United States, ³Psychiatry Neuroimaging Laboratory, Brigham and Women’s Hospital, Harvard Medical School, Boston, United States, ⁴cBRAIN, Department of Child and Adolescent Psychiatry, Psychosomatics, and Psychotherapy, LMU University Hospital, Ludwig-Maximilians-Universität, Munich, Germany, ⁵Center for Clinical Spectroscopy, Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, United States, ⁶Department of Biostatistics, Boston University School of Public Health, Boston, United States, ⁷Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, United States, ⁸Department of Psychiatry, Massachusetts General Hospital, Boston, United States, ⁹Department of Neurology, Mayo Clinic College of Medicine, Mayo Clinic, Scottsdale, United States, ¹⁰Cleveland Clinic Lou Ruvo Center for Brain Health, Las Vegas, United States, ¹¹Department of Neurology, University of Washington, Seattle, United States, ¹²Department of Neurology, NYU Grossman School of Medicine, New York, United States, ¹³Department of Population Health, NYU Grossman School of Medicine, New York, United States, ¹⁴Department of Ophthalmology, NYU Grossman School of Medicine, New York, United States, ¹⁵Department of Neurology, Boston University Alzheimer’s Disease Research Center and CTE Center, Boston University Chobanian & Avedisian School of Medicine, Boston, United States, ¹⁶Graduate School of Systemic Neurosciences, Ludwig-Maximilians-Universität, Munich, Germany, ¹⁷Chambers-Grundy Center for Transformative Neuroscience, Pam Quirk Brain Health and Biomarker Laboratory, Department of Brain Health, School of Integrated Health Sciences, University of Nevada Las Vegas, Las Vegas, United States, ¹⁸Banner Alzheimer’s Institute and Arizona Alzheimer’s Consortium, Phoenix, United States, ¹⁹Department of Psychiatry, University of Arizona, Phoenix, United States, ²⁰Department of Psychiatry, Arizona State University, Phoenix, United States, ²¹Neurogenomics Division, Translational Genomics Research Institute and Alzheimer’s Consortium, Phoenix, United States, ²²Department of Anatomy and Neurobiology Boston University Chobanian & Avedisian School of Medicine, Boston, United States, ²³Department of Neurosurgery, Boston University Chobanian & Avedisian School of Medicine, Boston, United States, ²⁴Department of Software Engineering and Information Technology, École de technologie supérieure, Université du Québec, Montréal, Canada

Introduction: Exposure to repetitive head impacts (RHI) is associated with brain changes and the risk of developing chronic traumatic encephalopathy (CTE). Here, we utilize structural magnetic resonance imaging to identify sulcal morphological changes associated with CTE neuropathology in former American football players and further examine the associations of age and exposure factors relating to RHI.

Methodology: Data from the DIAGNOSE CTE Research Project was used to compare the depth and width of cerebral sulci in former American football players (n=170) to unexposed asymptomatic controls (n=54) in CTE-related regions (superior frontal and temporal lobe) using CalcSulc toolbox. A generalized least square model compared groups and assessed interactions with age, age of first exposure to tackle football, and estimates of cumulative head impact index (CHII) scores of frequency, linear acceleration, and rotational force. Age, body mass index, education, race, imaging site, Apolipoprotein E4 carrier status, and total intracranial volume were used as covariates.

Results: Former American football players showed shallower sulcal depth in the left superior frontal region (p=.04) compared to controls. Furthermore, we observed an age-by-group interaction in the sulcal width of the left hemisphere temporal lobe region indicating increased width as age increased in former American football players compared to controls (p=.04). Finally, we found a significant association between wider sulcal width and younger age of first exposure (p=.02) and higher CHII-linear acceleration (p=.037) within the left hemisphere temporal lobe area.

Conclusion: These findings suggest that in vivo sulcal morphometry may serve as a valuable marker for detecting the neurostructural consequences of RHI, and a link to CTE neuropathology can be explored with postmortem studies.

O020 NX210c peptide promotes functional recovery and repair in a rat cervical clip-compression model of spinal cord injury

Dr Sighild Lemarchant,¹ Nayaab Punjani,^2,3 Svetlana Altamentova,² Jonathon Chio,^2,3 Jian Wang,² Dr Yann Godfrin,^1,4 Dr. Michael Fehlings ^2,5

¹Axoltis Pharma, Lyon, France, ²Genetics and Development, Krembil Research Institute, University Health Network, Toronto, Canada, ³Institute of Medical Science, University of Toronto, Toronto, Canada, ⁴Godfrin Life-Sciences, Caluire-et-Cuire, France, ⁵Division of Neurosurgery and Department of Surgery, University of Toronto, Toronto, Canada

Introduction: Regaining arm/hand and bladder functions are the most important priorities for cervical spinal cord injured (SCI) patients, yet with no satisfying therapeutic solutions. NX210c is a drug candidate peptide derived from a glycoprotein involved in axonal guidance during brain development. The aim of this study was to evaluate its efficacy to promote functional recovery and tissue repair in a cervical SCI model.

Methods: Adult female Wistar rats were subjected to a C6/C7 clip compression-contusion injury and treated once daily with intraperitoneal injections of NX210c (8mg/kg) or its vehicle from 4h or 8h post-injury (n=16–17/group). Sham rats received a laminectomy with vehicle treatment from 4h post-injury (n=12). Neurobehavioral tests were performed for up to 8 weeks post-injury, and rats were then sacrificed for histological assessments.

Results: Early administration of NX210c at 4h increased forelimb grip strength from 3 weeks post-injury (p<0.05) and improved several static/dynamic aspects of locomotion including interlimb coordination (regularity index or base of support of the forelimbs; CatWalk). When delaying first administration to 8h, NX210c promoted weight gain, accelerated bladder control recovery from 14 to 9 days post-injury, and improved trunk balance (inclined plane) from 1 week post-injury (p<0.05). Regardless of the therapeutic window, more SCI rats with weight support were observed following NX210c treatment, however a higher percentage of rats with weight support at 8 weeks post-injury were observed when the first administration was delayed. Greater white matter preservation and reduced cavity size at the injury epicenter were observed when NX210c first administration was delayed (n=6/group, p<0.05).

Conclusions: NX210c improves motor function, bladder control, and white matter preservation, with more benefits observed at the later initial injection timepoint. This study provides a strong proof of concept for the use of NX210c as an innovative treatment, injected by intravenous route, to start the clinical development in patients.

O021 Endothelial cell-specific knock-out of Ephrin-B2 signaling aggravates blood-spinal-cord-barrier disruption after experimental spinal cord injury

Katharina Kersting,¹ Laurens Roolfs,¹ Emily von Bronewski,¹ Lilly Waldmann,¹ Lea Meyer,¹ Melina Nieminen-Kelhä,¹ Dipl.-Biol. Irina Kremenetskaia,¹ Dr. Adnan Ghori,¹ Prof. Dr. Andre Rex,² Prof. Dr. Frank Heppner,³ Prof. Dr. Michael G. Fehlings,⁴ Prof. Dr. Peter Vajkoczy,¹ Dr. Vanessa Hubertus ¹

¹Department of Neurosurgery and Berlin Institute of Health, Charité Berlin, Germany, Berlin, Germany, ²Department of Experimental Neurology, Charité Berlin, Germany, Berlin, Germany, ³Department of Neuropathology, Charité Berlin, Germany, Berlin, Germany, ⁴Toronto Western Hospital, University Health Network and University of Toronto, Division of Neurosurgery and Krembil Neuroscience Center, Toronto, Canada, Toronto, Canada, ⁵Berlin Institute of Health, Berlin, Germany

Objective: The disruption of the blood-spinal-cord-barrier (BSCB) is a leading pathophysiology in traumatic spinal cord injury (SCI) and its restitution plays a crucial role in spinal cord regeneration. The guidance molecule Ephrin-B2 promotes cell-cell-contacts in the neurovascular unit. However, its specific role in endothelial cells following SCI is unknown. With this study, we characterize posttraumatic BSCB disruption and restitution dependent on an endothelial cell-specific knock-out (KO) of Ephrin-B2 up to 28d post SCI.

Methods: Adult CDH5-CreERT2-Efnb2lox/lox (KO) and littermate wildtype (WT) mice (m/f, n=94) underwent mid-thoracic (Th6/7) clip-compression SCI or Sham-injury (two-level laminectomy). Endothelial cell-specific Ephrin-B2 KO was induced via the promotor CDH5 in a cre-lox-system. Neurobehavioral analysis was performed using Catwalk® gait analysis and Basso Mouse Scale at 1, 3, 7, 14 and 28d post SCI. Additionally, in vivo 7Tesla-MRI was performed, and individual specimens were sacrificed for qualitative histological analysis (LFB+H&E; CD31+Evans-Blue, n=49). BSCB disruption was quantified using Evans-Blue fluorescence assay up to 14d post SCI (Tecan, n=45).

Results: Quantitative assessment of EVB-extravasation displayed a significantly increased BSCB disruption in KO SCI animals compared to WT SCI animals at 1d and 3d but not at 7 and 14d post SCI (1d: p=0,0097, 3d: p=0,0002, 7 + 14d: p>0.05). Neurobehavioral analyses showed no differences between KO SCI and WT SCI animals, with regaining of some hindlimb function in the course up to 28d post SCI in both groups, without between-group differences (1–28d: p>0.05).

Conclusion: Ephrin-B2 signaling plays a significant role in BSCB maintenance and repair in the acute injury phase, as the posttraumatic BSCB disruption is aggravated by the endothelial cell-specific KO of Ephrin-B2, leading to aggravated secondary injury. Whilst there is no significant difference in functional regeneration, further analysis must be performed to evaluate the regenerative effect and opportunities of the aggravated BSCB disruption.

O022 Establishing a novel model for sepsis post spinal cord injury in rodents

Dr. Krithika Iyer,¹ Karianne Zamiar,¹ Alyson Galvan-Lara,² Sidney Rippy,² Dr. Timothy Butterfield,³ Dr. Hiroshi Saito,² Dr. Samir Patel ¹

¹Spinal Cord & Brain Injury Research Center, Department of Physiology, University Of Kentucky, Lexington, United States, ²Departments of Surgery and Physiology, University of Kentucky, Lexington, United States, ³Athletic Training and Clinical Nutrition, University of Kentucky, Lexington, United States

Introduction: Spinal cord injury (SCI) causes motor and sensory loss below the level of injury which significantly compromises quality of life physically, mentally and socioeconomically. Systemic infection/sepsis post-SCI is a predominant secondary complication that further impairs spontaneous functional recovery and increases mortality. Sepsis can be classified as primary (occurs within 48hrs of SCI) or secondary (occurs after 48hrs) based on its onset. To date there are no clinically relevant experimental model(s) available to study complications in sepsis survivors after SCI. The current study is designed to develop a novel rodent model that mimics the long-term complications in sepsis survivors post-SCI.

Methodology: Rats were randomly divided in 4 groups – Sham, Sepsis, SCI and SCI + Sepsis. Rats in SCI and SCI + Sepsis received T10 laminectomy and contusion (200kDyn) using Infinite Horizon impactor. Sepsis was induced by injecting cecal slurry (3ml) intraperitoneally immediately after SCI. All animals received fluid resuscitation and antibiotic 8 hours after SCI and/or sepsis induction, then twice daily for 5 days. Outcome measures includes: body weight, blood parameters, weekly behavior, in vivo muscle strengths testing and histological analysis.

Results: Decreased survival was observed in SCI+Sepsis (∼39.3%) and Sepsis (∼55%) compared to SCI and Sham (100%). SCI+sepsis resulted in significantly impaired hindlimb locomotor recovery compared to all other groups. At 12 weeks post-injury, rats in SCI+Sepsis group were able to stand or walk without support (BBB∼9) whereas rats in SCI alone group walked with occasional/frequent coordination (BBB∼12–13). In vivo muscle-strength test also showed significant muscle weakness in SCI+Sepsis versus SCI. Ongoing studies are assessing blood cytokines and histological change to corelate with BBB and skeletal muscle strength.

Conclusion: In summary, this study is the first step towards understanding underlying mechanisms of sepsis post-injury and paving the way to elucidate therapeutic strategies for SCI.

O023 Systemic Application of IL-4 induces local immunomodulation after traumatic spinal cord injury in rats

Obada Alhalabi,¹ Stefan Heene,¹ Dr Guoli Zheng,¹ Xaiwei Zha,¹ Dr. Johannes Walter,¹ Prof. Thomas Skutella,² Dr. Raban Heller,^3,4,5 Prof. Klaus Zweckberger,⁶ Prof. Sandro M. Krieg,¹ Pd Dr. Med. Alexander Younsi ,¹ Prof. Andreas W. Unterberg¹

¹Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Germany, ²Department of Neuroanatomy, Institute for Anatomy and Cell Biology, Heidelberg University, Heidelberg, Germany, ³Center for Orthopaedics, Trauma Surgery and Spinal Cord Injury, Trauma and Reconstructive Surgery, Heidelberg University Hospital, Heidelberg, Germany, ⁴Bundeswehr Hospital Berlin, Department of Traumatology and Orthopaedics, Septic and Reconstructive Surgery, Berlin, Germany, ⁵Department of General Practice and Health Services Research, Heidelberg University Hospital, Heidelberg, Germany, ⁶Department of Neurosurgery, Brunswick City Hospital, Brunswick, Germany

Introduction: Traumatic spinal cord injury (SCI) elicits local and systemic inflammation cascades, limiting neuroregeneration and impeding functional recovery. We determined the systemic effect of immunomodulation with Interleukin-4 (IL-4), on the local immune reaction and regeneration in injured spinal cord tissue, in addition to the systemic cytokine landscape after SCI in rats.

Methodology: After T9/10-laminectomy, 120 female Wistar rats were randomized for thoracic clip compression/contusion SCI or sham. SCI animals received intraperitoneal IL-4 or placebo twice daily for up to 7 days post-injury (dpi). Rats were sacrificed at different timepoints (1, 3, 7, 14 and 28 dpi) with immunohistochemistry used to assess macrophage polarization, cellular neurodegeneration, and astrogliosis of explanted spinal cords. High throughput seromics were applied to measure levels of 22 cytokines in rat serum which were also compared to SCI patient serum data.

Results: IL-4-treated rats showed a significantly higher abundance of IBA1+/ARG1+ and IBA1+/CD206+ M2-macrophages compared to IBA1+/iNOS+ M1-macrophages versus placebo-treated rats. Furthermore, astrogliosis and post-traumatic cyst size were significantly reduced under IL-4 at 28 dpi. While APC+ oligodendrocytes showed a significantly higher cell count in the IL-4 group, no significant difference was noted in NeuN+ neurons compared to the placebo-treated group at 28 dpi. Seromics revealed significantly higher levels of pro-inflammatory serum cytokines in placebo-treated rats compared to sham rats in the acute to subacute post-injury phase (1, 3 and 7 dpi), which were effectively suppressed after IL-4 treatment. SCI patients with low serum levels of the same cytokines reduced under IL-4 in our rat model showed better functional outcomes.

Conclusions: Systemic IL-4 application induced systemic and local immune-modulatory effects after experimental SCI in rats, which were associated with neuroregeneration and functional recovery. Further preclinical and ultimately translational studies on IL-4 in the context of SCI should be considered.

O024 Promoting spinal cord injury recovery through LPS, IL-4, and TGF-β enhanced macrophage activation

Mr. Xiaowei Zha ,¹ Hao Wang,¹ Obada Alhalabi,¹ Guoli Zheng,¹ Maryam Hatami,² Thomas Skutella,² Sandro M. Krieg,¹ Andreas Unterberg,¹ Alexander Younsi¹

¹Department of Neurosurgery, Heidelberg University Hospital, Heidelberg, Germany, ²Department of Neuroanatomy, University of Heidelberg, Heidelberg, Germany

Introduction: Spinal cord injury (SCI) elicits an inflammatory cascade that often culminates in chronic impairment. This research explores how a triad treatment using lipopolysaccharide (LPS), Interleukin-4 (IL-4), and Transforming Growth Factor-beta (TGF-β) modulates macrophage phenotypes to support post-SCI regeneration.

Methodology: Bone marrow-derived macrophages (BMDMs) from Wistar rats were cultured and exposed to LPS, IL-4, and TGF-β combinations for 24 hours to induce a regenerative macrophage phenotype (iBMDMs). The anti-inflammatory and regenerative markers of iBMDMs were evaluated through qPCR, immunocytochemistry (ICC), and proteomic analyses. Neuron co-culture and phagocytosis assays determined the functional implications in vitro. Rats with a T9/10 clip-compression SCI model received iBMDM injections or inducers directly into the lesion site. The ensuing effects on inflammation, neuroregeneration, and functional recovery over 28 days were analyzed and compared to placebo-treated controls.

Results: Post 24-hour stimulation with LPS, IL-4, and TGF-β, iBMDMs showed significant upregulation of Arg1 and IL-10 mRNA, with protein levels visually confirmed by ICC. Proteomic profiling revealed gene expressions favoring metabolic activity, immune modulation, and enhanced phagocytosis. In neuron co-cultures, iBMDMs reduced neuronal death significantly more than untreated BMDMs and demonstrated superior phagocytosis. Post SCI and iBMDM-transplantation in vivo, rats showed elevated Arg1 expression, reduced iNOS and IL-1β levels after 3 days. Furthermore, neuroinflammation in the injured spinal cord was decreased and neuroregeneration improved after 14 days, accompanied by enhanced functional recovery. Of note, iBMDMs outperformed both the inducer treatment and placebo groups.

Conclusions: Induction of BMDMs into an anti-inflammatory, regenerative phenotype via LPS, IL-4, and TGF-β suggests a viable therapeutic strategy for SCI.

O025 Default mode network hyperconnectivity is associated with abnormal plasma pTau217 in retired professional rugby players

Ms Jessica Hain ,^1,2 Doctor Emma-Jane Mallas,^1,2 Doctor Thomas Parker,^1,2,3 Doctor Karl Zimmerman,^1,2 Ms Erin Rooney,^1,2,4 Doctor Ying Lee,^1,2,4 Ms Martina Del Giovane,^1,2 Doctor Lucia Li,^1,2 Professor Paresh Malhotra,^1,2 Mr Daniel Friedland,¹ Doctor Neil Graham,^1,2,5 Doctor Richard Sylvester,^4,6 Professor David Sharp^1,2,5

¹Department of Brain Sciences, Division of Medicine, Imperial College London, London, United Kingdom, ²UK Dementia Research Institute, Centre for Care, Research and Technology, London, United Kingdom, ³Department of Neurodegenerative Disease, The Dementia Research Centre, UCL Queen Square Institute of Neurology, London, United Kingdom, ⁴Institute of Sport, Exercise and Health, University College London, London, United Kingdom, ⁵Centre for Injury Studies, Imperial College London, London, United Kingdom, ⁶Acute Stroke and Brain Injury Unit, National Hospital for Neurology and Neurosurgery, Queen Square, London, United Kingdom

Introduction: Exposure to repetitive head impacts (RHI) and the effect this has on brain function, neurodegenerative pathophysiology and cognitive outcomes is poorly understood. Long-term outcomes associated with RHI include impaired memory and executive dysfunction. Altered default mode network (DMN) connectivity, measured by resting-state fMRI (rsfMRI), is a feature of traumatic brain injury (TBI) and Alzheimer’s disease, and is associated with cognitive impairments. This study tested the hypotheses that in ex-rugby players exposed to RHI: i) DMN connectivity will be altered and ii) this will be associated with cognitive outcomes and blood-based biomarkers of neurodegeneration and TBI.

Methodology: 181 retired elite rugby players (median age 44, 9.4% female) and 30 non-sporting controls (median age 46.5, 23.3% female) were included in this analysis. rsfMRI, blood-based biomarker (ptau217, glial fibrillary acidic protein (GFAP), Aβ42/40, neurofilament light (Nf-L)), neuropsychological performance (Trail-making test, D-KEFS and WMS-IV) and RHI exposure (concussion history, years of professional play and player position) data were analysed. Connectivity within the DMN was derived using dual regression of a seed-based ROI in the ventral posterior cingulate cortex (PCC). General linear models corrected for age and sex investigated the relationship between DMN connectivity, elite rugby participation, blood-based biomarkers, and cognitive performance.

Results: An area of the posterior DMN including the dorsal PCC showed hyperconnectivity in rugby players compared to controls. Hyperconnectivity was significantly associated with high levels of plasma pTau217 in rugby players and poorer performance on a test of executive function. DMN connectivity was not significantly associated GFAP, NfL concentration, Aβ42/40 ratio, or RHI exposure.

Conclusions: Rugby players show hyperconnectivity within the DMN compared to controls, and this is associated with increased pTau217 concentration and executive dysfunction. DMN hyperconnectivity and elevated pTau217 may be associated with increased risk of neurodegeneration following repetitive head impacts.

O026 Effects of CSF Shunting Timing on Cranioplasty Outcomes

Mb Netta Urvas ,¹ Dr. Tommi Korhonen,¹ Dr. Sami Tetri,¹ Dr. Angelos Kolias,² Dr. Ivan Timofeev,² Dr. Adel Helmy,² Dr. Peter Hutchinson²

¹University Of Oulu, Oulu, Finland, ²University of Cambridge, Cambridge, United Kingdom

Background: Hydrocephalus can occur as a complication after acquired brain injury, especially following decompressive craniectomy. However, cranioplasty may alleviate existing hydrocephalus, and in turn, hydrocephalus or CSF shunting may affect cranioplasty outcome. Optimal timing of CSF shunting remains uncertain, however single-stage shunting is associated with an increased rate of complications. We assessed the effect of shunt surgery timing (pre- or post-cranioplasty) on cranioplasty survival.

Methodology: The data was retrospectively identified from the Cambridge University Hospital and Oulu University Hospital patient databases. The primary and secondary outcome were implant removal and complications, respectively.

Results: 379 patients underwent 433 cranioplasties aged at mean 43 years (SD 17,6) with a mean ASA grade of 2,5 (SD 0,7). The median follow-up time was 9 months (IQR 21,7), during which 61 (14%) cranioplasties were removed. 68/433 (16%) cranioplasties were conducted to patients requiring shunting, of which 43 (63%) were conducted before, three (4%) during and 22 (32%) after cranioplasty. CSF shunting predicted shorter implant survival (HR 2,22, 95% CI 1,15–4,27) and implant failure during the follow-up time (OR 2,73, 95% CI 1,30–5,74). Complication rates were 51,2% and 40,1% among those shunted before and after cranioplasty, respectively (OR 1,51, 95% CI 0,54–4,28). Patients who were shunted pre-cranioplasty had slightly more SSIs (26% vs. 18%), post-operative haematomas (9% vs. 4%) and CSF leaks (0% vs. 7%) than those shunted post-cranioplasty, but overall implant removal rates were similar (26% vs. 32%, respectively, OR 0,74, 95% CI 0,24–2,28).

Conclusions: Patients with CSF shunts had higher cranioplasty removal rates and shorter implant survival than those without shunts. Patients who were shunted pre-cranioplasty had more cranioplasty infections (26% vs. 14%) and post-operative haematomas (9% vs. 4%) than those shunted post-cranioplasty. Nevertheless, implant failure rates were similar between these groups. Delaying shunting until after cranioplasty may be beneficial.

O027 Exposure to intimate partner violence related physical, but not emotional or sexual abuse, is associated with poorer cognitive performance and reduced structural brain volumes in older adults

Ms Natalie Jenkins ,¹ Dr Donald Lyall,¹ Prof. Graciela Muniz-Terrera,^2,3 Prof. William Stewart^1,4

¹University Of Glasgow, Glasgow, United Kingdom, ²University of Edinburgh, Edinburgh, United Kingdom, ³University of Ohio, Athens, United States of America, ⁴NHS Greater Glasgow and Clyde, Glasgow, United Kingdom

Objectives: Intimate partner violence (IPV) is a major global public health concern. Approximately 30% of women will experience physical or sexual IPV in their lifetime, frequently resulting in repeated exposure to TBI. Whilst the association between traumatic brain injury and neurodegenerative disease is widely understood, the long-term impact of IPV on brain aging in later life have yet to be explored. To address this, we examine neurocognitive performance and structural brain volumes in older adults who have experienced physical, sexual, or emotional IPV compared to individuals with no exposure.

Methods: 18,555 participants from UK BioBank were included in these analyses. 5,225 reported IPV of which, 2,257 reported physical IPV, 2,507 emotional IPV, and 461 sexual IPV. All participants completed neuropsychological testing and brain MRI. Participants with and without exposure to IPV were compared on eight tests of neuropsychological function, as well as structural brain volumes.

Results: The mean age of participants was 65.43 (SD=7.58) years. Physical IPV, but not emotional or sexual IPV, was associated with reduced hippocampal (p=.027) and white matter volumes (p=.033), and poorer cognitive performance in fluid intelligence(p=.009), trail making (p=.004), and digit symbol substitution (p=.010) compared to participants with no exposure. Higher frequency of physical abuse was associated with reduced hippocampal volumes (p=.040) and poorer performance in all neurocognitive tasks including fluid intelligence (p<.001), trail making task(p<.001), matrix completion(p<.001), digit symbol substitution (p=.003), tower arranging (p<.001), and pairs matching (p=.017).

Conclusion: Our results show for the first time reduced structural and functional brain changes in older adults associated specifically with physical, and not emotional or sexual IPV compared to individuals with no exposure. Furthermore, a dose-response was observed whereby increased frequency of physical IPV was associated with poorer structural and functional brain changes. These results highlight the importance of considering the long-term effects of IPV-related TBI on brain ageing and neurodegenerative disease.

O028 Chronic traumatic brain injury induces an impairment of tau clearance

Mr Luther Loose ,¹ Dr Gloria Vegliante,¹ Ms Francesca Tribuzio,¹ Dr Martina Violatto,² Dr Laura Talamini,² Dr Ada De Luigi,² Dr Francesca Pischiutta,¹ Dr Federico Moro,¹ Dr Roberto Chiesa,³ Dr Mario Salmona,² Dr Paolo Bigini,² Dr Elisa Roncati Zanier¹

¹Mario Negri Institute For Pharmacological Research IRCCS - Department of Acute Brain and Cardiovascular Injury, Milan, Italy, ²Mario Negri Institute For Pharmacological Research IRCCS - Department of Biochemistry and Molecular Pharmacology, Milan, Italy, ³Mario Negri Institute For Pharmacological Research IRCCS - Department of Neuroscience, Milan, Italy

Introduction: We previously found that severe TBI in mice induces the emergence of self-templating tau that spreads throughout the brain and contributes to neurodegeneration¹. However, it is unknown whether TBI also alters proteostatic mechanisms, which may contribute to the accumulation of pathogenic tau intermediates. Therefore, we investigated whether severe TBI affects tau clearance over time and whether aging has a synergistic effect.

Methodology: Male adult (2 months-old) and aged (15 months-old) mice were exposed to severe TBI or sham injury and injected at 1-week (w), 6 or 12 months (m) with recombinant monomeric human full-length tau bound to a Cy5 fluorophore (tau-Cy5, 1µg/2µL) in the contralateral hippocampus (n=5/group/time point). In vivo fluorescent molecular tomography imaging (Lumina III, PerkinElmer) was performed longitudinally at 2 days post-tau-Cy5 injection (dpt) and then weekly up to 42 dpt. After background subtraction, fluorescence was measured as radiant efficiency (p/s/[µW/cm2])/(pixel areaX10e4).

Results: The peak signal intensity was detected at 2dpt and was similar in all groups (mean±SD:11±7.7). Adult TBI mice showed a rapid decrease of signal intensity that reached the lowest value at 21dpt when injected 1w (TBI:1.5±1.2, sham:0.6±0.5) or 6m (TBI:4.2±2.3, sham:2.9±0.8) post-injury, with no difference compared to sham. However, a significant clearance defect emerged at 12m post-injury, with TBI mice showing higher Tau-Cy5 signal compared to sham up to 42dpt (TBI:12±6 vs sham:4.9±1.9, p<0.05). Aged TBI mice showed significant anticipation in the emergence of clearance impairment already detectable at 6m post-injury (TBI:7±0.9 vs sham:2.8±1.7, p<0.05).

Conclusion: Our data show that TBI and aging have synergistic effects on tau clearance, impeding its removal and promoting the accumulation in brain parenchyma.

O029 Longitudinal volume loss in military-relevant TBI: a LIMBIC-CENC study

Dr Emily Dennis ,¹ Dr Carrie Esopenko, Dr Nicholas Tustison, Dr Mary Newsome, Dr Brian Avants, Dr Jessica Gill, Dr Sidney Hinds II, Dr Kimbra Kenney, Dr Hannah Lindsey, Dr Sarah Martindale, Dr Mary Jo Pugh, Dr Jared Rowland, Dr Randall Scheibel, Dr Pashtun-Poh Shahim, Dr Robert Shih, Dr James Stone, Dr Maya Troyanskaya, Dr William Walker, Dr J Kent Werner, Dr Gerald York, Dr David Cifu, Dr David Tate, Dr Elisabeth Wilde

¹University Of Utah, Salt Lake City, United States

Introduction: Traumatic brain injuries (TBI) are the “signature injury” of the U.S. operations in Iraq and Afghanistan with many potential chronic sequelae. Leveraging the multi-site LIMBIC-CENC sample, we examined longitudinal alterations in brain structure in individuals with military-relevant TBI.

Methodology: We included a subset of LIMBIC-CENC: 335 participants (37F/298M, age M=41.6 years, SD=10.4) with longitudinal MRI data (interval M=2.2 years, range=0.4–7.1) from 8 sites. Lifetime possible concussive events were assessed with structured interviews and classified as mTBI versus not, during deployment or not, and as blast-related or non-blast. PTSD symptoms were measured with the PCL-5. Exposure to combat experiences was measured with the DRRI-2 section D. We used tensor-based morphometry (TBM) to measure voxelwise volume change from baseline to follow-up.

Voxelwise linear mixed effects models were run covarying for age and gender, with site as a random effect. We examined several group comparisons: blast-related mTBI vs. none, deployment mTBI vs. none, both covarying for PTSD severity, and PTSD vs. none. We additionally examined PCL-5 and DRRI-2 section D scores continuously.

Results: There was volume loss in blast-related and deployment-related mTBI in the bilateral middle temporal white matter, in regions corresponding to the inferior longitudinal fasciculus (ILF). PTSD showed no associations with volume change. There were negative associations between DRRI-2 D and volume change in the left cerebellum lobule V, right caudate, and bilateral frontal white matter, and a positive association with the left hippocampus.

Conclusions: We report volume loss in regions overlapping with the ILF in deployment-related and/or blast-related mTBI. As a key structure in the ventral visual stream, the ILF is important for object/face recognition and visuo-emotional function. Combat exposure was associated with volume loss in cerebellar lobule V, associated with inhibition processing.

O030 Traumatic axonal injury on early MRI and associations to outcome in children with moderate and severe traumatic brain injury

MD Anne-Mari Holte Flusund ,^1,2 Mari Olsen,^1,3 Oddrun Sandrød,^1,4,5,6 Joakim Stray Andreassen,^1,4 Turid Follestad,⁵ Anne Vik,^1,4 Toril Skandsen,^1,3 Kent Gøran Moen^4,7,8,9

¹Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ²Department of Radiology, Møre and Romsdal Hospital Trust, Molde Hospital, Molde, Norway, ³Department of Physical Medicine and Rehabilitation, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, ⁴Department of Neurosurgery, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, ⁵Department of Public Health and Nursing, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ⁶Department of Anesthesiology and Intensive Care Medicine, St. Olavs Hospital, Trondheim University Hospital, ⁷Department of Circulation and Medical Imaging, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ⁸Department of Radiology and Nuclear Medicine, St.Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, ⁹Department of Radiology, Vestre Viken Hospital Trust, Drammen Hospital, Drammen, Norway

Introduction: In a prospective pediatric cohort with moderate and severe traumatic brain injury (TBI) and early MRI, we explored the occurrence and burden of traumatic axonal injury (TAI), and its associations to outcome.

Methodology: 63 children (0–18 years) with moderate (n=33) or severe (n=30) TBI and MRI within 6 weeks were prospectively included. TAI lesion location, number and volumes were registered on fluid attenuated inversed recovery (FLAIR) and location and numbers on T2* gradient echo (T2*GRE) or susceptibility weighted imaging (SWI). Outcome was dichotomized into good outcome (Glasgow Outcome Scale Extended [GOSE] 7–8 at 12 months), or disability (GOSE ≤6).

Results: Median age was 13.7 years, (IQR 9.2–16.5), 70% were boys and 19% had evacuation of mass lesion. Median time from injury to MRI was 8 days (IQR 1–32). TAI was detected in 83% of the patients with severe TBI and 65% of the moderate, the proportion of patients with TAI grade 3 was 40% and 18% respectively (p=0.130). The volumes of TAI lesions on FLAIR were higher in patients with severe TBI (p=0.007). At 12 months, 48% of the patients with severe TBI had disability (GOSE ≤6) compared to 4% of the patients with moderate TBI (p<0.001). Patients with disability (GOSE≤6) had higher volumes of TAI on FLAIR (p=0.010) and higher numbers of TAI on T2*GRE/SWI (p=0.048). Bilateral TAI lesions in deep brain structures (basal ganglia, thalami, or brainstem) were only found in patients with disability (GOSE ≤6) at 12 months.

Conclusions: TAI on early MRI was a common finding in our pediatric TBI cohort. Patients with severe TBI had higher burden of TAI and worse outcome compared to patients with moderate TBI where nearly all had a good outcome at 12 months. Bilateral TAI in deep brain structures was only found in patients with disability.

O031 Establishing a gyrencephalic model of paediatric diffuse head injury in ferrets

A/Prof Frances Corrigan ,^1,2 Mr Justin Krieg,^1,2 Mr Carl Hooper,^1,2 Dr Rebecca George^1,2

¹Head Injury Laboratory, University Of Adelaide, Adelaide, Australia, ²Translational Neuropathology Laboratory, University of Adelaide, Adelaide, Australia

Introduction: Traumatic brain injury (TBI) is a leading cause of death and disability in children, with the developing brain responding to injury differently compared to adults. Pre-clinical models of paediatric TBI that replicate clinical features, including meaningful functional outcome measures and similar blood biomarker profiles are needed to understand injury evolution and identify new therapeutic targets. Gyrencephalic models provide a key intermediate step from rodents, as the presence of gryi increases brain deformation with impact. Here a diffuse TBI model was developed in 3–4-month-old ferrets, with brain development equivalent to an early elementary aged child.

Methodology: Ferrets (11–16 weeks) were allocated to sham or TBI groups with injury induced with the Closed-Head Injury Model for Engineered Rotational Acceleration at 17J. Serum was collected at 30 mins, with a group perfused at 24h (n/6 group) and at 3d (n=8–9/group) with a terminal serum sample collected. Motor and cognitive performance were assessed in the 3d group on the ladderwalk and puzzlebox. Serum GFAP and NFL were investigated independently using a Simoa HD-X Analyzer.

Results: Serum GFAP levels increased significantly at 30mins (p<0.01) and 24h (p<0.05), returning to sham level at 3d post-injury (p=0.91). In comparison NFL levels increased from 24h to 3d post-injury (p<0.05) with significant differences at 3d (p<0.001), but not 24h (p=0.07) compared to shams. Injured ferrets exhibited motor and cognitive deficits, taking longer to traverse the ladder (15.52±6.20 vs 9.55±4.26 secs, p<0.05), requiring more steps (30.5(25–39) vs 24(19–31), p<0.05) and having an overall increase in latency to reward across all trials (straw, foam and wooden plug obstacles) on the puzzlebox (p<0.05)

Conclusion: Acutely paediatric ferret TBI demonstrated biomarker patterns akin to clinical observations, with early elevation of GFAP and a subsequent NFL increase, with associated motor and cognitive deficits. Further work is needed to investigate more chronic time-points.

O032 Determinants for reduced health-related quality of life in paediatric traumatic brain injury

Sophie M. Coffeng ,¹ Manon Out,² Harm J. van der Horn,² Roos M.D. van der Jagt,² Zwany Metting,³ Joukje van der Naalt²

¹Department of Emergency Medicine, University Medical Center Groningen, Groningen, the Netherlands, ²Department of Neurology, University Medical Center Groningen, Groningen, the Netherlands, ³Department of Paediatric Neurology, University Medical Center Groningen, Groningen, the Netherlands

Introduction: Traumatic brain injury (TBI) is one of the most common causes of morbidity and mortality amongst children with a tremendous impact on their lives. However, the knowledge about health-related quality of life (HRQoL) after TBI is limited for paediatric patients. Therefore, the aim of this study was to determine the HRQoL and its risk factors in children several years after TBI.

Methodology: Data were obtained from a prospectively follow-up cohort study of paediatric TBI patients admitted to the emergency department of the UMCG between 2008–2016. Longterm HRQoL as primary outcome was measured by the PedsQL 4.0 questionnaire between 2015–2017. Patient and trauma characteristics were collected from digital patient files and posttraumatic complaints were evaluated by a separate health questionnaire.

Results : In total 416 children completed the PedsQL and the health questionnaire comprising 69 (16%) minor TBI, 271 (65%) mild TBI, 33 (8%) moderate TBI, and 43 (10%) severe TBI. Patients with a minor TBI were younger at time of accident (4 years (8 interquartile range (IQR) p<0.01). Almost half (52%) of the children experienced concentration problems, 52% had headache problems and 39% of the children experience long term memory problems. Memory problems were more present in the more severe TBI group (p<0.01). The median total PedsQL score for the total group of children was high: 88.0 (IQR16.3) with a comparable score for parent-proxies (87.4 (18.5 IQR)). Lower HRQoL was associated with severe TBI (B 8.13 (95%CI 1.13 to 15.12)), posttraumatic memory problems (B -7.3 (95%CI -10.1 to -4.6)), concentration problems (B -6.7 (95%CI -9.3 to -4.1) and headache (B -4.2 (95%CI -6.6 to -1.8)).

Conclusions: Most paediatric TBI patients score their HRQoL as good, but posttraumatic cognitive complaints and headache have a significant negative influence on HRQoL several years after injury. This effect was more pronounced in severe TBI.

O033 Associations between microdialysis metabolites, intracranial pressure and brain tissue oxygenation in paediatric TBI

Dr Devon van Eck ,¹ Dr Thandani Mlambo,¹ Dr Nqobile Thango,¹ Professor Johannes Enslin,¹ Dr Andrew Appiah-Baiden,¹ Professor Ursula Rohlwink,¹ Professor Anthony Figaji¹

¹University of Cape Town, Cape Town, South Africa

Introduction: Because paediatric physiology changes considerably over the first 12 years of life, extrapolating adult recommendations for care may be inappropriate. Monitoring brain microdialysis (MD) metabolites may complement more widely practiced monitoring techniques and assist in their interpretation. To explore this, we examined MD metabolites for associations with intracranial pressure (ICP) and brain tissue oxygen (PbtO2) in children with TBI.

Methods: We analyzed the clinical and physiological data of children ≤ 12 years old who underwent brain MD monitoring analysis using the ISCUSFlex bedside analyzer (Mdialysis, Sweden) along with ICP and PbtO2 monitors, over the duration of MD monitoring, typically up to 4 days.

Results: There were 59 patients: ages 9 months-12 years, most (63%) were males, and 75% were road traffic accident victims. Mortality was 13.5%. Median ICP was 11.5mmHg (IQR 8.8–12.6) and median PbtO2 was 32.9mmHg (IQR 26.5–39.5). MD metabolites did not correlate with ICP, although when comparing ICP groups dichotomized at 15mmHg, LPR (20.5 vs 18.3), glutamate (1.7 vs 1.2 µM) and glycerol (83 vs 72 µM) were slightly higher when median ICP was ≥15mmHg. Median PbtO2 had a weak negative correlation with LPR (r=-0.27; p=0.047). LPR was useful in graphically displaying a threshold falloff of PbtO2 values as LPR increased. Surprisingly, glucose did not correlate with PbtO2, but correlated significantly with other metabolites, LPR (r=-0.59) and glycerol (r=-0.47).

Conclusion: This is the first study to examine MD metabolites in association with ICP and PbtO2 in children. The negative correlation between LPR and PbtO2 is expected because both are influenced by perfusion of the brain. The poor correlation between MD metabolites and ICP may reflect the variability of causes of increased ICP in children which confound developing accurate ICP targets for treatment. Our results show that MD adds information to existing monitoring, and therefore potential value.

O035 Comparative Analysis of Spinal Cord-Derived and Induced Pluripotent-Derived Neural Stem & Progenitor Cells for SCI Therapy

Dr Ryan Sandarage ,¹ Dr Ahmad Galuta, Dr Eve Tsai

¹University of Ottawa, Ottawa, Canada

Introduction: Induced pluripotent stem cells (iPSCs) have revolutionized spinal cord injury (SCI) treatment by enabling the creation of neural stem/progenitor cells (NSPCs). However, a comprehensive understanding of how iPSC-derived NSPCs compare to authentic spinal cord NSPCs in molecular and functional terms remains elusive. Our study aims to provide a comprehensive characterization of bona fide spinal cord NSPCs and their isogenic iPSC-derived counterparts, specializing in the spinal cord (iPSC-SC) and the brain (iPSC-Br).

Methods: We obtained human spinal cord and skin tissue with ethics approval to establish primary NSPC cultures. From these primary cells, we derived iPSCs and differentiated them into iPSC-SC and iPSC-Br NSPCs. Assessments included differentiation, proliferation capabilities, immunostaining, and differential gene expression through RNA sequencing.

Results: Significant differences were identified in the functional and transcriptional properties of bona fide NSPCs compared to iPSC-SC and iPSC-Br. Bona fide and iPSC-SC NSPCs exhibited spinal cord regionalization, whereas iPSC-Br displayed a dorsal forebrain regionalization. Notably, iPSC-derived NSPCs shared functional and transcriptional features reminiscent of early developmental stages, including embryonic patterning genes and increased proliferation rates. Moreover, differentiation profiles were most similar between bona fide and iPSC-Br, while substantial distinctions were observed between bona fide and iPSC-SC.

Conclusion: Our study unveils unique regional, developmental, and functional characteristics differentiating bona fide spinal cord NSPCs from iPSC-derived NSPCs. Addressing these disparities holds promise for enhancing the clinical effectiveness of iPSC-derived NSPC therapies for spinal cord injuries. This investigation sheds light on the distinct attributes of these two cell types, contributing to a deeper understanding of their potential applications in the realm of spinal cord injury treatment and regenerative medicine.

O036 Peripheral nerve stimulation via a novel, bio-adhesive graft-antenna to improve outcomes following traumatic spinal cord injury

Mr Ryan Michael Dorrian ,¹ Mr Jayden Troy Christie,¹ Mr Yi Feng Liu,¹ Ms Srisankavi Sivasankar,¹ Dr Carolyn Fay Berryman,² Dr Antonio Lauto,³ Dr Anna Victoria Leonard¹

¹Spinal cord injury research group, University of Adelaide, Adelaide, Australia, ²Impact in Health, University of South Australia, Adelaide, Australia, ³School of Science, Western Sydney University, Penrith, Australia

Introduction: Peripheral nerve stimulation (NS) represents a promising intervention for spinal cord injury (SCI), having demonstrated functional improvements and neuropathic pain relief. However, current NS devices are large, invasive, and incorporate complex circuitry, subsequently risking device failure, restricting clinical applications and creating access barriers. We have developed the graft-antenna, a biodegradable device that facilitates wireless NS. Whilst having successfully promoted regeneration following nerve transections, the impact of NS via the graft-antenna has not been investigated post-SCI. We hypothesised that NS via the graft-antenna would alleviate neuropathic pain and improve motor and bladder function post-SCI by modulating neuroinflammation and promoting tissue regeneration.

Methodology: Male Sprague Dawley rats (11-weeks-old) were randomised into groups (naïve, sham, SCI-only, SCI + unilateral NS [US], SCI + bilateral NS [BS]), and endpoints (3 days/8 weeks post-SCI, n=9/group/timepoint). Graft-antenna’s (US/BS) or an inactive adhesive (SCI-only) were implanted on the sciatic nerve one week before T10 SCI induction (200kdyne, Infinite Horizon). NS was administered immediately post-injury and weekly thereafter (1Hz, 1hr/antenna). Animals were assessed for neuropathic pain (von Frey, place-escape-avoidance-paradigm [PEAP]), motor function (BBB open-field, Horizontal Ladder), and bladder function (Void spot assay, retained urine weight).

Results: Preliminary results suggest the graft-antenna may alleviate neuropathic pain, with SCI-only animals exhibiting a lower foot withdrawal force (von Frey: -14.09% change from baseline) than NS animals (US: -3.26%, BS: 0.81%) and greater pain-avoidance behaviour (PEAP percentage time in white box: SCI = 32.06%, US = 7.72%, BS = 9.72%). However, improvements to motor or bladder function have not been detected in either NS group. Statistical analysis will be completed when datasets are appropriately powered. Ongoing analysis is evaluating potential mechanisms (neuroinflammation, tissue regeneration) via immunofluorescence and lightsheet microscopy.

Conclusions: The graft-antenna offers smaller, simpler, less invasive NS, and may represent a novel intervention for individuals who experience SCI-induced neuropathic pain.

O037 Can neuro-navigated Transcranial Magnetic Stimulation be used as an additional outcome measure in an SCI trial?

Mr Azharul Khan,¹ Mr Jose Pedro Lavrador, Ms Naomi Roopnarine, Professor Elizabeth Bradbury, Dr Ana Miravella Pescador, Mr Francesco Vergani, Mr Aminul Ahmed

¹King's College Hospital, London, United Kingdom, ²King's College London, London, United Kingdom

Introduction: For any clinical trial to have the best chance of success, we need to explore and validate new assessment methods. In this feasibility study, we used neuro-navigated Transcranial Magnetic Stimulation (nTMS) combined with Tractography as an adjunct outcome tool to standard European Multicenter Study about SCI (EM-SCI) assessments. This was to determine if nTMS can be used as an additional assessment in SCI trials.

Methodology: We recruited 10 chronic (minimum of 6 months from injury) SCI subjects to carry out EM-SCI ‘core’ and ‘additional’ assessments. Subjects were a mix of thoracic and cervical injuries and with ASIA scores of A-D. We performed nTMS with tractography measuring area and volume of the functional cortical areas for each limb, and the resting motor threshold (RMT). This allowed functional characterisation and density measurements of this tract in SCI patients.

Results: Overall, there was an increase in the area and volume of the functional cortical areas and a decrease of the RMT from ASIA A to ASIA D in both upper and lower limbs. This suggests a higher functional representation with lower energy transfer required to elicit motor responses from ASIA A to ASIA D.

Conclusion: nTMS-tractography assessment can be used as an adjunct outcome tool with EM-SCI assessments for a potential SCI trial. Specifically, nTMS-tractography derived metrics can objectively measure the motor function and integrity of motor pathway, giving a novel tool for stratifying and selecting patients for SCI trials.

O038 Bench to Bedside and Back to the Bench: Spinal Cord Injury

MD, PhD Ann Parr ¹

¹University of Minnesota-Twin Cities, Minneapolis, United States

Introduction: Spinal cord injury (SCI) is devastating and has no clinically available treatments. There is likely no single cure and a toolbox of treatments should be explored including combination therapies.

Methodology: Our laboratory has focused on 3 different therapies: induced pluripotent stem cell (iPSC) derived regionally specific neuronal stem cell progenitors to create a relay system in chronic SCI, a 3D printed matrix to create spinal cord organoids/assembloids, and epidural stimulation to encourage appropriate connectivity.

Results: We have created a fast and replicable method of producing regionally specific ventral and dorsal iPSC derived human neuronal cells. We have developed a new method of 3D printing these cells. We have tested the effects of epidural stimulation on these cells after transplantation in a rat model. We also have a human clinical trial of epidural stimulation in concert with these studies.

Conclusions: Spinal cord injury is complex and a combinatorial therapy is likely needed. We have discovered that while epidural stimulation is beneficial to most patients in a clinical setting, a lack of sensory/proprioceptive function remains a problem. Further, some of our patients have demonstrated neuroplasticity in that they retain function after the stimulation is off. Thus, we have further studied this in our rat model in an attempt to elucidate mechanism. Our takeaway message is that there is a crucial interplay between basic science and clinical research that is crucial to advancement in the field.

O039 Multi-modal neuroimaging reveals radiological biomarkers of sensory, motor, and autonomic changes in distal spinal cord and brain regions following spinal cord injury and intensive neurorehabilitation

Miss Georgia Bright ,^1,2 Ms Angela Walls,^3,4,5 Dr Wickramaarchchigeige Lakshantha,^4,5,6 Prof Patrick Stroman,^7,8,9 Dr Claudio Pizzolato,^10,11 Dr Dinesh Palipana,^10,11 Prof Jillian Clark,^1,2 Dr Ryan O’Hare Doig^1,2,4,5

¹School of Medicine, Faculty of Health and Medical Sciences, The University Of Adelaide, Adelaide, Australia, ²Neil Sachse Centre for Spinal Cord Research, Lifelong Health Theme, South Australian Health and Medical Research Institute (SAHMRI), Adelaide, Australia, ³Jones Radiology Clinical Research and Imaging Centre (CRIC), SAHMRI, Adelaide, Australia, ⁴Preclinical Imaging and Research Laboratories (PIRL), SAHMRI, Adelaide, Australia, ⁵National Imaging Facility, South Australian Node, Adelaide, Australia, ⁶Australian Cancer Research Foundation (ACRF) Molecular Theranostics Laboratory, Centenary Institute, Sydney, Australia, ⁷Centre for Neuroscience Studies, Queens University, Kingston, Canada, ⁸Department of Biomedical and Molecular Sciences, Queens University, Kingston, Canada, ⁹Department of Physics, Queens University, Kingston, Canada, ¹⁰Griffith Centre of Biomedical and Rehabilitation Engineering (GCORE), Griffith University, Gold Coast, Australia, ¹¹School of Health Sciences and Social Work, Griffith University, Gold Coast, Australia

Introduction: Individuals with spinal cord injury (SCI) often experience chronic secondary health outcomes alongside permanent sensorimotor impairments. Quantitative MRI (qMRI) techniques, including diffusion tensor imaging (DTI) and functional MRI (fMRI), may provide a more comprehensive understanding of functional and microstructural changes post-SCI. Unfortunately, conducting qMRI of the lesion site post-injury poses significant challenges due to metal stabilisation artefacts. Therefore, we aimed to develop a novel multi-modal qMRI protocol to investigate post-SCI sensory, motor, and autonomic changes in distal spinal cord and brain regions.

Methodology: Brain and spinal cord fMRI and DTI pilot data were acquired from healthy volunteers and individuals with chronic traumatic SCI across two Australian clinical trials (ACTRN12622000733774; ACTRN12622001272785). A brain and thoracolumbar spinal cord fMRI paradigm was developed to interrogate autonomic (sexual) dysfunction using CONN Toolbox and Pantheon software. A semi-automated brain and whole-cord DTI protocol was also developed to derive DTI and tractography metrics using DSI Studio and Spinal Cord Toolbox software. Finally, we assessed the utility of our optimised DTI protocol via longitudinal tractography to evaluate therapeutic-related outcomes of a parallel thought-driven electromechanical assistive neurorehabilitation device program.

Results: Our fMRI pipeline revealed significant changes in sexual (dys)function-related activity among SCI compared to healthy participants. The number of sexual function-related brain regions varied notably between SCI and healthy participants. Importantly, preserved neuronal activity was observed below the level of injury in complete (ASIA A) participants, associated with significant white matter microarchitectural changes. Longitudinal tractography demonstrated significant microarchitectural differences associated with reclassification of neurological levels of injury, somatosensory changes and patient-reported outcomes.

Conclusions: The biological mechanisms underlying neurological dysfunction and recovery are still poorly understood. Our newly developed multi-modal qMRI protocol can detect SCI-induced functional and microarchitectural changes in the clinical setting. Such radiological biomarkers may assist researchers and clinicians in diagnosing and targeting functional recovery post-SCI.

O040 Plasma biomarkers in chronic single moderate – severe traumatic brain injury

Dr Gershon Spitz ,^1,2 Dr Amelia Hicks,¹ Dr Stuart McDonald,^2,3 Dr Vincent Dore,^4,5 Dr Natasha Krishnadas,^4,5 Professor Terence O’Brien,^2,3,6 Dr William O’Brien,² Dr Lucy Vivash,^2,3,6 Professor Meng Law,^2,7 Professor Jennie Ponsford,¹ Professor Christopher Rowe,^4,5 Professor Sandy Shultz^2,3,8,9

¹Monash-Epworth Rehabilitation Research Centre, School of Psychological Sciences, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia, ²Department of Neuroscience, School of Translational Medicine, Faculty of Medicine, Nursing and Health Sciences, Monash University, Melbourne, Australia, ³Department of Neurology, The Alfred, Melbourne, Australia, ⁴Florey Department of Neuroscience and Mental Health, University of Melbourne, Melbourne, Australia, ⁵Department of Molecular Imaging and Therapy, Austin Health, Melbourne, Australia, ⁶Departments of Medicine and Neurology, The University of Melbourne, Royal Melbourne Hospital, Melbourne, Australia, ⁷Department of Radiology, Alfred Health, Melbourne, Australia, ⁸Health Sciences and Human Services, Vancouver Island University, Nanaimo, Canada, ⁹Centre for Trauma and Mental Health Research, Vancouver Island University, Nanaimo, Canada

Introduction: Blood biomarkers are an emerging diagnostic and prognostic tool that reflect a range of neuropathological processes following traumatic brain injury (TBI). Here, we examine plasma biomarkers in the chronic period following TBI and their association amyloid and tau positron emission tomography, white matter microarchitecture, brain age, and cognition.

Methodology: We recruited 90 TBI participants ≥40 years of age who had suffered a single moderate-severe TBI ≥10 years previously, along with 32 non-TBI control participants. We measured plasma biomarkers using single molecule array technology (UCH-L1, NfL, tau, GFAP, P-tau181); PET tracers to measure amyloid-beta ([18F]NAV4694) and tau neurofibrillary tangles ([18F]MK6240); MRI to assess white matter microstructure and brain age; and the RAVLT to measure verbal-episodic memory.

Results: Plasma UCH-L1 (β = 1.67, p = 0.018, pAdjust = 0.044, CI95% [1.10, 2.55]) and P-tau181 (β = 1.24, p = 0.011, pAdjust = 0.044, CI95% [1.05, 1.46]) were significantly elevated in TBI participants compared to controls. Amyloid and tau PET were not elevated in TBI participants. Higher concentration of plasma P-tau181, UCH-L1, GFAP, and NfL were significantly associated with worse white matter microstructure but not brain age in TBI participants. For TBI participants, poorer verbal-episodic memory was associated with higher concentration of P-tau181 (p = 0.043, CI95% [-4.28, -0.07]; Long delay: p = 0.020, CI95% [-4.71, -0.41]), tau (Immediate memory: p = 0.014, CI95% [-11.14, -1.30]), and UCH-L1 (Immediate memory: p = 0.048, CI95% [-4.26, -0.01]).

Conclusions: Elevated plasma markers related to neuronal damage and accumulation of phosphorylated tau suggest the presence of ongoing neuropathology in the chronic phase following a single moderate-severe TBI. Plasma biomarkers were associated with measures of microstructural brain disruption on MRI and disordered cognition, further highlighting their utility as potential objective tools to monitor evolving neuropathology post-TBI.

O041 Acute plasma Vascular endothelial growth factor A (VEGF-A) after traumatic brain injury is associated with clinical features of blood brain barrier breakdown

Dr Lucia Li ,^1,3 Dr Amanda Heslegrave,² Dr Eyal Soreq,^1,3 Dr Giovanni Nattino,⁴ Dr Margherita Rosnati,¹ Dr Elena Garbero,⁴ Dr Karl Zimmerman,¹ Dr Neil Graham,^1,2 Dr Federico Moro,⁴ Dr Deborah Novelli,⁴ Dr Primoz Gradisek,⁵ Dr Sandra Magnoni,⁶ Prof Ben Glocker,¹ Prof Henrik Zetterberg,^2,7,8 Dr Guido Bertolini,⁴ Prof David Sharp^1,3

¹Imperial College London, London, United Kingdom, ²UCL UKDRI, London, United Kingdom, ³UKDRI CR&T at Imperial College London, London, United Kingdom, ⁴Mario Negri Institute, Ranica, Italy, ⁵University of Ljubljana, Ljubljana, Slovenia, ⁶Santa Clara Hospital, Trento, Italy, ⁷University of Gothenburg, Gothenberg, Sweden, ⁸Sahlgrenska University Hospital, Mölndal, Sweden

Introduction: Vascular endothelial growth factor A (VEGF-A) is produced from almost all cell types in the brain and is upregulated in the brain after traumatic brain injury (TBI). After TBI, it is involved in neuronal survival, peripheral immune cell recruitment and increased blood brain barrier (BBB) permeability. Here we report the effect of TBI on plasma VEGF-A and its relationship to clinical outcomes.

Methods: In a group of 1147 TBI patients admitted to intensive care during the European CREACTIVE study, we assess acute (day 0 and day 5 of ITU admission) plasma VEGF-A levels, and its clinical associations. VEGF-A levels was measured using the OLINK® Target 96 Inflammation panel, a high-dimensional protein assay platform testing 92 inflammation-related proteins.

Results: Plasma VEGF-A levels increased from D0 to D5, mirroring previous findings. Day 0 VEGF-A levels were higher in patients with intracerebral haemorrhage. Day 5 VEGF-A levels were higher in patients who developed refractory intracranial hypertension. Further, VEGF-A levels correlated with volumes of intraparenchymal haemorrhage (rs=0.22, p<0.001), extra-axial haemorrhage (rs=0.20, p<0.001) and perilesional oedema (rs=0.24, p<0.001) measured on initial CT scan. LASSO regression of the whole OLINK® Inflammation panel identified that Day 0 interleukin-6 (IL6) and Stem Cell Factor (SCF) levels, but not VEGF-A levels, show moderate association with injury severity (pre-hospital GCS and Marshall CT category). Day 0 IL6 was moderately correlated with VEGF-A levels (rs=0.3, p<0.001), and volume of intraventricular haemorrhage (rs=0.20, p<0.001). Both D0 and D5 VEGF-A levels were strongly related to 6-month Glasgow Outcome Scale-Extended, with higher VEGF-A levels associated with worse outcome, after controlling for age and other covariates.

Conclusions: Acute plasma VEGF-A levels are associated with intracranial haemorrhage, cerebral oedema, the development of refractory intracranial hypertension and 6-month functional outcome. This may reflect the role of VEGF-A in post-injury BBB breakdown.

O042 Serum biomarkers as adjuncts to NICE head injury guidelines (NG232, 2023) when selecting traumatic brain injury patients for CT: a CENTER-TBI study

Dr Daniel Whitehouse ,¹ Dr Ana Mikolić,^2,3 Dr Endre Czeiter,^4,5 Dr Sophie Richter,¹ Professor András Büki,⁶ Professor Kevin K Wang,^7,8 Professor Ewout Steyerberg,⁹ Professor Andrew Maas,^10,11 Professor David Menon,¹ Professor Fiona Lecky,¹² Dr Virginia Newcombe¹

¹Perioperative, Acute, Critical Care and Emergency Medicine (PACE), Department of Medicine, University of 7 Cambridge, Box 93, Addenbrooke’s Hospital, Hills Road, CAMBRIDGE, United Kingdom, ²Department of Psychology, University of British Columbia, Vancouver, Canada, ³Rehabilitation Research Program, GF Strong Rehabilitation Centre, Vancouver, Canada, ⁴Department of Neurosurgery, Medical School, University of Pecs Ret u. 2, H - 7623, Pecs, Hungary, ⁵Neurotrauma Research Group, Szentagothai Research Centre, University of Pecs, Ifjusag utja 20, H - 7624, Hungary, ⁶Department of Neurosurgery Faculty of Medicine and Health Örebro University SE 7018, Sweden, ⁷Department of Neurobiology, Center for Neurotrauma, Multiomics & Biomarkers (CNMB) Neuroscience institute, Morehouse School of Medicine (MSM), Atlanta, USA, ⁸Program for Neurotrauma, Neuroproteomics and Biomarker Research, Departments of Emergency Medicine, Psychiatry and Neuroscience, University of Florida, McKnight Brain Institute, Gainesville, USA, ⁹Department of Biomedical Data Sciences, Lei den University Medical Center, Leiden, Netherlands, ¹⁰Department of Neurosurgery, Antwerp University Hospital, Edegem, Belgium, ¹¹Department of Translational Neuroscience, Faculty of Medicine and Health Science, University of Antwerp, Belgium, ¹²Centre for Urgent and emergency care REsearch (CURE), Health Services Research Section, School of Health and Related Research (ScHARR), University of Sheffield, Sheffield, UK

Introduction: To explore the diagnostic performance of a panel of six biomarkers (GFAP, NFL, NSE, S100B, total tau, UCH-L1) in the context of UK emergency medicine practice and the “2023 National Institute for Health and Care Excellence (NICE) Head injury: assessment and early management”. Emphasis is given to those in whom the decision for CT is particularly challenging (commonly known as medium risk subjects).

Methodology: The area under the Receiver Operating Characteristic curve (AUC) was used to assess the diagnostic performance of a panel of six biomarkers (GFAP, NFL, NSE, S100B, total tau, UCH-L1) to identify those with 1) CT abnormality (including skull fracture) or 2) potential neurosurgical lesion in reference to the 2023 NICE head injury Guideline NG232 risk categories. Biomarkers were assessed both as the observed biomarker level sampled any time up to 24hrs following injury, and as an adjusted biomarker level to provide an estimate of the biomarker level at 6 hrs following injury.

Results: In medium risk subjects (n=401) the median biomarkers concentration was higher in those with traumatic CT abnormalities for all biomarkers aside from NSE (all p<0.05). When sampled within 24 hours of injury GFAP demonstrated the best diagnostic performance for any CT abnormality (AUC 0.8 [0.75 to 0.84]), with NFL and t-tau showing moderate performance. Similar results were seen when using the estimated biomarker level when sampled at 6 hours following injury, with improved performance seen from S100B, t-tau and UCH-L1 at the shorter sampling time.

Conclusions: Serum biomarkers demonstrate incremental value if used alongside the NICE Guideline NG232 criteria in discrimination of CT abnormality, with potential reductions in CT image requirements in those classified as medium risk.

O043 Autoregulatory dysfunction is associated with increasing serum natriuretic peptide levels following severe traumatic brain injury: a preliminary study

Emöke Hegedüs,^1,2 Andras Czigler ,^3,4 Dominika Lendvai-Emmert,^3,4 Zsofia Dina Magyar-Sumegi,^1,3,5 Krisztina Amrein,^3,6 Endre Czeiter,^3,6 Levente Stankovics,³ Andras Buki,^3,7 Peter Toth^3,4,8,9

¹Doctoral School of Clinical Neurosciences, Medical School, University of Pecs, Pecs, Hungary, ²Department of Anaesthesiology and Intensive Therapy, Medical School, University of Pecs, Pecs, Hungary, ³Department of Neurosurgery, Medical School, University of Pecs, Pecs, Hungary, ⁴Institute for Translational Medicine, Medical School, University of Pecs, Pecs, Hungary, ⁵Department of Psychiatry and Psychotherapy, Medical School, University of Pecs, Pecs, Hungary, ⁶Szentagothai Research Centre, University of Pecs, Pecs, Hungary, ⁷Department of Neurosurgery, Faculty of Medicine and Health, Orebro University, Orebro, Sweden, ⁸Department of Public Health, Semmelweis University, Budapest, Hungary, ⁹Department of Neurosurgery, Oklahoma Center for Geroscience and Healthy Brain Aging, University of Oklahoma Health Sciences Center, Oklahoma City, USA

Introduction: Traumatic brain injury (TBI) leads to dysfunction of cerebral blood flow (CBF) autoregulation, which determines the outcome of patients. However, the underlying mechanisms are not known. Atrial and brain natriuretic peptides (ANP, BNP) were demonstrated to be produced following TBI, and are shown to modulate vascular tone and fluid balance. We tested the hypothesis that ANP and BNP levels are associated with autoregulatory dysfunction in severe TBI patients.

Methods: We examined 25 patients with severe traumatic brain injury (TBI) at the Department of Neurosurgery of the University of Pecs, Hungary. We assessed autoregulatory function with non-invasive TCD-based monitoring. Vascular pressure reactivity was monitored by obtaining pressure reactivity index (PRx) using the ICM+ software. We analyzed the levels of natriuretic peptides (ANP, BNP) in blood samples collected within the first day post-TBI with ELISA. The mean flow index (Mx) and PRx were correlated with NP levels.

Results: Both ANP and BNP levels correlated significantly with the TCD-based autoregulatory parameter Mx (ANP vs. Mx: R=0.71, p=0.016 and BNP vs Mx: R=0.56, p=0.09). Interestingly no significant correlation was observed between PRx and NP levels.

Conclusion: Our results suggest that ANP and BNP are associated with autoregulatory dysfunction following severe traumatic brain injury. It is unlikely that NPs affect autoregulation via changes of cerebrovascular pressure reactivity. Further studies should establish the role of NPs in autoregulatory dysfunction after TBI, with special focus on fluid homeostasis and volume regulation.

O044 Trajectories of vascular injury-related biomarkers after traumatic brain injury

Dr. Andrea Schneider ,¹ Sonia Jain,² Xiaoying Sun,² Catherine Demos,³ Jacob Wohlstadter,³ Nikhil Padmanabhan,³ Taron Gorham,³ George Sigal,³ John Yue,⁴ Raquel Gardner,⁵ Lindsay Nelson,⁶ Joseph Giacino,⁷ Michael McCrea,⁶ David Okonkwo,⁸ Claudia Robertson,⁹ Pratik Mukherjee,⁴ Ava Puccio,⁸ Kevin Wang,¹⁰ Nancy Temkin,¹¹ Geoffrey Manley,⁴ Ramon Diaz-Arrastia,¹ TRACK-TBI Investigators

¹University Of Pennsylvania, Philadelphia, United States, ²University of California San Diego, ³Meso Scale Diagnostics, ⁴University of California San Francisco, ⁵Sheba Medical Center, ⁶Medical College of Wisconsin, ⁷Harvard University, ⁸University of Pittsburgh, ⁹Baylor College of Medicine, ¹⁰Moorehouse University, ¹¹University of Washington

Introduction: Cerebral microvascular injury represents a potentially treatable endophenotype of traumatic brain injury (TBI). Biomarkers which reflect this underlying pathology and change with injury recovery are needed. Our objectives were to describe trajectories of vascular-related biomarkers over 6-months post-injury and to evaluate for differences in trajectories by age, sex, and TBI severity.

Methodology: A subset of TRACK-TBI Study participants with TBI (n=196 GCS 13–15; n=194 GCS 3–12) had plasma samples collected at 1-day, 2-weeks, and/or 6-months post-injury. A panel of 15 vascular-related biomarkers (VEGF-A, VEGF-C, VEGF-D, Tie2, Flt1, P1GF, bFGF, E-selectin, ICAM-3, P-selectin, thrombomodulin, ICAM-1, VCAM-1, Ang2, and vWF) were measured using Meso Scale Diagnostics electrochemiluminescence immunoassays. Biomarkers were log transformed and standardized for analyses. Linear mixed-effects models (adjusted for age, sex, TBI severity, and timepoint) with random intercepts and slopes were used to estimate biomarker trajectories over 6-months post-injury. Interactions between each timepoint and age, sex, and TBI severity were used to determine if trajectories differed by these factors.

Results: Participants had a mean age of 39.4 (SD=16.1) years and 29.0% were female. Tie2 was associated with the greatest mean increase in standardized log units over 6-months regardless of TBI severity (GCS 13–15=1.34 [95%CI=0.42,1.06], GCS 3–12=1.01 [95%CI=1.01,1.67]). The greatest mean decrease in standardized log units over 6-months was observed in vWF for individuals with GCS 13–15 (-0.75 [95%CI=-0.96,-0.53]) and in Flt1 for individuals with GCS 3–12 (-2.25 [95%CI=-2.63,-1.88]). Trajectories of all biomarkers except VCAM-1 differed by TBI severity; P1GF, Ang2, and vWF also differed by age (all p-interaction<0.05). None differed by sex (all p-interaction>0.05).

Conclusions: Of 15 vascular-related injury biomarkers evaluated, Tie2, vWF, and Flt1 had the greatest change in levels from 1-day to 6-months following TBI, with evidence for differences in trajectory by TBI severity. Future work evaluating these biomarker trajectories in relationship to clinical outcomes is warranted.

O045 Using the Cloud for Big Data Analytics: A Case Study for Cerebral Autoregulation Analysis

Ethan Moyer ,¹ Justin Moore,¹ Rachel Thomas,² Humberto Mestre,³ Dmitriy Petrov,² Dick Moberg,¹ Ramon Diaz-Arrestia²

¹Moberg Analytics, Philadelphia, United States, ²Penn Medicine, Philadelphia, United States, ³University of Pennsylvania, Philadelphia, United States

Cerebral autoregulation refers to the mechanism by which cerebral vasculature can maintain adequate blood flow independent of the body’s systemic arterial blood pressure. Following acute primary injury, the brain is at risk for secondary injury partly due to impairment of this mechanism. Computational methods have been proposed for assessing this process, including the autoregulation surrogate, pressure reactivity index (PRx). This can then be used to derive an optimal cerebral perfusion pressure (CPPopt), representing an individualized therapeutic target based on a patient’s autoregulatory status.

As advanced neuromonitoring becomes more accessible, there is an opportunity to study the relationship between these endophenotypes and other phenomena, such as spreading depolarizations and other electrocortical activity. This necessitates a system that can handle the data required for these complex analyses. This abstract highlights how the cloud was leveraged for this purpose.

A retrospective cohort of patients sustaining a traumatic brain injury (TBI) that were monitored with surface or depth electrodes using the Moberg CNS Monitor from Penn Presbyterian Medical Center were included in this study. They were anonymized using the Moberg Analytics Client and uploaded to the Moberg Cloud Platform. Patients were automatically analyzed to find PRx, CPPopt, ULA, and LLA metrics at one minute intervals using a multi-window weighting algorithm. Electrocorticography data was filtered according to predefined filters. All data was exported to downloadable open-source formats (e.g., CSV, HDF5, etc) for further analysis.

In total, 142 monitoring sessions across 59 patients were anonymized and uploaded to the cloud, summing approximately 9146 days of monitoring time and 463.327 GB. All sessions were successfully analyzed with the described methods.

This study demonstrates the feasibility of using the cloud for large-scale analysis of neuromonitoring data. We intend to continue this work by evaluating the congruence of our analytic implementation with state-of-the-art systems, such as ICM+.

O046 Predict the In-hospital mortality for traumatic Brain Injury (TBI) patient with new simplified model of TBI score-From MIMIC-IV Database to Clinical validation

M.D. Kuo-Hsing Liao ,^1,2 Ph.D. Shu-Fen Liao,³ M.D., Ph.D. Min-Huei Hsu,⁴ Student Cian Shen Liao,⁵ M.D., Ph.D. Wen-Ta Chiu⁶

¹Department of Neurosurgery, Taipei Medical University-Wan Fang Hospital, ²Department of Neurotraumatology and Intensive Care, Taipei Neuroscience Institute, Taipei Medical University, ³Department of Medical Research, Wan Fang Hospital, Taipei Medical University, ⁴Graduate Institute of Data Science, College of Management, Taipei Medical University, ⁵School of Medicine, National Taiwan Univeristy, ⁶Graduate Institute of Injury Prevention and Control, College of Public Health, Taipei Medical University

Abstract not published

O047 Prevalence, risk factors and prognosis of ultra-early, early and late seizures after traumatic brain injury: a CENTER-TBI study

Mr Edoardo Viaroli ,¹ Mr Angelos G Kolias,¹ Mr Thomas A. van Essen, Prof Peter J.A. Hutchinson¹

¹University of Cambridge, Department of Clinical Neurosciences, Division of Neurosurgery, Cambridge, United Kingdom, Cambridge, United Kingdom

Objectives: To determine the prevalence and risk factors of ultra-early, early and late PTS and its association with outcome in patients with TBI.

Background: Post-traumatic seizures (PTS) are a clinically relevant complication of traumatic brain injury (TBI) with uncertain incidence and risk factors.

Methods: CENTER-TBI study, that was conducted in 65 European trauma Centres from 2014–2017. We identified patients with PTS. Prevalence of PTS was stratified into ultra-early (prehospital and ED), early (< 7 days) and late ( > 7 days). The association of baseline variables with early and late PTS were modelled in multivariable logistic regression to determine risk factors for PTS. The association of ultra-early and early PTS with in-hospital endpoints (length-of stay [days], complications [neurological and systemic], and time to obey commands [days]), late PTS, functional outcome (Glasgow Outcome Scale Extended [GOSE]) and quality of live (Quality of Life After Brain Injury [QOLIBRI]) were modelled in appropriate multivariable regression respectively.

Results: 456 of 4509 patients (10.1%) developed at least one seizure one year post-TBI. Ultra-early PTS occurred in 4.3% (197/4509), early PTS in 4.5% (204/4509) and late PTS in 3.5% (160/4509). In admitted patients (ward or ICU), the prevalence of early PTS was 5.5% (204/3679). Risk factors for early PTS were: pre-existing severe systemic disease on the American Society of Anesthesiologists classification (OR 3.9, 95% CI 1.6–9.6); pre-hospital insults (OR 9.1, 95% CI 4.5 – 18.3); cortical and basal subarachnoid haemorrhage (OR 4.1, 95% CI 1.6–10.1).PTS were independently with a lower GOSE after case-mix adjustment (proportional odds common OR, 95% CI) but showed no association with QOLIBRI. Also, PTS were associated with a more complicated clinical course.

Conclusions: To date, this is the first international cohort studying PTS. They seem associated with pre-existing systematic disease, pre-hospital insults and cortical and basal subarachnoid haemorrhage.

O048 HEalth And Dementia outcomes following Traumatic Brain Injury (HEAD-TBI)

Dr Emma Russell ,¹ Dr Donald Lyall,² Professor William Stewart^1,3

¹School of Psychology and Neuroscience, University Of Glasgow, Glasgow, United Kingdom, ²School of Health and Wellbeing, University of Glasgow, Glasgow, United Kingdom, ³Department of Neuropathology, Queen Elizabeth University Hospital, Glasgow, United Kingdom

Introduction: While traumatic brain injury (TBI) is acknowledged as a risk for neurodegenerative disease, robust, population level data allowing adjustment for multiple confounders are lacking. To address this, HEalth And Dementia outcomes following Traumatic Brain Injury (HEAD-TBI) interrogates whole population electronic health record data to explore the risk of dementia and wider neurodegenerative disease in individuals with TBI when compared with matched general population controls.

Methods: Scottish, population level electronic medical and death certification records were analysed for individuals with a documented history of TBI, and compared with those of a matched general population control cohort selected on a 3:1 basis; that is, for every individual with TBI history, three general population controls were identified matched by year of birth, sex, and socioeconomic status. Cox proportional hazard regression models were run to compare risk of neurodegenerative disease within the cohort.

Results: Preliminary data support increased risk of neurodegenerative disease mortality in individuals with history of traumatic brain injury, when compared with their matched general population controls. Further, risk of neurodegenerative disease varies by subtype.

Conclusions: We show increased neurodegenerative disease mortality among individuals with TBI history, when compared to matched general population controls. Ongoing analyses within the HEAD-TBI program will assess the complex interaction between TBI and the lifelong health outcomes contributing to dementia risk among this population.

O049 Comparative effectiveness of decompressive craniectomy versus craniotomy for traumatic acute subdural hematoma (CENTER-TBI): an observational cohort study

Dr. Thomas van Essen ,^1,2,3,4 MSc Inge van Erp,^1,2 Prof. dr. Hester Lingsma,⁵ Dr. John Yue,⁷ MSc Dan Pisica,^5,6 MSc Ranjit Singh,^1,2 Dr. Jeroen van Dijck,^1,2 Dr. Victor Volovici,^5,6 Dr. Alexander Younsi,⁸ Dr. Angelos Kolias,^3,9 MSc Lianne Peppel,¹⁰ Dr. Majanka Heijenbrok-Kal,¹⁰ Dr. Gerard Ribbers,¹⁰ Prof. dr. David Menon,¹¹ Prof. dr. Peter Hutchinson,^3,9 Prof. dr. Geoffrey Manley,⁷ Prof. dr. Bart Depreitere,¹² Prof. dr. Ewout Steyerberg,¹³ Prof. dr. Andrew Maas,¹⁴ Prof. dr. Wilco Peul^1,2

¹Leiden University Medical Center, Leiden, Netherlands, ²University Neurosurgical Center Holland, Leiden University Medical Center, Haaglanden Medical Center, HAGA Teaching Hospital, Leiden & the Hague, Netherlands, ³Division of Neurosurgery, Department of Clinical Neurosciences, University of Cambridge and Addenbrooke's Hospital, Cambridge, United Kingdom, ⁴Department of Surgery, Division of Neurosurgery, QEII health Sciences Centre and Dalhousie University, Halifax, Canada, ⁵Center for Medical Decision Making, Department of Public Health, Erasmus MC - University Medical Center, Rotterdam, Netherlands, ⁶Department of Neurosurgery, Erasmus MC - University Medical Center, Rotterdam, Netherlands, ⁷Brain and Spinal Injury Center, Department of Neurological Surgery, Zuckerberg San Francisco General Hospital, University of California, San Francisco, United States of America, ⁸Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Germany, ⁹NIHR Global Health Research Group on Neurotrauma, University of Cambridge, Cambridge, United Kingdom, ¹⁰Rijndam Rehabilitation and Department of Rehabilitation Medicine, Erasmus MC - University Medical Center, Rotterdam, Netherlands, ¹¹Division of Anaesthesia, Addenbrooke's Hospital, University of Cambridge, Cambridge, United Kingdom, ¹²Department of Neurosurgery, University Hospital KU Leuven, Leuven, Belgium, ¹³Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, Netherlands, ¹⁴Department of Neurosurgery, Antwerp University Hospital and University of Antwerp, Edegem and Antwerp, Belgium

Introduction: Limited evidence existed on the effectiveness of decompressive craniectomy (DC) versus craniotomy for evacuation of traumatic acute subdural hematoma (ASDH) until the recently published randomised trial RESCUE-ASDH. In this concurrent study, we aimed to determine practice patterns and compare outcomes of primary DC versus craniotomy.

Methods: We included patients with an ASDH who underwent acute neurosurgical evacuation within the multicentre, observational CENTER-TBI, which enrolled patients throughout Europe. In an instrumental variable analysis, we compared outcomes between centres according to treatment preference, measured by the case-mix adjusted proportion DC per centre. The primary outcome was 6-month Glasgow-Outcome-Scale-Extended, estimated with ordinal regression as a common odds ratio (OR), adjusted for prespecified confounders. Variation in centre preference was quantified with the median odds ratio (MOR).

Results: Between 2014 and 2017, 4509 patients with traumatic brain injury were enrolled of whom 336 (7%) underwent acute surgery for ASDH evacuation; 91 (27%) DC and 245 (63%) craniotomy. The proportion DC within acute surgery cases ranged from 6 to 67% with an interquartile range (IQR) of 12–26% among 46 centres; the odds of receiving a DC for prognostically similar patients in one centre versus another randomly selected centre were trebled (adjusted median odds ratio 2.7, p < 0.0001, Figure). Higher centre preference for DC over craniotomy was not associated with better functional outcome (adjusted OR per 14% [IQR increase] more DC in a centre = 0.9 [95% CI 0.7–1.1], n = 200). Primary DC was associated with more follow-on surgeries and complications [secondary cranial surgery 27% vs. 18%; shunts 11 vs. 5%].

Conclusion: The substantial practice variation in the employment of DC over craniotomy for ASDH did not result in different functional outcome. Primary DC should be restricted to patients in whom immediate replacement of the bone flap is not possible due to brain swelling.

O050 A propensity-matched comparison of the in-hospital mortality of traumatic brain injury patients who have undergone decompressive craniectomy versus craniotomy

Dr Arjun Pant ,¹ Dr Sarthak Sinha,¹ Dr Gopalakrishnan M.S.¹

¹Jawaharlal Institute Of Postgraduate Medical Education And Research, Puducherry, India

Introduction: Traumatic brain injury (TBI) is a significant public health concern, with various surgical approaches available for treatment, including decompressive craniectomy (DC) and craniotomy. We aimed to compare in-hospital mortality of DC with craniotomy to confirm if DC has a higher rate after propensity score matching for CRASH 14-day mortality. Additionally, we sought to compare mortality rates after excluding secondary DC patients and adjusting for hospital stay duration.

Methodology: We conducted a retrospective cohort study at a government-run tertiary care center in South India, comparing patients who underwent DC (n=139) and craniotomy (n=133) for TBI. We employed propensity score matching using the Corticosteroid Randomisation After Significant Head injury (CRASH) prognostic scores, adjusting for differences in patient characteristics. We also performed additional matching based on the duration of the hospital stay.

Results: In-hospital mortality was significantly higher in the DC group (25.4%) compared to the craniotomy group (7.5%; CI [1.76 to 10.02]) after propensity score matching (p<0.001). This trend persisted even after excluding two secondary DC patients. However, when the duration of hospital stay was included in the matching process, the difference in in-hospital mortality became non-significant (p=0.171), probably due to a reduction in the sample size due to the additional matching criterion.

Conclusion: Our findings suggest that DC may be associated with higher in-hospital mortality compared to craniotomy in TBI patients. However, this difference could be confounded by the duration of the hospital stay. Further research is needed to elucidate the complex relationship between surgical choice, hospital stay, and mortality outcomes in TBI patients.

O051 Traumatic brain injury in Cameroon: a prospective observational study in a level 1 trauma center

Dr FRANKLIN CHU BUH ,¹ Prof Irene Ule Ngole Sumbele,¹ Prof Andrew I. R. Maas,¹ Prof Mathieu Motah,¹ Prof Jogi V. Pattisapu,¹ Mr Eric Youm,¹ Mr Basil Kum Meh,¹ Dr Firas H. Kobeissy,¹ Prof Kevin W. Wang,¹ Prof Peter J. A. Hutchinson,¹ Prof Germain Sotoing Taiwe¹

¹University of Buea, Buea, Cameroun

Background: About 14 million people will likely suffer a traumatic brain injury (TBI) per year by 2050 in sub-Saharan Africa. Studying TBI characteristics and their relation to outcomes can identify initiatives to improve TBI prevention and care.

Objective: The objective of this study was to define the features and outcomes of TBI patients seen over 1 year in a level-I trauma center in Cameroon.

Materials and Methods: Data on demographics, causes, clinical aspects, and discharge status were collected over 12 months. The Glasgow Outcome Scale-Extended (GOSE) and the Quality-of-Life Questionnaire after Brain Injury (QoLIBRI) were used to evaluate outcomes six months after TBI. Comparisons between two categorical variables were done using Pearson’s chi-square test.

Results: A total of 160 TBI patients participated in the study. The age group 15–45 years was most represented (78%). Males were more affected (90%). A low educational level was seen in 122 (76%) cases. Road traffic incidents (RTI) (85%), assaults (7.5%), and falls (2.5%) were the main causes of TBI, with professional bike riders being frequently involved (27%). Only 15 patients were transported to the hospital by ambulance, and 14 of these were from a referring hospital. CT-imaging was performed in 78% of cases, and intracranial traumatic abnormalities were identified in 64% of cases. Financial constraint (93%) was the main reason for not performing a CT-scan. Forty-six (33%) patients were discharged against medical advice (DAMA) due to financial constraints. Mortality was 14% (22/160) and high in patients with severe TBI (46%). DAMA had poor outcomes with QoLIBRI. Only four patients received post-injury physical therapy services.

Conclusions: TBI in Cameroon mainly results from RTIs and commonly affects young adult males. Lack of pre-hospital care, financial constraints limiting both CT scanning and medical care, and a lack of acute physiotherapy services likely influenced care and outcomes adversely.

O052 Trauma Transfers in the North-Western Region of Pakistan: A subgroup analysis of patients with brain and spine injury

Mr William Nabulyato,^2,5 Dr Laura Hobbs ,^1,4 Dr Brandon Smith,^1,4 Dr Katharina Kohler,^1,4 Sara Halimah,³ Dr Tom Bashford^1,4

¹International Health Systems Group, University Of Cambridge, Cambridge, United Kingdom, ²Royal London Hospital, Barts Health NHS Trust, United Kingdom, ³World Health Organization, Trauma Operational Advisory Team, ⁴NIHR Global Health Research Group on Acquired Brain and Spinal Injury, ⁵University of Cambridge

Introduction: Trauma remains a substantial contributor to morbidity and mortality globally, driving the imperative for establishing comprehensive trauma care systems. Pakistan has a high burden of traumatic injury, along with a significant rural population and a fragmented healthcare infrastructure. Leveraging a dataset routinely compiled for intrahospital trauma transfers, we performed a sub-group analysis on patients with neurosurgical injuries (including spinal injuries), to better contextualise current neurotrauma pathways and facilitate targeted resource allocation in the Khyber Pakhtunkhwa (KPK) region.

Methodology: By utilising observational healthcare data sourced from KPK and accessible through the World Health Organization Eastern Mediterranean Regional Office (WHO EMRO), we conducted an analysis of summarized emergency referral records from December 2020 to September 2021. Following extensive data cleansing procedures, we employed descriptive statistical techniques to scrutinise the demographic and clinical characteristics of patients categorized under 'neurosurgery', 'back/spine' injuries, and 'paralysis'.

Results: Neurosurgical trauma represented the greatest burden on trauma transfers, constituting 29.9% of the 9366 patients transferred. A substantial proportion were clinically unstable (67.5%), young men (median age 25years, 78.1% male), requiring an 86km transfer (median travelling distance) for definitive care (66.9%). Alarmingly, 58% of patients with neurosurgical injuries did not receive any treatment at the initial presenting hospital. Back/spinal injuries accounted for 1.3% of trauma transfers, whilst 2.1% were coded as 'paralysis'. Notably, patients with back/spinal injuries exhibited greater stability compared to those with neurosurgical injuries or paralysis, with 57.2% classified as stable upon presentation.

Conclusions: Neurotrauma represents a significant injury burden in Pakistan. It predominantly affects those who are economically active (young men), thereby engendering profound social and economic repercussions. Whilst caution should be exercised when interpreting routine collected data, this data set provides valuable insights into the prevailing trauma system in Pakistan, and the pathways patients with brain and spinal injuries, take to access specialised care.

O054 Intracranial hypertension and ICU mortality after Decompressive Craniectomy from a University Hospital in Montevideo, Uruguay

Student Camila Landaboure,¹ Student Ignacio Salgado,¹ Student Juan Cedrés,¹ Student Sofía Casal,¹ Student Mauricio Eizmendi,¹ Student Victoria Oviedo,¹ Engineer Bernardo Yelicich,¹ Professor Alberto Biestro,¹ Associate Professor Corina Puppo,¹ PhD Leandro Moraes ¹

¹Neurocritical care research group, Hospital de Clinicas, Montevideo, Uruguay

Background: Decompressive craniectomy (DC) is a potent therapeutical measure applied in selected severe TBI patients. Long and short-term mortality from Latin-American centers is higher than in high income countries. Results from our unit were unknown until this study.

Methods: Retrospective, observational, cross-sectional, and analytical study. Data extracted from digital database between 2005 and 2022.

Results: 67 decompressed patients suffering from severe TBI were included and 68,2% of them had primary DC. The median (IQR) age was 37 (23 – 51) and 79,1% were men. The median initial GCS was 3 (3 – 4). 27 (42,9%) patients developed ICH after DC and 48,4% of them died. The median anteroposterior diameter was 12 cm (11,2 – 12,8). The global ICU mortality rate was 31,3% (21/67). After multivariate analysis, only ICH after DC was associated with ICU mortality (p= 0,002; OR 44 (CI 4,1 – 468,7).

Conclusion: The vast majority of the patients were young males. ICH post DC was the most important prognostic factor. 50% of the patients had a DC diameter below the size proposed by international guidelines. Long-term outcome is currently under revision.

O055 Intervention with a Medical Multi-Nutrient in Traumatic Brain Injury – A Feasibility Trial

Dr Isabell Nessel ,¹ Dr Simon C. Dyall,² Dr Laus M Broersen,³ Arnoud Carol,³ Dr Ardy van Helvoort,³ Prof Adina T Michael-Titus,¹ Dr Christopher Uff^1,4

¹Centre for Neuroscience, Surgery and Trauma, Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University Of London, London, United Kingdom, ²School of Life and Health Sciences, University of Roehampton, London, United Kingdom, ³Danone Nutricia Research, Utrecht, The Netherlands, ⁴Department of Neurosurgery, Royal London Hospital, London, United Kingdom

Aims: Traumatic brain injury (TBI) is a major cause of mortality and morbidity, including long-lasting cognitive impairment – linked to disrupted brain connectivity. Phospholipids (PL) are key components of cellular membranes, also essential for synaptic structural integrity. It has been shown that the multi-nutrient combination Souvenaid® which contains PL precursors, improves neurological outcomes in experimental TBI. We hypothesised that this intervention could support brain recovery, and we tested the feasibility of providing this multi-nutrient to TBI patients.

Methods: In a randomised, placebo-controlled, feasibility trial, adult patients with TBI, at the Royal London Hospital, London, UK, were recruited within 7 days of injury to receive either the medical multi-nutrient Souvenaid or an isocaloric placebo, daily for 6 months (NUTRA-TBI; NCT04418440). Cognitive tests (CANTAB battery for TBI), and blood sample collection were performed at baseline, 3 and 6 months.

Results: 16 patients completed the study, which was not formally powered - therefore, results are exploratory. The omega-3 index improved significantly in the intervention group and was significantly higher than in the placebo group at the end of trial, indicating compliance. Similarly, other nutritional markers improved over time. Neurofilament L (NFL) and interleukin 6 (IL-6) plasma levels decreased significantly over time in the Souvenaid-treated group. Intervention patients, more than placebo patients, improved over time in several cognitive measures.

Conclusion: The medical multi-nutrient was well tolerated, indicating feasibility of this intervention in TBI. Although exploratory, the results indicate that Souvenaid improved nutritional status post-TBI and there are potential benefits of the intervention on functional recovery.

O056 A randomised Trial of Timing to Restart Direct Oral Anticoagulants after Traumatic Intracranial Haemorrhage (Restart tICrH)

Miss Catherine McMahon ,¹ Miss Ellie Edlmann, Professor Michael Jenkinson

¹Central Manchester Neuroscience Centre, W1 - British, United Kingdom

Research question: In adults with a traumatic intracranial haemorrhage (tICrH) who are taking oral anticoagulants (OAC) when is the safest time to restart a direct oral anticoagulant (DOAC): early (1 week) or late (4 weeks)?

Background: Head injury in older people falling from a standing height affects ∼25000 per year in the UK. Approximately 17–35% are taking OACs. OACs are stopped when CT scan confirms tICrH. The decision to restart balances the risk of worsening tICrH versus stroke/thromboembolism. OAC prescribing has changed from warfarin to direct OACs (DOACs). There are no studies assessing the safest time to restart DOAC in this setting. Many patients are unaware of the importance of restarting to prevent future stroke/thromboembolism.

Design: Multi-centre, randomised controlled trial

Subjects: 1084 adults on OAC/DOAC who sustain tICrH

Objectives:

Primary Objective:

• Determine the proportion of patients who meet the composite end-point of critical haemorrhagic and thrombotic events at 12 weeks

Primary economic Objective:

• Estimate cost effectiveness of restarting DOAC

Secondary Objectives:

Determine:

• Patient function & quality of life

• Time to death & mortality rate

• Patient/carer attitudes to restarting DOAC

Research Plan: Samples size / recruitment:

The time to critical haemorrhagic/thrombotic event is the primary outcome based on a composite event rate of 12% in the group restarting at 4 weeks. When the sample size in each group is 483, with 80 events required, a 0.05 level two-sided log-rank test for equality of survival curves has 90% power to detect a 6% absolute reduction in survival probability, the difference between a proportion at 12 weeks of 0.88 and a proportion of 0.94 (constant hazard ratio of 2.066).

This study has successfully been funded by the NIHR and recruitment is due to commence in July 2024

O057 Rotational thromboelastometry may improve identification of clinically significant coagulopathy in traumatic intracranial haemorrhage

Mr Abhiram Hiwase ,^1,2 Dr Ahad Sabab,¹ Dr Christopher Ovenden,¹ Ms Lola Kaukas,¹ Mr Michael Laden,³ Dr Ngee Foo,^1,2 Dr Mark Finnis,² A/Prof Benjamin Reddi,^1,2 Dr Adam Wells,^1,2 A/Prof Daniel Ellis^1,2

¹Royal Adelaide Hospital, ²Adelaide Medical School, University of Adelaide, ³Flinders Medical School, Flinders University

Abstract not published

O058 Targeted cervical cooling for severe traumatic brain injury: a protocol and rationale for SELETHERM 2

Dr Erta Beqiri,¹ Mr Alexis Joannides,¹ Dr Andrea Lavinio ¹

¹University Of Cambridge, Cambridge, United Kingdom

Introduction: Severe traumatic brain injury (TBI) presents a significant challenge in neurocritical care, with intracranial pressure (ICP) management and brain temperature regulation being pivotal for patient outcomes. Existing whole-body cooling methods are limited by systemic side effects and restricted patient mobility. SELETHERM 2 builds on the novel SELETHERM technology, aiming to evaluate the efficacy and mechanisms of a cervical external cooling device for targeted brain temperature control in severe TBI.

Methodology: This randomized controlled trial will compare the novel cervical cooling device against standard care in adult severe TBI patients requiring ICP management. Primary outcomes focus on brain fever burden, with secondary outcomes including therapy escalation needs and biomarkers of brain injury. The study will also explore the device's mechanism of action on patient physiology, using continuous brain and core temperature monitoring alongside biomarker analysis.

Results: The SELETHERM 1 study demonstrated the feasibility and safety of the cervical external cooling device, showing controlled ICP and selective brain temperature management without adverse venous obstruction or intracranial hypertension. The pilot study indicated a potential for improved neuroprotection, providing a foundation for the larger SELETHERM 2 trial.

Conclusions: SELETHERM 2 aims to validate the clinical efficacy and safety of targeted cervical cooling in severe TBI, potentially offering a new standard for managing brain temperature with fewer systemic side effects. This study could significantly impact TBI management strategies, improving patient outcomes and reducing healthcare resource utilisation.

O059 Return of intracranial beta oscillations and traveling waves with recovery from traumatic brain injury

Dr. Alex Vaz ,¹ Dr. Connor Wathen,¹ Dr. Stephen Miranda,¹ Dr. Samuel Tomlinson,¹ Dr. John Arena,¹ Dr. Kamila Bond,¹ Dr. Sanjana Salwi,¹ Dr. Sonia Ajmera,¹ Dr. Ludovica Bachschmid-Romano,² Dr. James Gugger,¹ Dr. Danielle Sandsmark,¹ Dr. James Schuster,¹ Dr. Ramon Diaz-Arrastia,¹ Dr. Ashwin Ramayya,³ Dr. Isaac Chen,¹ Dr. Dmitriy Petrov¹

¹University Of Pennsylvania, Philadelphia, United States, ²Institute of Neurobiology of the Mediterranean, Marseille, France, ³Stanford University, Palo Alto, United States

Traumatic brain injury (TBI) remains a pervasive clinical problem associated with significant morbidity and mortality. However, TBI remains clinically and biophysically ill-defined, and prognosis remains difficult even with the standardization of clinical guidelines and advent of multimodality monitoring. Here we leverage a unique data set from TBI patients implanted with either intracranial strip electrodes during craniotomy or quad-lumen intracranial bolts with depth electrodes as part of routine clinical practice. By extracting spectral profiles of this data, we found that the presence of narrow-band oscillatory activity in the beta band (12–30 Hz) closely corresponds with the neurological exam as quantified with the standard Glasgow Coma Scale (GCS). Further, beta oscillations were distributed over the cortical surface as traveling waves, and the evolution of these waves corresponded to recovery from coma, consistent with the putative role of waves in perception and cognitive activity. We consequently propose that beta oscillations and traveling waves are potential biomarkers of recovery from TBI. In a broader sense, our results support neurophysiological models that explain emergence from coma as resulting from a recovery of thalamo-cortical interactions that coordinate cortical beta rhythms.

O060 Real-time measurement of potassium and glucose by cerebral continuous online microdialysis in severe TBI patients: a pilot study

Dr. Sarah Svirsky,¹ Dr. Hansen Deng,¹ Dr. Andrea Jaquins-Gerstl,² Dr. David Okonkwo,¹ Dr. Adrian Michaels,² Ava Puccio ¹

¹Dept. of Neurological Surgery, University Of Pittsburgh, Pittsburgh, United States, ²Dept. of Chemistry, University Of Pittsburgh, Pittsburgh, United States

Introduction: Minimizing secondary injury is key for acute care of patients with severe traumatic brain injury (sTBI). Expanding the toolkit of multi-modal neuromonitoring allows the neurocritical care team additional real-time data to address care. Microdialysis using electrochemical biosensors has the potential to identify secondary injury, namely spreading depolarizations (SD), in real-time.

This pilot study aims to characterize detected SDs, defined as potassium (K+) transients, alongside cerebral glucose levels as a metric of metabolic demand, observed real-time in patients with sTBI.

Methods: Participants were enrolled between 2019–2023 under an IRB-approved protocol. Inclusion criteria was >18 years, initial Glasgow Coma Scale (GCS) score of 3 − 10 who received intracranial pressure and brain tissue oxygen catheters (least affected hemisphere) as a standard of care. Real time monitoring of K+ and glucose in the dialysate stream was performed by continuous online microdialysis.

Results: 17 participants were enrolled and analyzed (Mean age: 50±19yrs, 59% male; median GCS: 6). Mean monitoring time was 77.35hrs (Range 34.5–137.42hrs) and median SD detected was 3 (Range 0–8). Three SD phenotypes emerged: high K+/glucose decline, high K+/glucose increase, and high K+/no change. The temporal profile of SD events varied from multiple SDs within 1hr (clusters) or multiple SDs across the monitoring period. Blood glucose and K+ levels decreased over monitoring time.

Conclusions: SD phenotypes reveal K+ spikes occur independent of glucose levels and ongoing efforts aim to quantify K+ transients to further understand phenotypic differences, in conjunction with existing oxygen and glucose monitoring. Future directions include incrementally collecting cerebral dialysate, cerebrospinal fluid and blood to measure metabolic biomarkers and provide a comprehensive assessment of SD secondary injury pathology. This pilot study paves the way for integration of this novel, innovative and safe neuro-monitoring method into standard of care for sTBI and other severe neurological injuries.

O061 Influence of apolipoprotein E genotype on the proteomic profile in cerebral microdialysis after human severe traumatic brain injury: a prospective observational study

Dr Caroline Lindblad ,^1,2,3 Dr Andrea Klang,^2,4 Dr David Bark,² Dr Cristina Bellotti,⁵ Associate Professor Anders Hånell,² Professor Per Enblad,² Associate professor Anders Lewén,² Associate professor Elham Rostami^2,5

¹Karolinska Institutet, Stockholm, Sweden, ²Department of Medical Sciences, Section of Neurosurgery, Uppsala University, Uppsala, Sweden, ³Department of Clinical Neurosciences, University of Cambridge, Cambridge, United Kingdom, ⁴Department of Rehabilitation Medicine, Uppsala University Hospital, Uppsala, Sweden, ⁵Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

Introduction: Pre-injury patient genetic factors account for a substantial amount of outcome prediction variance following TBI. Among implicated genetic variants, single-nucleotide polymorphism in apolipoprotein E (APOE) has been linked to worse prognosis following TBI, but the underlying mechanism is still unknown. We hypothesized that APOE genotype would affect the levels of structural, or inflammatory, proteins in cerebral microdialysate following severe TBI.

Methodology: We conducted a prospective, observational study of patients with severe TBI treated with invasive neuromonitoring including cerebral microdialysis at Uppsala University Hospital. All patients were characterized regarding APOE genotype. Utilizing fluid- and plate-based antibody arrays, we quantified 101 proteins (of which 89 were eligible for analysis) in cerebral microdialysate at one- and three-days following trauma. Statistical analysis included clustering and linear mixed modelling techniques.

Results: In total, 26 patients were included, and all relevant genotypes of APOE were represented in the data. Among all proteins tested, 39 proteins showed a time-dependent expression level. T-distributed stochastic neighbor embedding and hierarchical clustering demonstrated a weak clustering tendency in the data, but not primarily to genotype. Using linear mixed models, two proteins (the inflammatory protein CLM-1 and the neurotrophic protein GFRα1) were found to have protein levels concomitantly dependent upon time and genotype, albeit this effect diminished upon multiple testing correction. Neither microtubule-associated protein tau, amyloid-β (Aβ) peptide levels, nor the Aβ42/40 ratio were seen related to time from trauma or APOE genotype.

Conclusions: This is the first study in clinical severe TBI examining the influence of APOE genotype on microdialysate protein expression. Protein levels in cerebral microdialysate following trauma were dependent on time from trauma, corroborating previous work on longitudinal protein expression following TBI. We also identified protein expression level alterations dependent on APOE genotype, which might indicate that APOE-mediated protein-level alterations affect ongoing pathophysiology in the injured brain.

O062 Characteristics of microdialysis metabolites in paediatric traumatic brain injury

Dr Thandani Mlambo ,¹ Dr Devon Van Eck,¹ Dr Nqobile Thango,¹ Prof Johannes Enslin,¹ Dr Andrew Appiah-Baiden,¹ Prof Ursula Rohlwink,¹ Pro Anthony Figaji¹

¹University Of Cape Town, Cape Town, South Africa

Introduction: As with most advanced brain monitoring methodologies, little is known about bedside microdialysis (MD) characteristics in children, even though adult and paediatric physiology differ considerably. To describe these characteristics, we analyzed our experience with brain MD monitoring in children with severe traumatic brain injury (TBI) and examined associations with mortality and functional outcome.

Methods: We analyzed the clinical and physiological data records of children (aged ≤ 12 years) who underwent brain MD monitoring analysis using the ISCUSFlex bedside analyzer (Mdialysis, Sweden). The primary outcome was mortality, the secondary outcome was the Pediatric Glasgow Outcome Score Extended (PGOS-E) at 6 months.

Results: We had 59 patients in our study: the age range was from 9 months old to 12 years, most (63%) were males, and 75% were road traffic accident victims. MD monitoring was typically employed for a maximum of 4 days. Mortality was 13.5%. Of the survivours, 69% had a favourable outcome (GOS-E scores of 1–3 vs 4–7). From more than 4000 hourly metabolite recordings, median values for metabolites were: pyruvate 79.4µM, lactate 1.5mM, lactate/pyruvate ratio 19.5, glycerol 69.7µM, glutamate 1.45µM, and glucose 1.29mM. We found that the LPR was significantly higher in non-survivors than survivors (23.7 vs 18.3; p=0.03) and there was a trend to glutamate being higher in non-survivors (3.9 vs 1.2 µM; p=0.06). Amongst survivours, the LPR did not differ between favourable and unfavourable outcome groups (17.9 vs 18.9), but glycerol was higher with unfavourable outcomes (95.7 vs 55.2 µM, p=0.05).

Conclusion: This is the first large study to describe the typical characteristics of bedside microdialysis values in paediatric TBI and examine associations with outcome. Higher LPR values in patients with unfavourable outcomes were largely driven by patients who died, while glycerol tended to be higher in survivours with unfavourable functional outcomes.

O063 Integration of cerebral microdialysis with sensor technology for continuous brain metabolic monitoring

Mr Chisomo Zimphango ,¹ Dr Farah C. Alimagham,¹ Ms Monica J. Killen,¹ Dr Agnieszka P. Zakrzewska,¹ Dr Adam M.H. Young,¹ Dr Tanya Hutter,² Dr Keri L.H Carpenter,¹ Professor Peter J. Hutchinson¹

¹Division of Neurosurgery, Department of Clinical Neurosciences, University of Cambridge, Cambridge, United Kingdom, ²Walker Department of Mechanical Engineering, Austin, USA

Introduction: This work extends foundational research on the Quantum Cascade Laser Mid-Infrared (QCL-MIR) sensor to assess its practical application in a clinical setting, focusing on patients with severe traumatic brain injuries (TBIs) at Addenbrooke's Hospital's Neurocritical Care Unit (NCCU) [1, 2]. Initial studies demonstrated the sensor's potential in monitoring metabolic variations, laying the groundwork for its assessment in practical neurocritical care settings.

Methodology: Data were collected from five acute TBI patients undergoing neurocritical care, at Addenbrooke's Hospital NCCU. Patients received care following established protocols, with interventions based on intracranial and cerebral perfusion pressures. The QCL-MIR sensor was integrated with cerebral microdialysis to facilitate continuous monitoring of the brain's extracellular fluid, aiming for real-time metabolic analysis. Data analysis was conducted using MATLAB R2023a, focusing on glucose absorbance patterns and their implications on patient management.

Results: Continuous sensor monitoring revealed dynamic metabolic responses to TBIs. Notably, fluctuations in glucose levels were observed, with significant implications for patient care. These metabolic variations highlight the importance of continuous, real-time monitoring in informing treatment strategies for TBI patients.

Conclusions: The integration of the QCL-MIR sensor with cerebral microdialysis in a clinical setting provides a promising approach to neurocritical care, offering insights into the metabolic state of TBI patients. This study highlights the sensor's potential in enhancing patient management through continuous metabolic monitoring. However, challenges such as the impact of tubing length on measurement accuracy and the need for sophisticated data interpretation models were identified. Future research should focus on optimising these technologies and methodologies to improve the precision of prognosis and therapeutic interventions in TBI care.

O064 Do you see what eye see?: experiences of using explainable artificial intelligence techniques in a deep learning model for papilledema detection

Dr Brandon Smith ,^1,2 Dr Lekaashree Rambabu,^1,2,3 Mr Phupha Amornkijja,⁶ Mr Christos Antonopoulos,⁶ Mr Blendi Bylygbashi,⁶ Mr Roshen Sidhu,⁶ Mr Thomas Edmiston,² Dr Stasa Tumpa,¹ Dr Katharina Kohler,^1,2,5 Mr Angelos Kolias,^1,4 Prof. Peter Hutchinson,^1,4 Dr. Tom Bashford^1,2,5

¹NIHR Global Health Research Group on Acquired Brain and Spine Injury, University of Cambridge, Cambridge, United Kingdom, ²International Health Systems Group, Department of Engineering, University of Cambridge, Cambridge, United Kingdom, ³Department of Cardiovascular Sciences, University of Leicester, Leicester, United Kingdom, ⁴Division of Neurosurgery, Addenbrooke's Hospital, Cambridge University Hospitals NHS Foundation Trust, Cambridge, United Kingdom, ⁵Division of Anaesthesia, Addenbrooke's Hospital, Cambridge University Hospitals NHS Foundation Trust, Cambridge, United Kingdom, ⁶Department of Engineering, University of Cambridge, Cambridge, United Kingdom

Introduction: Papilledema, a condition characterised by optic disc swelling, is a useful indicator of potentially life-threatening intracranial hypertension secondary to intracranial pathologies such as traumatic brain injury. Early and accurate detection of papilledema using Artificial Intelligence (AI) can be used as a diagnostic aid, (1) particularly in resource-limited settings where access to specialist neurosurgical intervention may be limited. Due to the ‘black box’ nature of AI and the need for all clinical AI systems to be explainable, we explored the application of explainable AI (XAI) techniques, specifically Gradient-Weighted Class Activation Mapping (Grad-CAM), to gain insights into a deep learning model developed for papilledema detection.

Methodology: A convolutional neural network (CNN) model based on EfficientNetB3 architecture was trained on a dataset comprising publicly available retinal images of normal (n=641) and pathological eyes (papilledema, n=318; pseudopapilledema, n=269). Grad-CAM was employed to visualise the regions of interest the model ‘focused on’ during predictions as a heat map.

Results: Grad-CAM visualisations revealed that the model successfully focused on relevant regions, in close proximity to the optic disc border, in its role in papilledema detection, reflecting our current clinical understanding regarding the landmarks of interest in papilledema diagnosis. We propose that XAI techniques may also play a secondary role in informing further data acquisition or pre-processing stages to reduce redundant data. This may reduce the computational requirements for training and inference, which can be advantageous when working with limited hardware resources, for example, in contexts of ‘edge’ computing, resource-limited settings, or large datasets.

Conclusion: This study showcases the successful implementation of an AI model for papilledema detection, with promising results on test and validation sets. The integration of Grad-CAM provided valuable insights into the model's inner workings; XAI techniques will continue to be an important consideration in developing AI-enabled, interpretable diagnostics for brain injury.

O065 Head Injury Management in a LMIC Tertiary Trauma Care Center: Patient profile, Current lacunae & Solutions

Dr Karthigeyan Madhivanan ,¹ Dr Sunil Gupta, Dr Pravin Salunke, Dr Dhandapani Sivashanmugam

¹Postgraduate Institute Of Medical Education & Research (PGIMER), Chandigarh, India

Introduction: While the magnitude of traumatic brain injury/ head injury (HI) burden has been a significant concern in LMIC conditions, major studies addressing such an issue have been relatively less. Knowledge of specific epidemiological characteristics can guide appropriate intervention strategies, with a registry-based data adding to its value. This study, based on extensive volume data, provides an analytical-overview of the clinico-epidemiological data of HI-patients, and focuses on the prevailing challenges with reasonable solutions from the perspective of developing nations.

Methodology: This is a registry-based observational analysis of a large-volume cohort of 14,888 patients with HI in a tertiary trauma-care-center of North-West India. The various clinico-epidemiological parameters and risk-factors were analyzed in a multivariate model, with the severity of HI.

Results: Notably, 50% of the patient load belonged to mild HI, despite referral (90.3%) being the predominant source. Only about one-third (30.8%) had severe HI. Less than a third reached tertiary services within an optimal time. Higher age, male gender, road traffic accidents, helmet non-usage, alcohol influence, systemic injuries and certain imaging features showed an independent association with the severity of HI.

Conclusions: This epidemiological study in a lower-middle income country setting brings out an important aspect of suboptimal utilization of the existing peripheral health-care systems. Strengthening and integrating these facilities with the tertiary centers in a hub and enhanced-spoke model, task sharing and efficient back-referrals can potentially enhance neurotrauma care, while avoiding the overburden in the apex centers. Adherence to traffic legislations is an area that needs to be focused, to reduce the HI severity.

O066 Global Neurotrauma Outcomes Study Spine: An international, multi-centre, prospective cohort study on the injury profile, management and outcomes of traumatic spinal injury

Ms Saniya Mediratta ,¹ Stefan Yordanov,^1,2 Jibin Francis,^1,2 G Balamurali,³ Karol Budohoski,¹ Alexis Joannides,^1,2 Tariq Khan,^1,4 Radek Kindl,⁵ Michael Martin,⁶ Colum Nolan,⁷ Abenezer Tirsit, Sarah Woodrow,⁹ Peter Hutchinson,^1,2 Rikin Trivedi^1,2

¹NIHR Global Health Research Group on Acquired Brain and Spine Injuries, University of Cambridge, Cambridge, United Kingdom, ²Division of Neurosurgery, Department of Clinical Neurosciences, Addenbrooke's Hospital and University of Cambridge, Cambridge Biomedical Campus, Cambridge, United Kingdom, ³Kauvery Hospital, Chennai, Tamil Nadu, India, ⁴North West General Hospital & Research Centre, Peshawar, Pakistan, ⁵Epworth Eastern Specialist Centre, Victoria, Australia, ⁶Obex Technologies Ltd, Cambridge, UK, ⁷Department of Neurosurgery, National Neuroscience Institute, Singapore, ⁸Addis Ababa University & Tikur Anbessa Hospital, Addis Ababa, Ethiopia, ⁹Cleveland Clinic at Akron General Hospital, Akron, USA

Objectives: Traumatic spinal injury (TSI) accounts for a significant proportion of disability and death worldwide, with the majority of this burden affecting individuals in LMICs. The current global approach to TSI care is inconsistent with considerable geographical differences. GNOS Spine aims to provide a comprehensive summary of the case-mix, management and short-term outcomes of TSI worldwide.

Methods: NIHR-funded, international, multi-centre, prospective observational study. Data was collected via an anonymised, secure online platform by local study teams at any institution assessing patients with TSI. All adults with radiologically confirmed TSI were included, in any consecutive 30-day period. Countries were stratified by Human Development Index (HDI), a composite of life expectancy, education and income measures.

Results: 1049 patients from 88 hospitals across 28 countries were included between 30th July 2021 and 29th March 2023. Distribution of records by HDI tier was: very-high 574(55%), 170(16%), medium 254(24%), and low 51(5%). The most common mechanism of injury was fall from standing in very-high HDI countries (216 [38%]), fall from height in high HDI (59 [35%]) and medium HDI tiers (101 [40%]) and road traffic accident in low HDI countries (29 [57%]). The most common Frankel Grade on admission was E in very-high HDI (450 [78%]), high HDI (86 [51%]) and medium HDI tiers (116 [46%]), however, A in low HDI institutions (21[41%]). Thoracolumbar injuries were most commonly seen across all income groups. Surgical intervention was performed in 41% of patients in the low HDI tier, 44% in the medium HDI tier, 49% in the high HDI tier and 26% in the very-high HDI tier. Short-term outcome analysis used an ordinal mixed-effects model analysing change in Frankel Grade at discharge (or 6-weeks post-admission, whichever occurred first).

Conclusions: GNOS Spine provide the first global snapshot of the case-mix, management, and short-term outcomes of patients with TSI.

O067 Can the CRASH prognostic model’s prediction of the risk of 14-day mortality in traumatic brain injury be used to evaluate the performance of a trauma center? A prospective cohort study

Dr Sarthak Sinha ,¹ Dr Gopalakrishnan Madhavan Sasidharan¹

¹Jawaharlal Institute Of Postgraduate Medical Education And Research, Puducherry, Puducherry, India

Introduction: The Corticosteroid Randomisation After Significant Head Injury (CRASH) calculator can reliably predict 14-day traumatic brain injury (TBI) mortality. It uses age, GCS, pupil reactivity, and CT Findings apart from the country. We aimed to repurpose the calculator to assess the performance of a neurotrauma centre.

Methodology: We did a prospective cohort study to compare predicted 14-day mortality risk with their actual mortality. We risk-stratified them into low, moderate, high, or very-high categories, and we calculated their cumulative 14-day mortality risk scores. We followed up on these patients for 14 days and saw how much was the actual mortality in each of the four groups.

Results: With 19 low-risk patients (16%), 40 moderate and high-risk patients (34% each), and 19 very high-risk patients (16%), the overall 14-day mortality of 119 patients was 48%, and this matched with the overall predicted mortality of 47.9 % (CI: 43.3- 52.4). The individual groups had a mortality risk prediction of 11.3% (8.6–13.9), 32.9% (30.8–35), 62.2% (52.7–64.6), 85.8% (83.4–88.1) for the low, moderate, high, and very-high risk groups. Correspondingly, observed, real mortality rates were 10.5%, 35%, 55%, and 100%.

Conclusion: The individual groups and the overall TBI cohort showed good agreement with CRASH’s predictions except for the very-high-risk group indicating sub-optimal care for that group in our centre. We also recommend this methodology be used to compare neurotrauma care centres.

O068 Serum GFAP and NfL trajectories after sport-related concussion: Subgroups and clinical associations

Dr William O’Brien,¹ Miss Becca Xie,¹ Dr Steven Mutimer,¹ Professor Terence O’Brien,^1,2 Professor Sandy Shultz,^1,2 Dr Stuart McDonald ^1,2

¹Monash University, Melbourne, Australia, ²Alfred Health, Melbourne, Australia

Introduction: The risk of cumulative consequences from repeated sport-related concussion (SRC) underscores the need for objective tools for monitoring clinical and neurobiological recovery. Blood biomarkers GFAP and NfL show promise in tracking neurobiological recovery, but their trajectories and associations with clinical features and outcomes of SRC remain unclear.

Methodology: This study aimed to construct a temporal profile for serum GFAP and NfL after SRC in Australian football players, identify subgroups with distinct trajectories, and investigate associations between biomarker trajectories, presence of loss of consciousness (LOC), symptomatology, cognitive performance, and return to training times. Outcomes were assessed at 24h, and 1w, 2w, 4w, 6w, 8w, 12w and 26w.

Results: Eighty-one SRCs and 56 controls completed a total of 945/1057 eligible testing sessions. Compared to controls, SRC cases exhibited higher GFAP at 24h and 4w, and higher NfL from 1w to 12w. Growth mixture modelling identified two GFAP subgroups: 'extreme prolonged' (16%) and 'moderate transient' (84%). For NfL, three subgroups were identified: 'extreme prolonged' (7%), 'moderate prolonged' (15%), and 'minimal/no change’ (78%). SRC cases with LOC (33%) had higher GFAP at 24h, 1w, 2w and 4w, and higher NfL from 1w to 12w, compared to cases without LOC. Return to training times were longer in the 'extreme’ compared to ‘moderate’ GFAP subgroup, and the NfL ‘extreme’ and ‘moderate’ subgroups compared to the ‘minimal’ subgroup, and for individuals with LOC compared to those without LOC.

Conclusions: A subset of SRC cases, particularly those with LOC, showed heightened and prolonged increases in GFAP and NfL, that persisted for at least 4 weeks. Serial biomarker measurement could identify such cases, guiding individualized return to play decisions based on neurobiological recovery. The association between prolonged biomarker elevations and LOC may support the use of more conservative return to play timelines for athletes with this clinical feature.

O069 Concurrent symptom domains and associations with recovery timelines after sport-related concussion in the Ivy League-Big Ten Epidemiology of Concussion Study

Dr. Bernadette D’Alonzo ,¹ Andrea L.C. Schneider,¹ Ian J. Barnett,¹ Christina L. Master,² Roy H. Hamilton,¹ Douglas J. Wiebe³

¹University Of Pennsylvania, Philadelphia, United States, ²The Children's Hospital of Philadelphia, Philadelphia, United States, ³University of Michigan, Ann Arbor, United States

Introduction: Research into how individual concussion symptoms cluster into domains has been largely descriptive, and not considered associations with recovery timelines. Studies have also not investigated concurrent symptom domains, despite agreement that multiple, overlapping symptom domains are common in clinical presentations of concussion. We investigate associations between concurrent symptom domains and time to three recovery outcomes after concussion. We also consider the role of sex in these relationships.

Methodology: We used data from the Ivy League–Big Ten Epidemiology of Concussion Study, and included sport-related concussions across academic years 2015–16 through 2019–20 with complete symptom data (n=1160). We use “symptom profile” to characterize how athletes endorse concurrent symptom domains following concussion. Symptom data consisted of the 22 symptoms measured in the Sport Concussion Assessment Tool (SCAT). Outcomes were time (in days) from injury to 1)symptom resolution, 2)return to academics, 3)return to full play.

Results: Females were more likely to endorse headache, sensory, and affective symptom domains. Four classes/symptom profiles characterized how athletes endorse concurrent symptom domains (1: “low” on all domains, 2: “high” on headache+sensory domains, 3: “high” on vestibulo-ocular+cognitive+sleep domains, 4: “high” on all domains). Time to symptom resolution, return to academics, and return to full play was consistently shorter among class/symptom profile 4 compared to the other classes/profiles. Compared to class/profile 1, classes/profiles 2, 3, 4 were less likely to have symptoms resolve over time (Adjusted HR=0.74, 95%CI 0.63–0.88; HR=0.74, 95%CI 0.60–0.92; HR=0.50, 95%CI 0.43–0.57, respectively). Results were similar for return to academics and full play outcomes. We found no evidence of interaction by sex.

Conclusions: Four symptom profiles characterized how symptom domains co-occur in our sample of collegiate athletes with sport-related concussion, and we found differences in recovery timelines among these symptom-profile groups. Findings inform and underscore the need for targeted, symptom-specific interventions in concussion management.

O070 Acute Head-and-Neck Cooling and Symptom Severity in acute Boxing

Md, Phd Stud Ali Al-husseini ,¹ Emanuel Blomstrand,² Emeritus Professor Yelverton Tegner,³ Senior Neurosurgent and Professor Niklas Marklund⁴

¹Department of Clinical Sciences, Neurosurgery, Lund University., Lund, Sweden, ²Medical Programme, Lund University, Lund, Sweden, ³Department of Health, Education and Technology, Division of Health and Rehabilitation, Luleå University of Technology, Luleå, Sweden, ⁴Department of Clinical Sciences, Neurosurgery, Lund University, Skåne University Hospital, Lund, Sweden

Introduction: Repetitive impacts in contact sports, such as boxing, increase the risk of brain injury. These impacts occur at times of elevated core body and brain temperatures, induced by the exercise, that may exacerbate the injury. We hypothesized that acute head-and-neck cooling, recently shown to shorten return-to-play in concussed ice hockey players, applied acutely following a boxing bout is associated with attenuated symptom severity scores using the sport concussion assessment tool (SCAT5) during the 1st week post-fight.

Method: Healthy elite boxers aged ≥ 18 (n=30) were recruited, and randomly assigned to either receive acute head-and-neck cooling for 45 minutes or standard post-fight care. Head impacts were recorded from match videos. Symptom severity is assessed using the self-assessment symptom evaluation of the SCAT-5 tool at 45 minutes post-fight, day 3, and day 6.

Results: A distinct increase in post-fight symptoms was observed in all boxers (p <0.001). The number of direct impacts to the head slightly more associated with higher scores on the SCAT-5 (r = 0.19; p=0.29), than the number of indirect hits to the body or total strikes (r = (-0.05); p=0.78 and r =0.13; p=0.46, respectively). At day 6 post-fight, there were no significant differences in SCAT5 symptom severity score in boxers receiving head-neck cooling (-1± IQR (-6–0)) when compared to those that received routine management ((-6) ± IQR (-12) - (0); p= 0.15), recruitment ongoing.

Conclusion: In this first randomised trial, still ongoing, using acute head-neck cooling following a competitive boxing bout, the number of hits to the head associated weakly with symptom scores post-fight. While there was a slight trend of lower symptom scores by head-and neck cooling, this did not reach statistical significance. The results of a larger population of boxers will be presented at the meeting.

O071 Novel concussion assessment methods relevant to the immediate and acute post-injury periods; SCAT5, OVRT and salivary assessment data from the RESCUE-RACER concussion in motorsport study

Dr Naomi Deakin ,¹ Alex Kiderman,² Keri Carpenter,¹ Gwen Kennedy,³ Peter Hutchinson¹

¹University Of Cambridge, Cambridge, United Kingdom, ²Neurolign USA LLC, Pitsburgh, United States of America, ³University of Edinburgh, Edinburgh, Scotland

Introduction: Sport-related concussion (SRC) is one of the most topical health issues in sports medicine; however, its diagnosis is complex. “Motorsport” encompasses many events and vehicles, attributed relatively high rates of concussion - despite ongoing safety improvements. This presentation outlines promising assessment methods relevant to SRC and mTBI; multimodal capture of neurophysiological function and fluid biomarkers in saliva.

Methodology: Via two sport-specific participant identification centres (PICs) and a single secondary care enrolment site (UK), RESCUE-RACER participants (16yrs+) were recruited to baseline (CArBON) and/or post-injury (CARS) studies. N=50 current motorsport competitors without mTBI <6mnths completed a single assessment (CArBON). In CARS, participants were assessed trackside immediately post-incident (n=53); n=21 were selected for additional weekly follow-up (referral to established clinical service; 1–3wks). CARS participants were segregated by clinical status (exposed, concussed, recovered).

Results: RESCUE-RACER participants included adolescent/adult competitors with a range of nationalities and ethnicities. All were active closed/open-wheel motorsport competitors (amateur to World Championship). CArBON and CARS participants shared near-identical age ranges (15–52/3yrs) and were predominantly male (91–98%). Learning difficulties (esp.dyslexia) were comparably common (12–15%); attentional disorders were four-times more common in CARS (2%vs.8.5%). No participants were involved in/considering litigation. Trackside data is presented for SCAT5 and ocular, vestibular, and reaction time (OVRT) assessments, supported by analysis of protein biomarkers of brain injury (GFAP, NFL, total-tau, UCH-L1).

Conclusions: A) Immediate post-injury SCAT5 assessment does not easily identify concussion in this motorsport cohort. B) Sports participation resulted in enhanced performance (ocular, RT), captured with a novel 3D headset (Dx100); post-participation data may represent a more appropriate baseline (vs. office-based). Acceptable reliability was achieved (12 variables, 5 OVRT tasks; 24hrs). C) Salivary NFL levels were higher in concussed vs.exposed participants over time (concussion identification); salivary NFL, GFAP, and UCH-L1 levels were lower in recovered vs.concussed (monitoring recovery). Future analyses include microRNA.

O072 Arachnoid cysts in contact sports athletes: a systematic review of the literature and case series from a specialist sports concussion clinic

Mr Andrew Stevens ,^1,2 Dr Kamal Yakoub,¹ Mr David Davies,^1,2 Prof. Antonio Belli,^1,2 Mr Phil O’Halloran²

¹University Of Birmingham, Birmingham, United Kingdom, ²University Hospitals Birmingham, Birmingham, United Kingdom

Introduction: Arachnoid cysts (AC) are associated with a risk of rupture or haemorrhage following head impact, and pose a potential predisposing factor for significant complications of sport-related concussion. Despite a recognised association between ACs and intracranial haemorrhage/cyst rupture, the risk profile of participating in contact sports with AC is not well defined.

Methods: A systematic review of the literature was performed using Medline, EMBASE, Web of Science, Google Scholar, and reference lists of related works. Studies of all designs were included which reported cases or cohorts of athletes and sportspeople with AC. A case series of all athletes with AC treated at Birmingham Sports Concussion Clinic (BSCC) between 2017–2023 is also presented.

Results: 62 case reports detailed athletes presenting with structural brain injury associated with AC. 11/62 cases resulted in subdural hygroma from cyst rupture; 51/62 cases presented with subdural haematoma (chronic or subacute). 5/62 cases were managed conservatively, with remaining cases undergoing burr hole evacuation/craniotomy. No case resulted in unfavourable neurological outcomes. Prospective studies are limited to two paediatric cohort studies [1,2]. During follow up, the combined prevalence of structural brain injury was 5/301 (1.7%). 11 athletes from BSCC were identified, with 10/11 experiencing no previous complications with exposure to an average of 3.3 concussions (range 1–9) in their sporting careers.

Conclusions: AC is an incidental finding in athletes, though there is no prospective evidence from adult cohorts to inform the risk of contact sports participation. There are frequent reports of structural brain injury associated in athletes, yet based on paediatric cohort studies and the BSCC experience, complications of AC appear to be rare. This review has not identified evidence to suggest that participation in sports with AC is of significant risk, though individualised assessment and discussion of the potential risks of contact sports participation should be offered.

O073 Single-cell RNA sequencing reveals marked cell-specific differences in ferroptosis gene signatures across sex in a murine model of TBI

Dr Aditya Kumar ,¹ Dr Dhivyaa Rajasundaram,² Dr Pavan Upadhyayula,³ Taruna Vani Neelakantan,³ Chaim Sneiderman,² Zujian Xiong,² Keri Janesko-Feldman,² Dr Vincent Vagni,² Dr Brent Schlegel,² Dr Joshua Catapano,¹ Adam Eberle,¹ Margaux Miller,¹ Dr Sudhanshu Raikwar,¹ Dr Anupama Rani,¹ Dr Chia-Ling Phuah,¹ Dr Dennis Simon,¹ Dr Brent Stockwell,³ Dr Patrick Kochanek,² Dr Gary Kohanbash,² Dr Ruchira Jha¹

¹Barrow Neurological Institute, Phoenix, United States, ²University of Pittsburgh Medical Center, Pittsburgh, United States, ³Columbia University, New York, United States

Background: Ferroptosis is increasingly recognized as a potential contributor to early and secondary neuronal damage following traumatic brain injury (TBI). Inhibition of ferroptosis has been shown to decrease neuronal dysfunction and death in preclinical models. Differential regulation of ferroptosis genes across sex, neuronal and inflammatory cell subtypes may help identify precise targets in TBI. We sought to identify transcriptomic nodes of heterogeneity in these markers at the single-cell level (SC) in a murine model of severe TBI across sex.

Methods: TBI was induced via controlled cortical impact (CCI) in male and female mice. Pericontusional tissue or equivalent cortex in naïve (n=3–7/group, 24h) was dissociated into SCs for RNA sequencing (Novaseq6000). SC gene expression matrices (Seurat) yielded UMAPs of SC transcriptomic profiles. Ferroptosis-specific differentially expressed genes (DEGs) were analyzed (R-packages), with differences reported as log 2-fold change (log2FC) and p values BH adjusted for multiple comparisons.

Results: 177,794 sequenced cells were analyzed. We previously identified 23 cell-clusters including 8 microglial subtypes (MG1–8). Key genes regulating ferroptosis were interrogated, including SLC7A11, GPX4, HIF1a, PTGS2, ACSL4, and CHAC1. DEGs were unique depending on biological sex and cell type. For example, SLC7A11, canonically associated with inhibiting ferroptosis, was downregulated in females (vs males) in neurons (log2FC -0.5, p 4.02e-5) and NK cells (log2FC -0.6, p 1.38e-6), but upregulated in MG2 (log2FC 0.5, p 9.03e-28) and a subtype of ependymal cells (log2FC 0.5, p 1.13e-10).

Conclusion: Ferroptosis gene-expression signatures show marked variation across multiple cell types and are sexually dimorphic. This may inform sex-based differences in neuronal death, the immune micro-environment, and novel cell-specific targets. Further research is warranted to understand the biological significance of these signatures and their role in impacting secondary injury and outcome post-TBI.

O075 Post-injury treatment with 7,8-Dihydroxyflavone attenuates white matter pathology following focal traumatic brain injury in aged mice

Mr Georgios Michalettos ,¹ MD PhD Fredrik Clausen,² MD PhD Elham Rostami,^2,3 MD PhD Niklas Marklund³

¹Lund Brain Injury Laboratory for Neurosurgical Research, Department of Clinical Sciences, Neurosurgery, Lund, Sweden, ²Department of Medical Sciences, Section of Neurosurgery, Uppsala, Sweden, ³Department of Neuroscience, Karolinska institute, Stockholm, Sweden

Introduction: Traumatic brain injury (TBI) is associated with white matter pathology, leading to long-lasting behavioral disturbances. Pharmacological interventions which can preserve brain structure and promote plasticity may show promise for improving neurological outcome post-TBI. The brain-derived neurotrophic factor (BDNF)-mimetic small molecule 7,8 dihydroxyflavone (DHF) can efficiently cross the blood-brain barrier. We hypothesized that post-injury DHF treatment could influence injury to white matter components of the aging mouse brain using an experimental model of focal TBI.

Methods: We used the controlled cortical impact (CCI) model in 23-month-old male and female mice and DHF, or saline in age- and sex matched controls, was administered subcutaneously 30 mins after injury (5 mg/kg). Survival time-points were 2, 7 and 14 days post-injury (dpi). Injection of DHF or saline was repeated on 1 and 3 dpi. Mouse brains were histologically assessed for tissue loss using Luxol Fast Blue (LFB)/Nissl staining. Additionally, immunohistochemistry and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) were used to visualize axonal profiles as well as apoptotic oligodendroglia in the subcortical white matter.

Results: Quantification of Nissl-stained gray matter revealed no effect of DHF on cortical tissue loss at 2, 7 and 14 dpi. Quantification of LFB-stained area revealed a preservation of white matter tissue of DHF-treated mice on 2 and 7 dpi. Apoptotic CC1+/Olig2+ oligodendroglia were present on 2 dpi but no difference was observed between treated and untreated mice. SMI-31 immunostaining revealed extensive axonal swellings within the ipsilateral external capsule on 2 dpi.

Conclusions: DHF might exert a protective effect on the peri-contusional white matter through mitigation of axonal damage rather than attenuation of oligodendrocyte loss. While promising, more research is needed to establish the optimal time window, dosing and mechanisms for the protection observed when using DHF to attenuate white matter injury following TBI in the aged animal.

O076 Unraveling the impact: acute cation-chloride cotransporter disruption following severe TBI in the developing gyrencephalic brain

Dr. Alexandra Hochstetler ,¹ Ms. Ya'el Courtney, Mr. Benjamin Baskin, Ms. Tawny Stinson, Dr. Maria Lehtinen, Dr. Beth Costine-Bartell

¹Boston Children's Hospital / Harvard Medical School, Boston, United States

Introduction: Maintenance of brain ion and fluid homeostasis is critical for typical neurodevelopment in children. Traumatic brain injury (TBI) can precipitate dysfunction of ion and fluid homeostasis in neurons and glia leading to irreversible tissue damage. We previously demonstrated seizure burden and subarachnoid hemorrhage drives hypoxic-ischemic injury patterns in an age-dependent manner in swine. One explanation for these differences is the “GABA switch” from excitatory to primarily inhibitory GABA in neurons across brain regions during early post-natal development, characterized by differential expression and phosphorylation of the cation chloride cotransporters NKCC1 and KCC2. In severe TBI, we hypothesize the activation of the cation chloride cotransporters in our “toddler” piglets versus “infant” piglets may be differentially disrupted.

Methodology: We used a validated combination of injuries including cortical impact, midline shift, subarachnoid hemorrhage, traumatic seizures, and brief apnea and hypoventilation to cause spreading hypoxic-ischemic injury. A series of male, Yorkshire piglets aged 7 days “infant” and 30 days “toddler” underwent severe TBI injuries plus 24 hours of ICU, sham injuries plus 24 hours of ICU, or naïve (no injury, no ICU). Fresh brain tissues were collected for biochemical assays and fixed tissues were collected for histology.

Results: We created a map of NKCC1 and KCC2 expression and activation during immaturity in piglets with spatial and temporal resolution. We found the toddler piglets more susceptible to disruptions in the balance of NKCC1 and KCC2 following TBI, with a greater effect correlated with cortical hypoxic-ischemic damage versus sparing of subcortical structures.

Conclusions: We demonstrated evidence of the postnatal GABA switch in swine between 7 and 30 days of age and acute disruption of NKCC1 and KCC2 activation following severe TBI in an age-dependent manner. This has key implications regarding the clinical management of children with severe TBI and opens the doors for novel therapeutic paradigms.

O077 Alterations in the circadian circuitry following experimental TBI

Dr Ehsan Mirzakhalili,¹ Dr John Wolf,^1,2 Dr Victoria Johnson,¹ Dr Alexandra Ulyanova ^1,2

¹University Of Pennsylvania, Philadelphia, United States, ²Philadelphia VA Medical Center, Philadelphia, United States

Introduction: Disrupted sleep is a common and persistent symptom after traumatic brain injury (TBI), which can significantly complicate recovery and contribute to cognitive dysfunction. While sleep disturbances such as difficulty falling and staying asleep are often caused by abnormal circadian rhythm, the extent of circadian-related sleep disruption following TBI remains poorly understood. Circadian rhythms are generated by a highly synchronized activity of the GABAergic interneurons in the suprachiasmatic nucleus of the hypothalamus (SCN). Previously, it has been reported that interneurons in the cortex and hippocampus are selectively vulnerable following TBI. However, the effects of TBI on the circadian circuitry and its neuronal network have not been explored.

Methodology: Using a novel porcine model of controlled cortical impact (CCI) injury, we performed in vivo electrophysiological examination of the circadian network at 6hr post-CCI (n=3) versus shams (n=2). MRI-based neuronavigation (Brainsight system) was used to precisely placed multichannel silicone probes into the SCN (128 contacts, 40 μm spacing, 45 mm shank length). Chronic SCN implantations for continuous video EEG recordings were also performed in the sham (n=1) and CCI-injured (n=1) animals, with four ECoG screws placed on both hemispheres for sleep detection.

Results: Under anesthesia, firing rate of the SCN neurons significantly decreased while the spike amplitude significantly increased (p=0.0323). Interestingly, stimulation with light caused activation of sham (p=0.0024), but not CCI-injured SCN interneurons, despite more active cells being detected electrophysiologically. Oscillatory analyses revealed a significant shift in the peak (p<0.0001), but not power of the 8–16 Hz frequencies post-TBI.

Conclusions: We demonstrate that changes in the firing properties of the SCN interneurons may indicate hyperexcitability of the circadian network. Understanding how early changes in the circadian network might contribute to sleep disturbances over time post-TBI will be important to explore using video EEG monitoring in the awake behaving animals.

Support: W81XWH-22-1-0287

O078 Three types of post-traumatic hydrocephalus

Dr Zofia Czosnyka ,¹ Dr Afroditi Lalou, Dr Laurent Gergele, Dr Romain Manet, Mr Adam Pelah, Prof Peter Hutchinson, Dr Marek Czosnyka

¹University Of Cambridge, Cambridge, United Kingdom

Introduction: There are three forms of post-traumatic hydrocephalus (PtH). External hydrocephalus (EH) can be seen in the acute phase. After decompressive craniotomy, gradual ventricular dilatation is frequently observed. Finally, late hydrocephalus, which is difficult to differentiate from generalized brain atrophy, is commonly seen. Our objective was to characterize CSF dynamics in these three scenarios.

Method: 102 patients with traumatic brain injury were examined to detect radiological pictures of external hydrocephalus. Findings were compared to bedside monitoring data of ICP, CPP and PRx, from the first week of NCCU stay.

In patients with craniectomy, CSF dynamics were compared before and after cranioplasty (N=4).

In patients with late PtH, computerised infusion tests were performed. 33 patients with dilated ventricles, atrophic changes, diminished GCS, were compared to patients classified as typical idiopathic normal pressure hydrocephalus (iNPH).

Results: In external hydrocephalus, a pattern of primarily low and then rising ICP was observed before radiological signs of EH were detected (p<0.05). Patients with EH had a worse outcome at 6 months and more frequently needed a shunt at a later stage (p<0.05). The traumatic subarachnoid hemorrhage was the most important risk factor to develop EH.

CSF circulation is not disturbed in all cases of ventricular dilatation after craniectomy and not all patients require shunting. CSF dynamics should be assessed after cranioplasty, as it profoundly changes resistance to CSF outflow and elasticity.

In both late PtH and iNPH, baseline ICP was normal (around 9.5 mmHg). Resistance to CSF outflow was higher in iNPH (17.6+/-5.6 mmHg/(ml/min) than in PtH (13.4+/-6.1 mmHg/(ml/min); p<0.05).

Conclusion: Different types of hydrocephalus may develop after TBI at different timepoints. Each requires bespoke management. External Hydrocephalus - CSF drainage. Ventricular dilatation after craniectomy- cranioplasty and, if CSF dynamics remains disturbed, shunting. Late PtH - shunting when CSF dynamics are clearly disturbed.

O079 Estimating intracranial parameters using an inverse mathematical model with viscoelastic elements closely predicts complex ICP morphologies

Abed Nassir,² Professor Guy Rosenthal,¹ Yuliya Zadka,² Dr. Saadit Houri,¹ Dr. Omer Doron,³ Professor Ofer Barnea ²

¹Hadassah-hebrew University Medical Center, Jerusalem, Israel, ²Tel Aviv University, Israel, ³Massachusetts General Hospital, Boston, USA

Background: The relationship between arterial pressure and ICP waveforms has not been well-described. We hypothesized that adding viscoelastic elements that alter the propagation velocity of pressure waves to a model of intracranial physiology may help account for the observed morphology of the ICP waveform.

Methods: We modified our electrical analog model of the interactions between brain tissue, blood, and CSF within the closed cranium, replacing purely elastic elements with viscoelastic elements. We used an open database, the Cerebral Hemodynamic Autoregulatory Information System Database (CHARIS DB) of traumatic brain injury (TBI) patients. Arterial blood pressure (ABP) and intracranial pressure (ICP) waveforms of patients from the database were used to compare the viscoelastic model with our previous model which does not include viscoelastic elements.

Results: We studied 65 ICP and ABP waveforms in 13 patients. Incorporating viscoelastic elements into the model of intracranial physiology resulted in model-generated waveforms that demonstrated a substantially better fit between measured and simulated ICP waveforms compared to the model without viscoelastic elements. The similarity index increased 160-fold in the model that incorporated viscoelastic elements. Sensitivity analysis showed that the most important viscoelastic elements accounting for the observed ICP waveform morphology were the viscoelasticity of the ventricles, capillaries, and veins.

Conclusion: Our findings suggest that viscoelastic properties of the cerebral tissues and vasculature, and the interactions between them, may help account for the observed morphology of ICP waveforms in severe TBI patients.

O080 Assessing a possible dose or time dependent relationship between physiological signal entropy and outcome in traumatic brain injury

Stefan Yu Bögli ,¹ Ihsane Olakorede,¹ Erta Beqiri,¹ Claudia Ann Smith,¹ Marina Sandra Cherchi,¹ Aiden Chen,¹ Ari Ercole,² Peter Hutrchinson,³ Peter Smielewski¹

¹Brain Physics Laboratory, Division of Neurosurgery, Department of Clinical Neurosciences, Cambridge, United Kingdom, ²Division of Neurosurgery, Department of Clinical Neurosciences, University of Cambridge, Cambridge, United Kingdom, ³Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, United Kingdom

Introduction: Entropy quantifies the level of disorder within a system. Low entropy reflects increased rigidity of homeostatic feedback systems and consequently may reflect underperformance of mechanisms like cerebral autoregulation. In traumatic brain injury (TBI), low entropy of arterial blood pressure (ABP), heart rate (HR), and intracranial pressure (ICP) predict unfavorable outcome. We hypothesized that entropy dosage or duration below specific cutoffs might aid individualized outcome prediction.

Methodology: 575 TBI patients with available 6-month outcome (Glasgow Outcome Scale - GOS) were evaluated. ABP, HR, and ICP entropy was estimated as multiscale entropy (MSE). MSE aggregates entropy at several time scales (20 coarse graining, averaging, steps starting from 0.1 Hz). MSE was calculated repeatedly for consecutive, overlapping 3h segments. Percentage monitoring time (PMT) or dosage (duration*level) below prespecified MSE cutoffs (10, 15, 20) was calculated and evaluated against outcome (GOS 1 vs. 2–5 or GOS 1–3 vs 4–5). Multivariable logistic regression models were built including age, initial GCS, ICP and PRx.

Results: PMT at all 3 cutoffs of ABP, HR and ICP MSE was associated with mortality. The largest differences were found for a cutoff of 10: for ABP 5 [2–10] vs. 9 [3–22]; for HR 11 [5–20] vs. 18 [5–40]; for ICP 19 [8–31] vs. 31 [17–49] for survived vs. expired). Concerning dosage only a cutoff of 10 was associated with mortality (for ABP: 10 [1–27] vs. 17 [4–47], for HR 29 [6–74] vs. 46 [6–121], for ICP 47 [12–105] vs. 74 [16–134] for survived vs. expired). Similar results were found when assessing unfavorable outcome, and the metrics remained significant when added to multivariable models.

Conclusions: PMT of entropy exceeding prespecified cutoffs might aid individualized, dynamic, outcome prognostication, allowing for further exploitation of the extensive physiological data lakes acquired for each TBI patient within an intensive care environment.

O081 Reconstruction based anomaly detection for cerebrovascular autoregulation state monitoring from arterial blood pressure and intracranial pressure signals: a proof of concept study using experimental data

Bavo Kempen ,¹ Samuel Klein, Veerle De Sloovere, Maarten De Vos, Bart Depreitere

¹KU Leuven, Leuven, Belgium

Develop the first deep learning model using reconstruction based anomaly detection to dynamically assess cerebrovascular autoregulation (CA) from arterial blood pressure (ABP) and intracranial pressure (ICP) signals derived from a porcine cranial window CA data set.

Raw ABP and ICP signals were preprocessed and down sampled to 20 Hz. Individual piglets’ Laser Doppler flow (LDF), i.e. cerebral blood flow (CBF), was plotted against cerebral perfusion pressure (CPP) on which the lower -and upper limits (LLA and ULAs) were computed using segmented regression to attain quadriphasic CA state labels. The latter labels were projected on the ABP and ICP time series which were subsequently windowed per 300 seconds. Windowed concurrent ABP and ICP data belonging to the active CA state was used to optimize a TimesNet model to reconstruct the respective ABP and ICP input time series. Performance was evaluated on a separate test data set.

The study confirmed that the optimized model qualitatively reconstructed ABP and ICP segments derived from active CA, with negligible quantitative reconstruction error. ABP and ICP reconstruction errors steadily increased concurrently with LDF in -or decreases. A significant interaction between variable (i.e. ABP and ICP) and CA state showed that ICP reconstruction error increased more than ABP reconstruction error at low CPP inactive CA, while the converse was true for high CPP inactive CA. The developed model also offered improved discriminative ability with respect to the pressure – reactivity index. Input data derived from the four CA states was represented differently in the latent space.

The present work suggests that relevant CA state related information was present in the higher frequencies of ABP and ICP, which can be exploited using advanced representation learning models. In addition, the developed model captured differential behavior of CA with increasing versus decreasing CPP.

O082 Non-Invasive Mapping of Cerebral Autoregulation Using Near-Infrared Spectroscopy

Mr Amanjyot Singh Sainbhi ,¹ Ms Nuray Vakitbilir,¹ Dr. Alwyn Gomez,¹ Mr Kevin Y. Stein,¹ Mr Abrar Islam,¹ Dr Logan Froese,^1,2 Dr Frederick A. Zeiler^1,2,3

¹University Of Manitoba, Winnipeg, Canada, ²Karolinska Institutet, Stockholm, Sweden, ³University of Cambridge, Cambridge, United Kingdom

Introduction: Traditionally, in the neurotrauma literature, cerebral autoregulation (CA) has been assessed in low temporal resolution, limited by large, immobile, and costly neuroimaging platforms. Recent technological advancement has led to continuous methods of assessment, leveraging multi-modal cerebral physiologic devices. However, such continuous assessments have also been hampered by low temporal and spatial resolution systems, that are often reliant on invasive point estimations of pulsatile cerebral blood flow or cerebral blood volume using commercially available technology. To solve this problem, we have developed and tested an entirely non-invasive multi-channel functional near infrared spectroscopy (fNIRS) based CA mapping system, solving both spatial and temporal limitation of previous technologies.

Methodology: Using a custom created monitoring setup, recordings of 1.5 hours on 50 healthy volunteers were performed using a combination of multi-channel (8 channels) cerebral fNIRS and non-invasive arterial blood pressure, across a block-trial design that included breath-holding, neurocognitive testing, and orthostatic trials. Similarly, a user interface was created that visualizes CA metrics by converting them to heat maps drawn on a template of human brain.

Results: The custom monitoring setup was successful in recording all signals in HDF5 format along with sending selected signals to be simultaneously recorded using ICM+ software. The custom Python heat map module works in “offline” mode to visually portray the CA index per channel with the use of colourmap along with a preselected update frequency, facilitating assessment of regional differences.

Conclusion: The generation of the heat maps was entirely non-invasive, with high temporal and spatial resolution. The CA mapping system is in its initial stage and development plans are ready to transform it from “offline” to real-time heat map generation along with integrating it within the custom monitoring platform. Integration of these systems can be leveraged for more comprehensive assessments in injury state such as acute neural injury.

O084 Post-concussion symptom burden and dynamics: insights from a digital health intervention and machine learning

Mr Aimun Jamjoom ,^1,2 Dr Rebecca Blundell,³ Ms Christine d'Offay,² Dr Charles Hand,² Prof Matthew Reed,⁴ Dr David Gillespie^2,5

¹Queens Hospital, Romford, United Kingdom, ²HeadOn Health Ltd, Edinburgh, United Kingdom, ³The National Hospital for Neurology and Neurosurgery, London, United Kingdom, ⁴The Emergency Medicine Research Group Edinburgh (EMERGE), Royal Infirmary of Edinburgh, Edinburgh, United Kingdom, ⁵The Department of Clinical Neuroscience, Royal Infirmary of Edinburgh, Edinburgh, United Kingdom

Objectives: Individuals who sustain a concussion can experience a range of symptoms which can significantly impact their quality of life and functional outcomes. This study aims to understand post-concussion symptomology by applying an unsupervised machine learning approach to data captured from a digital health intervention (HeadOn).

Methods: HeadOn is a digital health intervention developed to support individuals with their recovery following a concussion. As part of the 35-day program, patients complete a daily symptom diary which rates 8 post-concussion symptoms. Symptom data were analysed using K-means clustering to categorize patients based on their symptom profiles.

Results: The study included 94 patients with average age 41(±16); 65.9% were female. During the study period, a total of 758 symptom diaries were completed by 84 patients (average 9 diaries per patient) equating to 6064 individual symptom ratings. Fatigue, sleep disturbance and difficulty concentrating were the most prevalent symptoms reported. A decline in symptom burden was observed over the 35-day period, with physical and emotional symptoms showing earlier rates of recovery. K-means cluster analysis identified three distinct patient clusters based on symptom severity. Cluster 0 (n=24) had a low symptom burden profile across all the post-concussion symptoms. Cluster 1 (n=35) had moderate symptom burden but with pronounced fatigue. Cluster 2 (n=25) had a high symptom burden profile across all the post-concussion symptoms. Qualitative analysis of patient thought diaries revealed themes related to recovery concerns and highlighted the importance of physical activity in participants’ recovery plans.

Conclusions: By leveraging digital ecological momentary assessments, a rich dataset of daily symptom ratings was captured allowing for the identification of symptom severity clusters. These findings underscore the potential of digital technology and machine learning to enhance our understanding of post-concussion symptomology and offer a scalable solution to support patients with their recovery.

O085 Quantitative electroencephalography reveals regional brain dysfunction in mild traumatic brain injury

Miss Caerwen Beaton ,^1,2 Andre Avila,^1,2 Melissa Papini,^1,2 Jacinta Thorne,^1,2 Aleksandra Gozt,^1,2 Francesca Buhagiar,⁵ Elizabeth Thomas,^3,4 Alexander Ring,^6,7 Glenn Arendts,^8,9 John Charles Iliff,^14,15,16,17 Antonio Celenza,^18,19 Sjinene Van Schalkwyk,¹¹ Philip Brooks,^12,13,14 Dan Xu,^3,14,21 Stephen Honeybul,^22,23 Gill Cowen,^1,14 Carmela Pestell,^1,5 Daniel Fatovich,^9,25,26 Ben Smedley,¹⁰ Ashes Mukherjee,²⁰ Michael Bynevelt,²⁴ Melinda Fitzgerald,^1,2 Sarah Hellewell^1,2

¹Curtin Health Innovation Research Institute, Faculty of Health Sciences, Curtin University, Perth, Australia, ²Perron Institute of Neurological and Translational Science, Perth, Australia, ³School of Population Health, Curtin University, Perth, Australia, ⁴Division of Pathology and Laboratory Medicine, School of Medicine, The University of Western Australia, Perth, Australia, ⁵School of Psychological Science, The University of Western Australia, Perth, Australia, ⁶Institute for Immunology and Infectious Diseases, Murdoch University, Perth, Australia, ⁷School of Physiotherapy and Exercise Science, Faculty of Health Sciences, Curtin University, Perth, Australia, ⁸Emergency Department, Fiona Stanley Hospital, Perth, Australia, ⁹Centre for Clinical Research in Emergency Medicine, Harry Perkins Institute of Medical Research, Perth, Australia, ¹⁰Emergency Department, Rockingham General Hospital, Perth, Australia, ¹¹Emergency Department, Joondalup Health Campus, Perth, Australia, ¹²Emergency Department, Saint John of God Midland Public Hospital, Perth, Australia, ¹³School of Medicine, The University of Notre Dame, Perth, Australia, ¹⁴Curtin Medical School, Curtin University, Perth, Australia, ¹⁵Emergency Department, Saint John of God Hospital Murdoch, Perth, Australia, ¹⁶Emergency Department, Royal Perth Hospital, Perth, Australia, ¹⁷Royal Flying Doctor Service- Western Operations, Perth, Australia, ¹⁸Emergency Department, Sir Charles Gairdner Hospital, Perth, Australia, ¹⁹Division of Emergency Medicine, School of Medicine, The University of Western Australia, Perth, Australia, ²⁰Emergency Department, Armadale Health Service, Perth, Australia, ²¹The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China, ²²Department of Health, Government of Western Australia, Perth, Australia, ²³Sir Charles Gairdner Hospital, Royal Perth Hospital and Fiona Stanley Hospital, Perth, Australia, ²⁴The Neurological Intervention & Imaging Service of Western Australia at Sir Charles Gairdner Hospital, Perth, Australia, ²⁵Emergency Medicine, Royal Perth Hospital, Perth, Australia, ²⁶The University of Western Australia, Perth, Australia

Introduction: Individuals with mild traumatic brain injury (mTBI) experience symptoms in domains like physical, psychological, cognitive and sleep-related, though it is unclear whether these symptoms reflect underlying brain dysfunction. This study used quantitative electroencephalography (qEEG) to detect regional brain dysfunction and determine correlation with symptoms following mTBI.

Methodology: 24 participants (36.56 years ±12.87; 10F) with diagnosed mTBI and 30 matched controls (33.88 years ±12.15; 14F) were recruited as part of CREST. 5 days post-injury participants completed the Post-Concussion Symptom Scale (PCSS) and Depression-Anxiety-Stress Scale (DASS-21). qEEG scans (resting-state, eyes-open, 5mins) were acquired using NeuroGuide with a 19-channel Electro-cap (standardised 10–20 placement) and MITSAR-EEG-BT amplifier. Artifact-free data was imported into NeuroNavigator to generate current source density (CSD) for Brodmann areas (BAs). T-tests were used for each BA to determine the effects of mTBI on CSD. Pearson correlations were performed to assess the relationship to PCSS and DASS-21 scores.

Results: Substantial bilateral increases in CSD were detected within the cingulate and retrosplenial cortices across multiple bands: delta, theta, beta and high beta. In the mTBI group, increased CSD in the delta band within the left cingulate was associated with PCSS total score (r = -0.48, p<0.05) and symptom severity (r = -0.53, p<0.05). In the beta band, bilateral increases in CSD in the cingulate (BAs 23, 31) and retrosplenial cortex (BAs 26, 29) positively correlated with DASS-21 anxiety and stress subscale scores (all r>0.44; p<0.05), while unilateral CSD increases correlated with depression scores: right-BA 23, r=0.47, p<0.05; left-BA 26, r=0.52, p<0.05; right-BA 29, r=0.63, p<0.001.

Conclusions: These findings suggest that the cingulate and retrosplenial cortices are specifically vulnerable to mTBI. Early CSD alterations may drive emotional symptom development post-injury, as they correlate positively with various emotional states and are linked to brain regions involved in anxiety, depression, and stress-related memory.

O086 Attention-related functional network changes after mild traumatic brain injury: a longitudinal study

Dr Eunkyung Kim,¹ Dr Min-Yong Lee,² Dr Han Gil Seo,¹ Dr Roh-Eul Yoo,¹ Dr. Byung-Mo Oh ¹

¹Seoul National University Hospital, Seoul, South Korea, ²National Traffic Injury Rehabilitation Hospital, Yangpyeong, South Korea

Introduction: To investigate the cross-sectional and longitudinal changes in functional network connectivity (FNC) and structural network in the dorsal attention network (DAN), ventral attention network (VAN), and default mode network (DMN) after mild traumatic brain injury (mTBI).

Methods: Thirty-one individuals with mTBI (43.7±14.2 years) were scanned twice: within a month and three months after injury. Matched controls (43.2±13.7 years) were scanned twice at similar intervals. After preprocessing the rs-fMRI data, two DAN, two VAN, and three DMN were defined in each hemisphere (Schaefer et al., 2018) and correlation coefficient r was transformed to z between each network (Jafri et al., 2008). Probabilistic tractography was used to construct structural network with the predefined DAN, VAN, and DMN as nodes. A linear mixed effect model was used to investigate changes in FNC and structural network, with group and time as fixed effects and subjects as random effects. A streamline density map corresponding to the structural connection was generated. The relationship between the performance of attention-related tasks and the FNC or fractional anisotropy (FA) of the streamlines was examined using correlation analysis.

Results: The FNC between the DMN and VAN were decreased in the mTBI group compared to the controls, during the acute and chronic phases after injury. In contrast, the FNC between the DMN and DAN were tended to increase in the mTBI group at chronic phase after injury compared to the controls. There was a significant correlation between the performance of attention-related tasks and FNC in the mTBI group compared to the controls. Moreover, in the mTBI group, the FA of the structural connection corresponding to the FNC was not correlated with the attention performance, unlike in the control group.

Conclusion: Our findings highlight the multifaceted nature of mTBI between functional or structural brain changes and attentional impairments.

O087 Specialist Healthcare Services for Concussion/Mild Traumatic Brain Injury in England: A Consensus Statement using the Modified Delphi Methodology

Ms Elika Karvandi ,¹ Mr Adel Helmy, Prof Peter Hutchinson, UK Head Injury Network,

¹University Of Cambridge, Cambridge, United Kingdom

Introduction: The purpose of this study was to establish a consensus on the structure and process of healthcare services for concussion patients in England to facilitate better healthcare quality and patient outcomes.

Methods: This consensus study followed the modified Delphi methodology with five phases; participant identification, item development, two rounds of voting, and a meeting to finalise the consensus statements. Clinical specialists in head injury practising in emergency medicine, neurology, neuropsychology, neurosurgery, paediatric medicine, rehabilitation medicine and sports and exercise medicine in England. All participants are involved in the UK Concussion Network. Two voting rounds were conducted using an online questionnaire, and a virtual meeting was held to finalise the statements. A pre-defined threshold for agreement was set at ≥70% for the two voting rounds. Items not reaching this threshold were eliminated and where reasonable suggestions were made, items were amended after each round of voting. The threshold for consensus was raised to ≥85% for the meeting to ensure robust results.

Results: A total of 55 items were voted on. 30 items were removed following the first round of voting and 3 items were removed following the second voting round. Items were amended where appropriate. The five rounds of this study resulted in a final 18 statements covering three main topics; care pathway to structured follow-up, prognosis and measures of recovery, and provision of outpatient clinics.

Conclusions: This work presents statements on how the structure and process of care for concussion patients in England could be improved to meet their health needs. Future work will seek to implement these into the clinical pathway.

O088 Association of frailty assessment tools and muscularity with GOS and 30-day mortality after TBI in elderly patients – a retrospective study in 1104 patients

Dr. Sebastian Niedermeyer,¹ Dr. Thomas Weig,¹ Mathias Leiber,¹ Aylin Gencer,¹ Professor Sophia Stoecklein,¹ Professor Nicole Terpolilli ¹

¹LMU Munich, Munich, Germany

Background: The number of traumatic brain injuries (TBI) in the elderly population is constantly growing. In this group of patients, frailty is common, since it derives from several age-related conditions.

Objective: This study aimed to compare the prognostic value of different frailty assessment tools and the role of muscularity/sarcopenia as surrogate for frailty in a TBI context.

Methods: In this single-center retrospective analysis, we examined clinical notes of patients aged 60 years and older admitted for traumatic brain injury between 1/2010 and 12/2022. Frailty at admission was assessed using the Clinical Frailty Score (CFS), Charlson Comorbidity Index (CCI), 5-factor modified Frailty Index (mFI-5), and 11-factor modified Frailty Index (mFI-11). Muscularity was quantified at different locations through CT imaging.

Results: A total of 1104 patients with a median age of 78 years (IQR 72–84) were identified. The overall mortality rate was 12.9% (n=137). Spearman correlation analysis revealed a weak correlation of muscle areas with age and frailty scores (Spearman`s rho ≤ 0.1). Multivariate logistic regression models identified the CFS (p<0.0001) and the CCI (p<0.0001) as predictive variables for short term mortality; muscle area measurements as surrogate markers of sarcopenia were not associated with outcome in our cohort. Implementing frailty as measured by frailty scores into prognostic models for short-term mortality increased their predictive power (increase of AUC from 0.897 to 0.939 with CFS and 0.935 with CCI).

Conclusion: Given the demographic change in most industrialized countries, the number of TBI patients with advanced age is constantly growing; age alone, however, is an inaccurate prediction factor. We found that frailty measured by CFS and CCI adds prognostic value, while muscularity at various locations (as assessed in CT imaging) had no effect on 30 day mortality after TBI.

O089 Traumatic brain injury and risk of motor neurone disease: results from the UK Clinical Practice Research Datalink

Dr Xingxing Zhu ,¹ Dr Donald Lyall,¹ Dr Emma Russell,² Dr William Stewart^2,3

¹School of Health and Wellbeing, University of Glasgow, Glasgow, UK, ²School of Psychology and Neuroscience, University of Glasgow, Glasgow, UK, ³Department of Neuropathology, Queen Elizabeth University Hospital, Glasgow, UK

Introduction: Despite extensive research, we have no effective therapies for motor neuron disease (MND). In part, this is may reflect the disease having been active for many years by diagnosis, at which stage the neurodegenerative process is established and irreversible. Attention, therefore, is turning to identifying those at risk of MND and developing strategies to prevent or delay disease onset. An initial aim of this study was to leverage existing population level datasets to explore the association between history of traumatic brain injury and subsequent risk of MND.

Methods: We conducted a matched cohort study using the UK Clinical Practice Research Datalink (CPRD). N=143,730 TBI-exposed individuals were compared with N=287,460 age-, sex- and deprivation-matched ‘control’ comparators (i.e. 2:1 ratio). All participants were aged 18 years or over and registered in CPRD between 1st January 2005 and 31st December 2020. By fitting Cox proportional hazards models, we estimated the relative risk of MND onset in the matched cohort, and how subsequent MND risk varied with time by segmenting follow-up.

Results: During a median of 11·37 years follow-up, 295 incident MND were observed (6.23 per 100,000 person-year). Cox regression showed people with historic TBI were at higher risk for overall MND compared to their comparators (HR = 2·62, 95%CI 2·06–3·35). The effect of TBI was time-dependent: being highest within the first two years following TBI (HR = 29·91, 95% CI 9·28–96·42), then declining incrementally until around 6 years after TBI (>2 to <=4 years: HR = 8·20, 95% CI 3·50–19·19; >4 to <=6 years: HR = 2·10, 95% CI 1·05–4·23; >6 years: ps > 0·05).

Conclusions: A significant association between TBI and subsequent MND incidence was observed, especially within the first two years following TBI. Prevention and effective management of TBI may play a significant role in reducing incidence of MND.

O090 Pre-existing frailty predicts 6-month functional recovery following traumatic brain injury in older-adults: preliminary findings from the TRACK-GERI TBI cohort

Dr. Roy Tzemah-Shahar ,¹ Yael Rosen Lang,¹ AM Puccio,² EL Yuh,³ Domenico Lombardi,⁴ John Boscardin,⁵ Russell Huie,⁴ Kristine Yaffe,^5,6 DO Okonkwo,² Ramon Diaz Arrastia,⁷ GT Manley,⁴ RC Gardner¹

¹Joseph Sagol Neuroscience Center, Sheba Medical Center, Ramat Gan, Israel, ²Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, USA, ³Department of Radiology, University of California San Francisco, San Francisco, USA, ⁴Department of Neurological Surgery, University of California San Francisco, San Francisco, USA, ⁵Department of Epidemiology and Biostatistics, University of California San Francisco, San Francisco, USA, ⁶Departments of Psychiatry and Neurology, University of California San Francisco, San Francisco, USA, ⁷Department of Neurology, University of Pennsylvania, Philadelphia, USA

Introduction: Geriatric traumatic brain injury (TBI) poses a significant public health burden. Acute management decisions are hampered by lack of evidence to support age-appropriate TBI guidelines. Because age and injury severity alone do not consistently predict recovery in older adults, we sought to investigate the association between pre-injury frailty and recovery at 6-months following geriatric TBI.

Methodology: TRACK-GERI TBI is an ongoing two-center prospective study of adults 65+ years old presenting to emergency departments (ED) within 72 hours of all-severity TBI. Baseline characteristics of frail (Groningen Frailty Index, GFI≥4) vs. robust individuals were compared, and the association between baseline GFI (total score) with 6-month full recovery (Glasgow Outcome Scale-Extended, GOSE=8) was tested using logistic regression models, adjusted for demographic and injury factors.

Results: A total of N=99 (age 76.7±7.9, 47% women, 92.9% Glasgow Coma Scale 13–15) individuals were included in the analysis (of which, 19 [19.2%] achieved full recovery at 6mo). No significant differences were observed between frail and robust individuals in baseline age, sex, Comorbidity-Polypharmacy Score, or Rotterdam score. Baseline GFI total score was significantly inversely associated with full recovery (p<0.001, OR 0.61, 95% CI: 0.44–0.83). Additional models, adjusting for age, sex, ED disposition, or Rotterdam, demonstrated similar results.

Conclusion: Pre-injury frailty, as measured by the GFI, was found to be a significant predictor of incomplete recovery in older adults 6 months following TBI, with each additional unit in GFI score being associated with 39% decreased odds of full recovery. While further research in larger international cohorts is needed, these findings highlight the importance of assessing pre-injury frailty in predicting TBI recovery in older adults. The findings additionally raise the testable hypothesis that interventions designed to optimally improve outcomes in this population should not only address the acute TBI itself but should also address pre-existing excess frailty.

O091 Impact of age and mean ICP on the morphology of the ICP pulse waveform and Implications for Outcome in TBI

Prof. Magdalena Kasprowicz ,¹ M.Sc.Eng. Cyprian Mataczyński,¹ PhD Agnieszka Uryga,¹ Prof. Marek Czosnyka,² M.Sc.Eng. Adam Pelah,² PhD Agnieszka Kazimierska¹

¹Department of Biomedical Engineering, Wroclaw University of Science and Technology, Wrocław, Poland, ²Division of Neurosurgery, Department of Clinical Neurosciences, Addenbrooke’s Hospital, University of Cambridge, Cambridge, UK

Introduction: Morphological analysis of the intracranial pressure (ICP) pulse waveform provides indirect information on cerebrospinal compliance, which may decrease with aging and during intracranial hypertension. We aim to investigate the relationship between the shape and amplitude of the ICP pulse waveform and both age and mean ICP in TBI. Additionally, we explore the correlation between these morphological parameters and outcome six months after the ictus.

Methodology: ICP recordings from 186 TBI patients (median age: 51, IQR: 31 years, M: 141) from the CENTER-TBI High-Resolution Sub-Study were retrospectively analysed. ICP pulse shape was assessed using AI–based pulse shape index (PSI) which ranges from 1 (normal pulse waveform) to 4 (pathological waveform). The amplitude of the ICP pulse (AmpICP) was calculated as the difference between the systolic and diastolic peaks within each cardiac cycle. The association between ICP pulse metrics and age and mean ICP was assessed using the Spearman correlation coefficient and main effects ANOVA (for categorized age and ICP levels). Mann Whitney test was used to investigate differences in ICP pulse metrics concerning outcome.

Results: Both age and mean ICP significantly affect PSI (F=25.4, p<<0.001 and F=3.9, p=0.02, respectively), with a stronger correlation observed with age (R=0.50, p<<0.001) compared to ICP (R=0.22, p=0.002). AmpICP is influenced by mean ICP (F=63.7, p<<0.001; R=0.48, p<<0.001) but not by age. Patients who died (n=33) demonstrated higher levels of both PSI and AmpICP (p<0.001 for both), independent of age.

Conclusions: In TBI, ICP pulse shape is affected by advancing age and increased ICP, while pulse amplitude depends solely on mean ICP. Alterations in arterial blood inflow related to stiffening of the vessels can contribute to changes in the waveform shape, suggesting that age should be taken into account while interpreting the ICP pulse morphology. The study was supported by National Science Centre (UMO-2019/35/B/ST7/00500).

O092 Geriatric Acute TBI and Dementias Differ in Blood Biomarker Profiles: Results from the HeadSMART Geriatric Feasibility Study

Dr Damon Kuehl ,¹ Danielle Sandsmark,² Katya Roscovsky,² Ramon Diaz-Arrastia,² Justin Weppner,³ Sally Boyd,⁴ Johanna Leinmueller,⁴ Maggie Barton,⁴ Nazanin Mirshahi,⁴ Timothy Van Meter⁴

¹Virginia Tech Carilion School of Medicine, Department of Emergency Medicine, Roanoke, US, ²Perelman School of Medicine, University of Pennsylvania, Philedelphia, US, ³Virginia Tech Carilion School of Medicine, Department of Internal Medicine, Roanoke, US, ⁴BRAINbox Solutions, Inc, Richmond, US

Objectives: Blood based biomarkers are an area of intense study for identifying neurological disease subtypes. Geriatric subjects evaluated after acute head injury and those with chronic neurological diseases could be evaluated with the same blood biomarkers when approved for clinical use. Assessing mild traumatic brain injury (mTBI) within the context of cognitive decline necessitates objective measurements tailored specifically to mTBI in the geriatric population. The HeadSMART Geriatric Study was designed to derive a multimodality point-of-care testing system (digital neurocognitive battery + biomarker multiplex) for elderly patients with suspected mild TBI (mTBI), irrespective of cognitive status.

Methods: Subjects (ages 65+) were enrolled at Carilion Medical Center (Roanoke, VA) and Penn Presbyterian Medical Center (Philadelphia, PA). Subjects (n=101 mTBIl) were assessed by neurocognitive/dementia tests, digitized neurocognitive battery (app), and immunoassays. Review for TBI diagnosis and pre-injury cognitive and dementia stage (CDR and FAQ informant sections) was performed. Dementias (n= 56), were compared with TBI for levels of 15 biomarkers, encompassing glial, neuronal, inflammatory, and vascular biology. Biomarker differences were analyzed by ANOVA, and diagnostic models distinguishing acute mTBI from dementia were derived in in R with random forest.

Results: Biomarkers that differed between geriatric TBI and dementias were Interleukin-6, ST2, Phospho-181Tau, pS217-Tau, and von Willebrand Factor (ANOVA, p <0.05). Random Forest models utilizing 3 biomarker subsets had clinically useful performance (e.g., AUCROC of 0.896, Sensitivity 85%, Specificity 83%) in distinguishing mTBI from dementia. All top classifiers included ST2 and phosphoT181-Tau in this study, and variably included GFAP, SNCA, or BDNF.

Conclusions: The initial findings need to be tested in larger sample sets and in independent cohorts. The next phase of the study has a longitudinal study design with DTI-MRI and, DCE-MRI to assess the relationship between axonal and vascular pathology blood biomarkers, trajectories of cognitive decline, and persistent symptoms post-injury.

O093 Traumatic brain injury specific alterations in plasma proteome and altered cholesterol metabolism in UK service personnel an average of 8 years post injury: The ADVANCE cohort

Miss Grace Blissitt ,^1,3,4 Mr Angelos Manolias,⁵ Dr Neil Graham,^1,2 Dr Kanta Chechi,^3,5 Dr Karl Zimmerman,^1,2 Dr Lucia Li,^1,2 Ms Susie Schofeild,³ Prof Nicola Fear,^6,7 Prof Christopher Boos,^4,7,8,9 Prof Anthony Bull,^10,11 Gp Capt Alexander Bennett,^4,11 Prof Marc-Emmanuel Dumas,^3,5 Prof David Sharp^1,2,10

¹Department of Brain Sciences, Imperial College London, London, United Kingdom, ²UK Dementia Research Institute Centre for Care Research and Technology, Imperial College London, London, United Kingdom, ³National Heart and Lung Institute, Imperial College London, London, United Kingdom, ⁴Academic Department of Military Rehabilitation, Defence Medical Rehabilitation Centre, Stanford Hall, Loughborough, United Kingdom, ⁵Department of Metabolism, Digestion and Reproduction, Imperial College London, London, United Kingdom, ⁶King's Centre for Military Health Research, King's College London, London, United Kingdom, ⁷Academic Department for Military Mental Health, King's College London, London, United Kingdom, ⁸Faculty of Health & Social Sciences, Bournemouth University, Bournemouth, United Kingdom, ⁹Department of Cardiology, University Hospitals Dorset NHS Foundation Trust, Poole Hospital, Poole, United Kingdom, ¹⁰Centre for Injury Studies, Imperial College London, London, United Kingdom, ¹¹Department of Bioengineering, Imperial College London, London, United Kingdom

Introduction: Traumatic brain injury (TBI) is associated with increased mortality, poor cardiovascular health and is a risk factor for neurodegenerative diseases. Proteomic analysis of blood samples in the chronic phase post-TBI may help clarify the cause of long-term effects of TBI.

Methodology: Fasted plasma samples from 1104 male military personnel, deployed to Afghanistan (2003–2014), were analysed using the SomaScan v.4.1 assay. 1160 aptamers with protein targets reported to have cis-pQTL when measured by either Soma or Olink platforms were analysed. Data was inverse rank normalised and ANCOVA models adjusted for age, military rank and ethnicity were used to compare protein levels across three groups; TBI (n=95), Polytrauma (n=458), No Injury (n=551) determined by medical history and Joint Trauma Theatre Registry data. Kruskal-Wallis test and Spearman’s correlation were used to compare cholesterol levels across groups and infer associations between TBI-associated aptamers and cardiovascular health markers respectively.

Results: 11 aptamers were significantly different between groups after multiple comparison correction (FDR<0.05). PCSK9 and PTK7 levels were significantly elevated in TBI compared to other groups. Plasma TIMP-2, CRAC1, CD248, Notch-3, EGFLA, Gelsolin, TGF-b Receptor-III, and ENPP5 levels were significantly reduced in Polytrauma compared to No Injury and TBI. NAR3 levels were significantly lower in Polytrauma and TBI. Blood low-density lipoprotein (LDL) levels were significantly different across groups, due to higher levels in TBI than Polytrauma. PSCK9 levels had a significant positive correlation with LDL, HDL, and total cholesterol. PTK7 was also significantly positively correlated with VEGF-A, and VEGF-C.

Conclusion: PCSK9 and PTK7 are involved in cardiovascular function. Elevated circulatory PCSK9 increases the degradation of LDL receptors resulting in increased plasma LDL. PTK7 acts as a co-receptor in multiple pathways including VEGF signalling pathways, which regulate blood brain barrier permeability. Results suggest altered proteomic signature ∼8 years after TBI may relate to increased cardiovascular risk.

O094 Exploring Neural Correlates Of Ibogaine In Special Forces Combat Veterans Through Multimodal Imaging

Dr. Azeezat Azeez,¹ Ms. Malvika Sridhar,¹ Mr. Andrew Geoly,¹ Dr. Afik Faerman,¹ Dr. Kirsten Cherian,¹ Dr. John Coetzee ,^1,2 Ms. Saron Hunegnaw,¹ Dr. Derrick Buchanan,¹ Dr. Jackob Keynan,¹ Dr. Ian Kratter,¹ Dr. Cammie Rolle,¹ Dr. Manish Saggar,¹ Dr. Maheen Adamson,^1,2 Dr Nolan Williams¹

¹Stanford University, Stanford, United States, ²VA Palo Alto Health Care System, Palo Alto, United States

Objective: This analysis sought to identify the neural mechanisms underlying the strong therapeutic results from a recent study that evaluated the safety and clinical impact of ibogaine in treating military veterans with traumatic brain injury (TBI). TBI is a leading cause of disability with sequelae of psychiatric symptoms such as post-traumatic stress disorder (PTSD), major depressive disorder (MDD), and generalized anxiety disorder (GAD).

Methods: We collected arterial spin labeling (ASL) and blood-oxygen-level-dependent (BOLD) functional magnetic resonance imaging (fMRI) data at three-time points pre and post-treatment on 30 Special Operations Veterans (SOV) who had voluntarily enrolled in tabernanthe iboga exposure at a clinic in Mexico. We used a multimodal whole-brain resting-state exploratory approach of examining changes to regional Cerebral Blood Flow, Functional Connectivity, and Network communication to characterize neural features that were altered post-ibogaine treatment.

Results: Significant changes were identified in blood flow (p<0.001, PFDR<0.05), functional connectivity (p<0.005), and networks of the limbic and sensory-motor system, regions associated with TBI and PTSD. We found associations between neuroimaging findings in the left hemisphere insula, anterior cingulate cortex, and hippocampus-dorsal attention network with clinical measures of disability index and PTSD symptomology.

Conclusions and Relevance: Our novel multimodal neuroimaging approach revealed potential mechanisms underlying the therapeutic benefits of ibogaine for SOV suffering from TBI with comorbid disability and psychiatric symptoms. Further research with larger and diverse populations would be beneficial to establish clinical and neuroimaging alterations from ibogaine on subjects without lifetime TBIs or combat-induced PTSD.

O095 Diffusion Along the Perivascular Space in Military Veterans with Post Traumatic Stress Disorder and Mild Traumatic Brain Injury – Associations with Sleep Quality Disturbances

Lisa Ha Nguyen ,^1,2 Dr. Philine Rojczyk,^1,2 Leonard Ben Jung,^1,2 Alberto Villigran,¹ Nicholas Kim,² Zheyuan Li,¹ Tashrif Billah,² Luisa Berger,^1,2 Ariel Hyunseo Kim,^1,2 Dr. Catherine Brawn Fortier,^3,4 Dr. David Salat,^3,4,5 Prof. Dr. William Milberg,^3,4 Prof. Dr. Martha Elizabeth Shenton,^2,4 Prof. Dr. Inga Katharina Koerte^1,2,4

¹cBRAIN, Department of Child and Adolescent Psychiatry, Psychosomatics, and Psychotherapy, Ludwig-Maximilians-Universität, Munich, Germany, ²Psychiatry Neuroimaging Laboratory, Department of Psychiatry, Brigham and Women’s Hospital, Boston, United States of America, ³Translational Research Center for TBI and Stress Disorders and Geriatric Research, Education and Clinical Center, VA Boston Healthcare System, Boston, United States of America, ⁴Department of Psychiatry, Harvard Medical School, Boston, United States of America, ⁵Athinoula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital, Charlestown, United States of America

Introduction: Mild traumatic brain injury (mTBI) and post-traumatic stress disorder (PTSD) are known as signature wounds of military service. Poor sleep quality is highly prevalent and associated with worse long-term outcome following mTBI and in PTSD. However, the underlying pathomechanisms are not fully understood. There is initial evidence that poor sleep may affect brain clearance. Here, we investigate whether purported imaging measures of the brain-wide clearance system, are associated with mTBI and PTSD.

Methodology: Diffusion Tensor Imaging (DTI) and clinical data were obtained from male veterans from the Translational Research Center for TBI and Stress Disorders (TRACTS) study. A total sample of n=176 participants was included in the analysis, categorized into groups of veterans with mTBI, n=24; PTSD, n=37; PTSD+mTBI, n=92; or controls, n=23. Diffusion along the perivascular space was assessed using the DTI Along Perivascular Spaces (DTI-ALPS) algorithm. The ALPS-index, defined as the ratio of the diffusivity along the perivascular space to the diffusivity perpendicular to both the major fiber tract and the perivascular space, was compared between the groups using analyses of covariance. Linear regression models were used to test the association between ALPS-index and sleep quality (Pittsburgh Sleep Quality Index).

Results: Veterans with mTBI, PTSD, or both PTSD+mTBI showed a decreased ALPS-index compared to controls (p < 0.001). Lower sleep quality was associated with a lower ALPS-index (p = 0.012).

Conclusion: Results of this study demonstrate that mTBI and PTSD are associated with lower ALPS, a purported measure of brain clearance. ALPS in turn, is associated with poor sleep quality. While mTBI has a minor impact on the ALPS-index, its presence alongside PTSD intensifies the effects, likely due to its long-lasting impact on sleep quality. Further research is needed to investigate sleep as a potential therapeutic target for brain clearance in the context of mTBI and PTSD.

O096 Sex-dependent differences in acute phase serum NfL levels in rats subjected to penetrating TBI

Mr Erik Lidin ,¹ Dr Johan Davidsson,² Dr Kaj Blennow,^3,4 Dr Henrik Zetterberg,^3,4,5 Dr Mårten Risling,¹ Dr Mattias K. Sköld^1,6

¹Experimental Traumatology, Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden, ²Vehicle Safety Division, Department of Applied Mechanics, Chalmers University of Technology, Gothenburg, Sweden, ³Department of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden, ⁴Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Stockholm, ⁵Department of Neurodegenerative Disease, Institute of Neurology, University College London, United Kingdom, ⁶Section of Neurosurgery, Department of Neuroscience, Uppsala University, Uppsala, Sweden

Introduction: Neurofilament light chain (NfL) is candidate biomarker of axonal damage and neurodegeneration after TBI. Multiple studies suggest female rodents have superior morphological outcome compared to male after TBI, and that female sex-hormones may be contributing factors. Whilst serum NfL have a role in injury detection and outcome prognosis after mild TBI, the influence of injury mechanism and severity on serum NfL remains unknown. The influence of biological sex on temporal dynamics and correlation to axonal damage in the acute phase of severe TBI are unknown but important to further understand physiological differences of relevance for injury amelioration and clinical translation.

Materials and Methods: 60 adult Sprague-Dawley rats divided based on biological sex and estrus cycle phase received a jugular vein catheter and a subsequent penetrating TBI or sham-surgery. Serum was collected prior to injury, and 24-, 72-, 120-hours post injury. Vaginal swabs were performed on female rodents in conjunction with serum sampling to determine estrus cycle phase at each timepoint. Serum NfL was quantified using Simoa.

Results: Serum NfL was higher in females subjected to penetrating TBI compared to male counterparts at every assessed timepoint following injury. The difference in serum NfL between the sexes increased over time. 120-hours after injury, females exhibited significantly higher serum NfL levels than male counterparts (p = 0.045, Welch’s unpaired t-test).

Discussion: As a candidate biomarker of TBI, understanding of the influence of subject characteristics and injury mechanism on serum NfL is warranted. Our findings suggest biological sex influences serum NfL levels following penetrating TBI, with females exhibiting higher serum NfL which may suggest differences in injury susceptibility, pathophysiology, or biomarker clearance.

O097 Machine Learning Approach to Identify the Pathobiology of Repeated Exposures to Subconcussive Level of Primary Explosive Blast

Professor Denes Agoston ,¹ Mr. Jesse McCullough, Ms. I-Hsuan Lin, Mr. Michael Eklund, Mr. Wallace Graves, Mr. Cyrus Dunbar, Dr. James Engall, Dr. Eric Schneider, Dr. Fabio Leonessa, Dr. Josh Duckworth

¹USUHS, Bethesda, United States

The widespread use of powerful, high-energy explosives during military conflicts and military training exercises causes a unique form of injury called blast induced neurotrauma (BINT) [1]. Primary BINT caused by exposure to the blast wave itself can cause endothelial stress / injury triggering complex downstream biological responses, vascular remodeling and neuroinflammation that may lead to chronic neurological and neuropsychiatric conditions.

The most frequent form of primary BINT is mild/sub-concussive but repeated sub-concussive blast exposure, (RSCBE) associated with heavy weapons training (HWT) and breaching exercises can induce long-term neuropsychological problems, however the pathobiology is currently unknown. In order to identify the molecular changes in response to RSCBE, we have analyzed blood samples collected at various post exposure time points for changes in protein biomarker levels of GFAP, IL-1B, IL-6, NFL, pTau, S100B, Tau, UCHL1, vWF, CHRNA7, CLDN5, IFNg, MMP9, OCL, TNFa, VEGFa, AQP4, HMGB1, HSP70, VCAM representing markers of neuro-, glia, axon, vascular/endothelial injuries, and inflammation [2]. We then have utilized a supervised decision tree machine learning algorithm to determine a combination of time point and biomarker rules that best separates out the various exposure groups. The training data set was used to train the model, which we then tested on the test data set using R programming language. We have generated decision trees PER post-injury timepoints that have shown a) time dependent evolution of the pathobiological responses to RSCBE b) led by vascular / endothelial injury and stress and inflammatory processes and c) indications that the pathobiological responses are “dose-dependent”.

Combined with functional outcome measures, frequency and amplitude of blast exposures, the biomarker data will enable to develop an algorithm aimed to optimize training protocols and to maintain and improve force readiness and to mitigate potential long-term adverse outcomes.

O098 Predicting subacute autoantibody production through acute immunological protein alterations in moderate to severe human traumatic brain injury

Miss Julie Cheung,¹ Dr Emma Hammarlund,^1,2 Dr Sofia Bergström,³ Dr Philipp Lassarén,¹ Dr Sei Yon Sohn,⁴ Professor Peter Nilsson,³ Associate Professor Eric Thelin,^1,5 Dr Edward J. Needham,^4,8 Dr Caroline Lindblad ^1,4,6,7,8

¹Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden, ²Section Perioperative monitoring and intensive care, Karolinska University Hospital, Stockholm, Sweden, ³Division of Affinity Proteomics, Department of Protein Science, SciLifeLab, KTH-Royal Institute of Technology, Stockholm, Sweden, ⁴Department of Clinical Neurosciences, University of Cambridge, Cambridge, United Kingdom, ⁵Medical unit Neurology, Karolinska University Hospital, Stockholm, Sweden, ⁶Department of Medical Sciences, Uppsala University, Uppsala, Sweden, ⁷Medical unit Neurosurgery, Karolinska University Hospital, Stockholm, Sverige, ⁸These authors contributed equally

Introduction: Traumatic brain injury (TBI) triggers neuroinflammation at both acute and chronic stages. Neuroinflammation affects patient outcome deleteriously and could lead to neurodegeneration. Interestingly, the link between early neuroinflammatory responses and subsequent chronic conditions, characterized by autoantibody production, is not fully understood. This study aimed to examine if acute immunological protein alterations can predict subacute autoantibody profiles following moderate to severe human TBI.

Methodology: We conducted a prospective observational cohort study analyzing serum samples from n = 19 patients with moderate to severe TBI and n = 12 healthy controls. Utilizing bead-based protein microarrays, we quantified n = 177 acute proteins and n = 79 IgM and IgG autoantibodies associated with neuroinflammation, in acute (day 0–3 post injury) and subacute (day 7 post injury) samples. Data were analyzed using raw median fluorescence intensity (MFI) and Z-score-converted values. Linear mixed-effects models and volcano plots were utilized to identify upregulated subacute autoantibodies, which were further assessed through multivariable regression analysis with principal components (PC) of acute proteins.

Results: We found significant elevations of IgM autoantibody expression at both acute and subacute phases compared with controls. Among the autoantibodies assessed, 23 were significantly increased. Notably, autoantibodies against α-synuclein, T-cadherin, immunoglobulin heavy constant gamma 1, and Gamma-aminobutyric acid type B receptor subunit 1 could be predicted utilizing PCs from the acute proteins. Particularly, α-synuclein autoantibodies, known to induce neuroinflammation by production of pro-inflammatory cytokines and associated with various neurodegenerative diseases including Parkinson’s and Alzheimer’s disease, were predicted by PC2 (p=0.05). T-cadherin autoantibodies, known markers of autoimmunity, were predicted by PC1 (p=0.04) of acute proteins.

Conclusions: This study demonstrates a relationship between acute protein alterations and subacute autoantibody levels following human TBI, highlighting early inflammatory responses as potential precursors to chronic neuroinflammation, of importance for future therapeutic investigations.

O099 Metabolic shifts are associated with NOX activity in proinflammatory microglia

Dr. Nathan Strogulski ,¹ Ms. Janeen Laabei,¹ Ms. Carly Douglas,¹ Ms. Sara V. Fraj,¹ Mr. Patrick Griffin,¹ Mr. Sahil Threja,¹ Dr. Gloria Vegliante,¹ Dr. David J. Loane¹

¹Neurotrauma and Neuroimmunology Research Group, School of Biochemistry and Immunology, Trinity College Dublin, Dublin, Ireland

NADPH-oxidase (NOX) 2 sustains microglial inflammation post-traumatic brain injury (TBI),and may rely on metabolic adjustments, particularly in the pentose-phosphate-pathway (PPP), to meet heightened NADPH demands. Understanding shifts in microglial metabolism may offer immunotherapeutic opportunities for TBI. Our objective was to identify metabolic changes required for microglial NOX activation using in vitro and in vivo TBI models.

Immortalized microglia (IMG) were stimulated with LPS (100 ng/mL) and/or pentose-phosphate pathway (PPP) inhibitor, 6-aminonicotinamide (6-AN; 200 μM), for 24 hours, followed by ATP (1mM) for 10 minutes. NOX and G6PD activities were measured by lucigenin chemiluminescence and colorimetric kit respectively, and cytokine release (TNF-α, IL-1β) by ELISA. TMRM explored mitochondrial membrane potential (MMP), and SeaHorse analysis assessed oxygen consumption (OCR) and glycolysis (ECAR). RT-qPCR evaluated transcripts of Hk2, Cybb, and G6pd2. Further, tamoxifen-treated control (Nox2Flox) and microglia-specific NOX2 conditional knockout (HexBCreERT2:Nox2Flox) mice were subjected to sham or moderate-level controlled cortical impact (CCI). Seven days post-injury, brain hemispheres were collected, and microglia isolated for gene expression (Hk2, Il1, Hk2 and G6pd2) and contralateral tissues for mitochondrial content and MMP using flow cytometry.

LPS+ATP stimulation of microglia increased ECAR, decreased MMP and OCR, and upregulated Hk2 and G6pd2 expression. Also, LPS+ATP increased both G6PD and NOX activity, proinflammatory cytokine release, and Cybb expression in microglia. 6-AN mitigated G6pd2 and Cybb upregulation, attenuating NOX activity, proinflammatory cytokine release and glycolytic metabolism, restoring OCR. A negative correlation between MMP and cytokine levels identified a link between immune response and microglial metabolism. In vivo, selective microglial NOX2 knockdown attenuated CCI-induced Il1b, G6pd2 and Hk2 upregulation, paralleled by increased mitochondrial content and preserved membrane potential, indicative of enhanced oxidative metabolism. These findings highlight PPP's crucial role in supporting proinflammatory microglial activation and NOX2 activity, shedding light on microglial metabolic dynamics post-TBI, which may be targeted to modulate post-traumatic neuroinflammation.

O100 Exploratory Study of Short Chain Fatty Acid Supplementation on Anti-Inflammatory Tricarboxylic Acid Cycle Metabolites after Traumatic Brain Injury

Dr Amanda Dave ,¹ Mona Chatrizeh,⁵ Dr. Hulya Bayir,² Keri Janesko-Feldman,³ Vincent Vagni,³ Dr. Jianmin Tian,⁴ Dr. Robert Clark,^1,3 Dr. Patrick Kochanek,^1,3 Dr. Michael Morowitz,⁴ Dr. Dennis Simon^1,3

¹Department of Critical Care Medicine, University of Pittsburgh School of Medicine, PITTSBURGH(PA), United States, ²Columbia University, New York City, United States, ³Safar Center for Resuscitation Research, University of Pittsburgh School of Medicine, Pittsburgh, United States, ⁴Department of Surgery, University of Pittsburgh School of Medicine, PITTSBURGH, United States, ⁵Department of Pathology, University of Pittsburgh School of Medicine, PITTSBURGH(PA), United States

Objectives: Short chain fatty acids (SCFA) are produced by microbial fermentation of dietary fiber and serve as an energy substrate for glial cells through the tricarboxylic acid (TCA) cycle. In the controlled cortical impact (CCI) model of traumatic brain injury (TBI), SCFAs are reduced after injury and SCFA supplementation is protective and anti-inflammatory. We hypothesized that SCFA supplementation increases production of the potent anti-inflammatory TCA metabolites itaconate and 2-hydroxyglutarate.

Methods: CCI was performed in adult male C57BL6/J mice (6m/s, 2.0mm). Shams received anesthesia and skin incision. Mice (n=3/group) were randomized to SCFA supplemented water (0.067M acetate, 0.047M butyrate) or standard drinking water provided ad libitum 14 days before CCI and continued until sacrifice 7 days post-injury. Pericontusional tissue was sent for relative quantification of TCA metabolites by liquid chromatography-mass spectrometry and data were normalized to vehicle treated shams (Sham-Veh).

Results: CCI and SCFA supplementation were associated with multiple changes in TCA metabolites. Itaconate level decreased by 16.67% after CCI (Sham-Veh 1.00±0.16 vs. CCI-Veh 0.81±0.05; p=0.1, T-test). SCFA supplementation restored itaconate levels after CCI (CCI-Veh 0.81±0.05 vs. CCI-SCFA 0.97±0.04; p<0.05). 2-hydroxyglutarate decreased 29% after CCI (Sham-Veh 0.99±0.02 vs. CCI-Veh 0.71±0.05; p<0.01). SCFA produced a trend toward increased 2-hydroxyglutarate levels after CCI (CCI-Veh 0.71±0.05 vs. CCI-SCFA 0.83±0.13; p=0.2).

Conclusions: In this exploratory study, SCFA supplementation restored the level of the anti-inflammatory metabolite itaconate to sham levels after CCI. Further studies are needed to confirm these results, delineate the pathway implications, and identify whether this impacts neurologic outcomes.

O101 HMGB1 localization following modeled traumatic brain injury: An acute and chronic response

Laura Siracusa ,^1,2 Dr. Eugene Park,² Dr. Elaine Liu,² Dr. Andrew Baker^1,2,3

¹Institute of Medical Sciences, University Of Torotno, Toronto, Canada, ²Trauma Research, Keenan Research Center, Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, Canada, ³Department of Anesthesia & Surgery, University of Toronto, Toronto, Canada

Introduction: Under normal physiological conditions high mobility group box protein 1 (HMGB1) stabilizes chromatin, controls transcription, and contributes to DNA repair. However, following injury and release from the nucleus, HMGB1 acts as a proinflammatory cytokine. Its many inflammatory functions make HMGB1 a candidate for the modulation of inflammation following traumatic brain injury (TBI). The objective of this study was to evaluate the role of HMGB1 in neuroinflammation following TBI. We began with characterizing the localization of HMGB1 at both acute and subacute time points following modeled TBI in rats. We sought to determine whether translocation coincided with pro-inflammatory activation at the area of injury.

Methods: Brain tissues from rats subject to fluid percussion injury (FPI) were collected for Western blot and immunofluorescent analysis. Morphological changes in HMGB1 expression and localization in various cell types was examined at 6 hours, 24 hours, 7 days, and 14 days after injury.

Results: There was a significant loss of HMGB1 from the nucleus of neurons at 6 and 24 hours with cells showing translocation to the cytoplasm. At 7 days post-injury, HMGB1 expression persisted within the cytoplasm, and HMGB1 loss was observed in the neuronal population lasting two weeks. Changes in cytoplasmic NF-κB was observed during the first week indicating pro-inflammatory activity.

Conclusion: A reduction in nuclear expression of HMGB1 was initiated by modeled TBI. Post-traumatic HMGB1 translocation within the cortex of animals was observed at acute and subacute time points. HMGB1 loss in neurons was seen up to 2 weeks following injury. These changes coincided with NFκB loss in cytoplasmic fractions. This suggests HMGB1 translocation is involved in the proinflammatory response and may be involved in a secondary injury mechanism following TBI. Therapeutic targeting of this protein may allow for modulation of inflammation and reduction in secondary injury.

O102 Endothelial inflammation is linked to intracranial physiology following human severe traumatic brain injury

Dr Caroline Linblad,^1,2,3,6 Ms Claudia Smith ,^1,6 Dr Erta Beqiri,¹ Dr Sofia Bergström,⁴ Dr Edward J. Needham,⁵ Dr Peter Smielewski,¹ Dr Peter Nilsson,⁴ Dr Eric Thelin,² Mr Adel Helmy¹

¹Department of Clinical Neurosciences, University Of Cambridge, Cambridge, United Kingdom, ²Karolinska Institutet, Stockholm, Sweden, ³Uppsala University, Uppsala, Sweden, ⁴SciLifeLab, KTH-Royal Institute of Technology, Stockholm, Sweden, ⁵Department of Anaesthesia, University Of Cambridge, Cambridge, United Kingdom, ⁶these joint first authors contributed equally

Introduction: Elevated intracranial pressure (ICP) and disturbed cerebrovascular reactivity, as indicated by pressure reactivity index (PRx), commonly occur after severe traumatic brain injury (sTBI) and indicate vulnerability to, and ongoing, secondary injury. The influence of neuroinflammation, particularly endovascular inflammation, during these processes is largely unknown. We aimed to investigate neuroinflammatory analytes from the blood of patients with sTBI, and associate this with ICP and PRx in the first week post injury.

Methodology: We analysed 179 proteins in plasma samples over the first week post ictus from 20 sTBI patients using an antibody-based suspension bead array (FlexMap 3D Luminex). Dimensionality reduction techniques were applied to the 179-analyte screen, with waveform level physiological data (ICP and PRx) available in a subset of 14 patients. Relationships between inflammatory analytes and physiological variables (mean, thresholds, and dose above thresholds) were investigated with Spearman’s correlation, principal component (PC) and regression models.

Results: PC analysis showed distinct clustering patterns with respect to ICP and PRx thresholds and dose. Further investigation of PCs implicated time from injury for PC1, while PC2 and 3 were significantly associated with PRx dose (p-values < 0.001) in a linear mixed effect model with patient and time as random effects. Notably, vascular inflammation, as previously defined by our group1 represented 20% of the top 50 proteins driving loadings of PC2, confirmed by supervised PCA which implicated complement inflammation in these processes. Dimensionality reduction of proteins associated with vascular inflammation indicated complement protein C9 as the main driver of PC2.

Conclusion: Endovascular inflammation is associated with dysregulated intracranial dynamics within the first week post ictus in sTBI. This preliminary investigation highlights the vascular inflammation as a possible contributor to elevated ICP and disturbed vascular reactivity in sTBI. Specifically, the complement pathway should be investigated in the context of brain physiology in sTBI patients.

O103 Reliability assessment of the Liverpool Head Injury Tomography Score (Liverpool HITS) for mild traumatic brain injury

Miss Sandra Sungailaite ,¹ Mr George Bonanos,¹ Prof. KS Manjunath Prasad,¹ Dr Laura Evans,¹ Dr Amy Verrinder,¹ Dr Tanmay Sukthankar¹

¹James Cook University Hospital, Middlesbrough, United Kingdom

Introduction: The Liverpool Head Injury Tomography Score (HITS) is a novel CT-based scoring system developed to be used by various healthcare professionals to determine surgically significant mild traumatic brain injuries (TBI) and potentially reduce inappropriate referrals to neurosurgery. By giving a numerical value for the radiological injuries demonstrated on the CT scan, patients are stratified into a non-surgically significant group which could be managed locally and a surgically significant group which would warrant neurosurgical referral. We aimed to assess Liverpool HITS score inter- and intra-observer variability among a neurosurgery registrar, emergency department consultant and radiology registrar.

Methodology: 53 referrals made to a tertiary neurosurgical centre with mild TBI (Glasgow Coma Scale on presentation 13–15) were identified on the on-call online referring system referrapatient.org. The sample aimed to include a variety of TBI. CT scans from the referrals were scored using the Liverpool HITS scoring system by a neurosurgery registrar, A&E consultant, and radiology registrar. Each observer evaluated the same set of scans twice, with a minimum of 6-week interval between assessment. The outcome of the referral was determined for all referrals. Inter-observer and intra-observer agreement were assessed by the intraclass correlation coefficient (ICC) using SPSS v24.0 (IBM, Armonk, NY, USA).

Results: There was a very high level of agreement among all observers (interobserver agreement) with the ICC measuring 0.86 (95% confidence interval). When comparing the first and the second sets of readings, the intra-observer agreement was almost perfect (0.89–0.99). All patients scored 2 or below by any of the observers at any point were managed locally.

Conclusions: The study demonstrates a high level of agreement when using the Liverpool HITS scoring system by several healthcare professionals. Our study results support the potential implementation of the system into local policies to reduce inappropriate referrals to neurosurgery.

O104 Outcome prediction after severe traumatic brain injury: developing a novel prognostic model by adding prehospital variables from the BRAIN-PROTECT Study

Christopher Ryalino,¹ Jelmer-Joost Lenstra ,¹ Dr. Bas Bossers,² Dr. Patrick Schober,² Prof. Dr. Joukje van der Naalt,¹ Prof. Dr. Anthony Absalom,¹ and BRAIN-protect investigators²

¹University Medical Center Groningen, Groningen, Netherlands, ²Amsterdam University Medical Centre, Amsterdam, Netherlands

Introduction: Severe traumatic brain injury (TBI) is a major public health problem with high mortality and disability rates that necessitate early identification of risk factors that are modifiable by treatment to prevent secondary brain injury and improve outcome. Two large prognostic models (IMPACT and CRASH) have identified important risk factors which can discriminate patients at risk for mortality and unfavorable outcome. The aim of the study was to develop a novel outcome prediction model for severe TBI including prehospital variables.

Methodology: We used data from the BRAIN-PROTECT study, a prospective, observational study that included 2.589 patients with severe TBI. Key outcome predictors of 30-day mortality, 6-month mortality, and 12-month extended Glasgow Outcome Scale (GOSE) were used. For each outcome measure, two prediction models were developed, a first model including only prehospital variables (age, GCS, pupil reactivity), and a second model expanded with intrahospital variables (Hb, glucose, pH, APTT, ISS, Rotterdam CT-score).

Results: Median [IQR] age of patients were 46 [24 – 65] years. Data on 30-day mortality, 6-month mortality, and 12-month GOSE were available for 2.101 (82%), 1,833 (70.8%) and 1.243 (48%) patients respectively. The expanded models performed better than the prehospital model in predicting 30-day mortality (AUC=0.918 vs 0.831, p <0.001), 6-month mortality (AUC=0.917 vs 0.824, p <0.001), and 12-month GOSE (AUC=0.921 vs 0.839, p <0.001). All models showed good discrimination (c-statistics >0.80).

Conclusions: Known key outcome predictors already showed good calibration and discrimination in prediction outcome after severe TBI. Addition of intrahospital variables resulted in some improvements in prediction performance. Our results may give clinicians more insights in providing early predictions of the prognosis of severe TBI.

O105 The use of automated CT analyses in the acute phase of patients with traumatic brain injury

MD Robin Knepflé ,¹ MD Hugo Den Boogert,² Jan Verheyden,³ MD PhD Godard de Ruiter,² MD Anke van der Eerden,⁴ MD PhD Thomas van Essen,⁵ MD PhD David Menon,⁶ PhD Hester Lingsma,⁷ MD PhD Ronald Bartels⁸

¹Department of Neurology, Maastricht University Medical Center, Maastricht, Netherlands, ²Department of Neurosurgery, Haaglanden Medical Center, The Hague, the Netherlands, ³Icometrix, Leuven, Belgium, ⁴Department of Radiology, Erasmus Medical Center, Rotterdam, the Netherlands, ⁵Department of Neurosurgery, Leiden University Hospital, Leiden, the Netherlands, ⁶Division of Anaesthesia, University of Cambridge and Addenbrooke's Hospital, Cambridge, United Kingdom, ⁷Department of Public Health, Erasmus Medical Center, Rotterdam, the Netherlands, ⁸Department of Neurosurgery, Radboud University Medical Center, Nijmegen, the Netherlands

Introduction: Neurosurgical decision-making in the acute phase of traumatic brain injury management (TBI) remains challenging and mostly based on non-contrast CT findings. Icobrain software developed by Icometrix, can automatically detect and quantify volumes of intracranial hematomas, basal cisterns and amount of midline shift, thereby increasing accuracy and reducing inter- and intraobserver variability. We aimed to assess the correlation between Icobrain derived measurements with acute neurosurgical treatment and 6 months outcome in patients with TBI.

Methodology: We included admitted patients from the Center-TBI study and obtained Icobrain values from the first scan after presentation at the emergency departments. Outcomes were the decision to undertake any acute neurosurgical intervention and functional outcome, using the dichotomized Glasgow Outcome Scale Extended (GOSE) at six months. Mean values of Icobrain values were compared between groups and the association with outcomes were analyzed using logistic regression.

Results: A total of 3224 patients were included. Acute surgery was performed in 484 patients, with favorable outcome at 6 months of 43%, compared to 64% in the non-surgical group. Patients with favorable outcome in the surgery group had less midline shift, ASDH and larger basal cisterns, but a larger volume of EDH. After correcting for baseline characteristics, multivariable regression analyses showed that larger volumes of EDH, ASDH, and more compression of the basal cisterns were associated with a higher likelihood of performing surgery in the acute phase.

Conclusion: Icobrain values showed important differences in volumes of intracranial hematomas and basal cisterns in the surgical vs non-surgical group, and also in relation to outcome. The possibility to quickly and accurately quantify cisternal volumes as a surrogate of basal cisternal compression could potentially be of added value in neurosurgical decision-making in TBI patients and warrants further research.

O106 Plasma lipid profiles on admission and one month after mild traumatic brain injury: results from an emergency department cohort

Dr Sarah Hellewell ,¹ Mr Harrison Szemray, Dr Aleksandra Gozt, Dr Chidozie Anyaegbu, Dr Nathan Lawler, Dr Melissa Licari, Dr Glenn Arendts, Dr Stephen MacDonald, Dr Swithin Song, Dr Ellen MacDonald, Dr Philip Vlaskovsky, Dr Sally Burrows, Professor Michael Bynevelt, Professor Carmela Pestell, Professor Daniel Fatovich, Dr Luke Whiley, Professor Melinda Fitzgerald

¹Curtin University, Perth, Australia, ²The Perron Institute for Neurological and Translational Science, Perth, Australia

Introduction: Mild traumatic brain injury (mTBI) causes structural and biochemical damage to the axonal cytoskeleton, shedding lipids into the blood. Lipids readily cross the blood-brain barrier and are emerging novel biomarkers of neuropathology. This study examined lipidomic profiles following mTBI to determine whether lipid alterations relate to cognitive outcomes, persistent symptoms and/or white matter structure.

Methodology: 30 participants with mTBI were recruited from Royal Perth Hospital. Plasma was sampled at <48h (inception) and 28d (follow-up), alongside assessment of cognition (Repeatable Battery for the Assessment of Neuropsychological Status, Trail Making Tests A&B), and mood (Depression Anxiety Stress Scales). 37 control participants underwent assessment at one timepoint. Lipid concentrations were determined by liquid chromatography-mass spectrometry and correlated with cognition and mood. A subsample of mTBI participants underwent MRI scans 30d post-injury, with exploratory examination of relationships between lipid species and FA in white matter tracts of relevance for mTBI.

Results: Lipid concentrations were similar to control at inception. At 28d follow-up there were decreased concentrations of monoacylglycerols(14:0, 18:3, 22:4) and diacylglycerol(20:0/20:20) (p<0.001 vs. inception and control). KEGG analysis indicated that decreases had potential impacts on circadian entrainment, endocannabinoid and chemokine signaling. In contrast, there were increased concentrations of ceramide (26:0), hydroxyceramides(14:0, 16:0 & d18:0/26:0), sphingomyelin(14:0) and phosphatidylcholine(18:0/18:2) (all p<0.001 vs. inception and control). KEGG analysis demonstrated potential consequences for sphingolipid metabolism and signalling and AMP-activated protein kinase signalling. Inception sphingomyelin(14:0) correlated with depression and anxiety scores and hydroxyceramides(d18:0/26:0 and 16:0) correlated with visuospatial perception. At follow-up, hydroxyceramide(d18:0/26:0) correlated with immediate memory and ceramide(26:0) correlated with processing speed. Linear regression revealed that concentrations of hydroxyceramide(14:0) could significantly predict FA alteration in the posterior limb of the internal capsule.

Conclusions: These data suggest that plasma lipid concentrations may be dynamic biomarkers for mood alteration, cognitive deficits and white matter alteration in the brain following mTBI.

O107 Timing Of venous thromboembolism Prophylaxis for adult patients with Traumatic Brain Injury (TOP-TBl): a pragmatic, randomised trial protocol

Dr Midhun Mohan ,¹ Dr Daniel Horner,² Mr Edoardo Viaroli,¹ Ms Sara Venturini,¹ Dr Harry Mee,⁵ Prof Mark Wilson,³ Prof Danny McAuley,⁴ Prof Jonathan Coles,⁵ Professor Garry Barton,⁶ Dr Tony Veenith,⁷ Prof Simon Stanworth,⁸ Dr Simon Bond,⁹ Dr Virginia Newcombe,⁵ Prof Peter JA Hutchinson,¹ Mr Angelos Kolias¹

¹University of Cambridge, Department of Clinical Neurosciences, Division of Neurosurgery, Cambridge, United Kingdom, Cambridge, United Kingdom, ²Salford Royal NHS Foundation Trust, Manchester, United Kingdom, ³Imperial College Healthcare NHS Trust, Department of Neurosurgery, London, United Kingdom, ⁴Wellcome Wolfson Institute for Experimental Medicine, The Queen's University Belfast, Belfast, United Kingdom, ⁵University of Cambridge, Department of Clinical Neurosciences, Cambridge, United Kingdom, Cambridge, United Kingdom, ⁶Norwich Medical School, University of East Anglia, Norwich, United Kingdom, ⁷University Hospitals Birmingham NHS Foundation Trust, Birmingham, United Kingdom, ⁸Radcliffe Department of Medicine, University of Oxford, Oxford, United Kingdom, ⁹Cambridge Clinical Trials Unit, Cambridge, United Kingdom

Introduction: Every year in the UK, an estimated 1.4 million people suffer a Traumatic Brain Injury (TBI) and 200,000 people with TBI are admitted to hospital. Following a TBI, patients are at considerable risk of morbidity and mortality for a number of reasons, including the development of venous thromboembolism (VTE). In hospitalised patients, national guidelines recommend early initiation of pharmacological VTE prophylaxis (PTP) for appropriate patient populations. However, in patients with TBI the optimal timing for initiation of PTP remains unclear. This trial aims to evaluate the clinical and cost-effectiveness of early PTP administration (<72 hours) versus late administration (>120 hours or not administered at clinical discretion) for adult patients with TBI.

Methodology: This will be a multi-centre, parallel-group, pragmatic, randomised superiority trial. The inclusion criteria is as follows: adult patients (≥ 16 years of age), acute TBI, (defined as acute traumatic changes on CT brain, either in isolation or in the context of polytrauma), patients admitted to hospital within 72 hours of injury. The primary outcome will be clinically relevant VTE within 30 days from randomisation, to include any confirmed diagnosis of symptomatic DVT, pulmonary embolism or death related to VTE. All centers who manage patients with TBI are eligible to participate.

Results: We will recruit 1512 patients in total (150 in the internal pilot, 1362 in substantive study). The study will be 60 months in total with the pilot phase due to start in the third quarter of 2024 and the study close planned for the third quarter of 2028.

Conclusions: This randomised trial will provide level 1 evidence on the optimal timing of VTE prophylaxis in patients with TBI. For further information, please contact Midhun Mohan (mm2446@cam.ac.uk) and Angelos Kolias (ak721@cam.ac.uk).

O108 Elevated Biomarker Expression in Infant Victims of Abusive Head Trauma

Laura Blackwell ,^1,2 Makda Mulugeta,² Meena Verma,² Mahwish Javed,² Guangzheng Cai,³ Daniel Rawad Arja,³ Firas Kobaissy,³ Kevin Wang,³ Andrew Reisner^1,2

¹Emory University School Of Medicine, Atlanta, United States, ²Children's Healthcare of Atlanta, Atlanta, United States, ³Center for Neurotrauma, MultiOmics & Biomarkers, Department of Neurobiology, Morehouse School of Medicine, Atlanta, United States

Introduction: Abusive head trauma (AHT) is an increasingly recognized cause of infantile mortality, worldwide. A prompt diagnosis of AHT is imperative to avoid returning the child home where repetitive and more severe injuries may occur. Difficulties in identifying AHT include inconsistent histories, non-specific symptoms, and inadequate awareness of AHT by overseeing agencies. Thus, there is an urgent need to develop an objective screening tool for AHT evaluation.

Objectives: Examine levels of well-known and novel blood biomarkers in children who sustained AHT vs. accidental trauma (AT).

Methods: Children <5 years presenting to the ED of a tertiary children's hospital with TBI. Blood samples were collected at admission and 3 daily follow ups. Following admission, children were classified as having confirmed or suspected AHT, identified by a Child Protection Team physician, or AT (e.g., motor vehicle accidents). Biomarkers were assayed on the Simoa platform (GFAP, UCHL1, NFL, Tau) and ELISA (Osteopontin, neuroinflammatory marker). Binary logistic regression and repeated measures were used to associate biomarker levels and mechanism of injury.

Results: A total of 68 patients were identified from a larger biorepository dataset (AHT=39; AT=29). AHT patients were slightly younger (mean 1.53 vs. 3.04 years) with comparable GCS (median 11.5 vs. 12.5). AHT patients had higher rates of mortality (16.7% vs 9.5%) and similar rates of anoxic brain injury as identified on neuroimaging (30% vs. 28.6%) compared to AT. OPN levels at admission were higher in the AHT group versus AT (302.52 ng/ml vs. 218.13 ng/ml). Trajectory of OPN also showed significant increase over time compared to AT F(49)=72.04,p<.01. No associations were found with levels of GFAP, UCHL1, NFL or Tau.

Conclusions: Admission OPN levels are overexpressed in children with AHT compared to AT. These results support further investigation of other neuroinflammatory biomarkers that may assist in screening of infants with suspected AHT.

O109 Biopsychosocial and neuropsychological outcomes in women who have experienced intimate partner violence with brain injury and probable PTSD

Dr Georgia Symons ,¹ Dr Jen Makovic Knight,^1,2 Ms Beatrice Duarte Martins,¹ Prof Jennie Ponsford,^1,2 Prof Sandy Shultz^1,3

¹Department of Neuroscience, Monash University, Melbourne, Australia, ²Monash-Epworth Rehabilitation Research Centre, Richmond, Australia, ³Health Sciences, Vancouver Island University, Nanaimo, Canada

Introduction: Intimate partner violence (IPV) is a significant public health concern affecting one in three women worldwide. IPV is associated with several long-term health challenges such as PTSD, however brain injuries (i.e. mild traumatic brain injury mTBI) within this context have commonly been underdiagnosed and overlooked. We aimed to better understand the effect of brain injury and PTSD within this setting.

Methodology: 35 women who have experienced IPV-related mTBI (IPV-BI) were recruited. They were compared to 15 IPV controls (with history of IPV but not mTBI); 9 mTBI controls (with history of mTBI but not IPV) and 12 healthy controls (with no history of either mTBI or IPV). Participants completed medical history, psychosocial questionnaires (i.e. Rivermead Post Concussion Symptom Questionnaire; RPQ, PTSD Checklist for DSM-5; PCL-5, Depression Anxiety and Stress Scale 21; DASS-21), neuropsychological assessment (i.e., TOPF, RAVLT) and blood collection to assess serum protein biomarkers (i.e., GFAP).

Results: Women who had experienced IPV both with and without mTBI (i.e. IPV+/- mTBI) had significantly greater PCL-5 scores, RPQ symptom severity, and anxiety symptoms on the DASS-21 compared to both healthy mTBI controls. IPV-BI alone had significantly greater symptom severity compared to healthy and mTBI controls. Interestingly, 56% of IPV+/- mTBI met the criteria for probable PTSD. Preliminary cognitive assessment revealed that irrespective of mTBI, women who had experienced IPV with probable PTSD had significantly poorer verbal learning on the RAVLT. Preliminary blood biomarker analysis revealed no significant differences in GFAP; however, the IPV-BI group had a large spread and warrants further analysis to assess whether different levels of exposure may influence this.

Conclusion: Early findings support the significant long-term psychological and mental health implications of IPV. Further understanding of the exposure of IPV-BI and interaction with probably PTSD may influence neuropsychological and blood biomarker outcomes.

O110 Potential Biomechanical Vulnerability of Cortical Sulcal Depths Following Experimental Traumatic Brain Injury

John Arena ,¹ John Wolf,¹ Alexandra Ulyanova,¹ Douglas Smith,¹ H. Isaac Chen,¹ D. Kacy Cullen,¹ Victoria Johnson¹

¹University of Pennsylvania, Philadelphia, United States

Introduction: Traumatic brain injury (TBI) is associated with the development of chronic traumatic encephalopathy (CTE), defined by perivascular tau pathologies at the sulcal depths. However, mechanisms of CTE pathogenesis and the unique location at the sulci are unknown. Although it has been hypothesized that cortical sulci undergo extensive deformation during rapid head motions in TBI, it remains unknown if this is reflected by selective acute pathology.

Methodology: Using a novel porcine model of cortical impact with focal and diffuse pathologies, we performed detailed histological examination of sulci at 30min (n=2) and 72hr (n=3) post-injury versus shams (n=2). Whole-brain coronal sections at the level of impact and remote from the site of focal injury were assessed for hemorrhage (H&E), blood-brain barrier (BBB) permeability (fibrinogen extravasation), neuronal degeneration (H&E, Fluorojade-C, SNTF) axonal injury (APP, SNTF) and microglial reactivity (IBA-1).

Results: Marked accentuation of BBB permeability was frequently observed at the sulcal depths post-TBI at 72hr (p<0.001 vs. sham), however, this was not observed at 30min post-TBI (p=0.819 vs. sham). The BBB disruption at the sulcal depths was observed across the rostral-caudal extent of the brain, including regions remote from the impact/contusion. Interestingly, at 72hr, foci of neurons and astrocytes demonstrated fibrinogen immunoreactivity, often in a patchy distribution around vessels with BBB permeability, and frequently at sulcal depths. Notably, detailed mapping failed to reveal sulcal-predominance of axonal/neuronal degeneration (H&E, APP, SNTF, Fluorojade-C) or microglial reactivity (IBA-1) in any group.

Conclusions: We demonstrate that sulci are important sites of vascular injury post-TBI, with associated cellular-uptake of serum proteins. The absence of other mechanically-induced pathologies preferentially at the sulcal depths suggests BBB permeability in this region is a delayed secondary event. Accordingly, understanding how early sulcal vascular pathologies might contribute to later degenerative processes will be important to explore.

Support: R01-NS123034, W81XWH2010838, W81XWH-20–1-0901, T32-NS043126

O111 Early MRI in moderate and severe traumatic brain injury: Quantitative lesion measures and associations to cause of injury

Miss Tiril Svaasand Eliassen ,¹ Anne-Mari Holte Flusund,^2,3 Joakim Stray Andreassen,^1,4 Anne Katharina Köster,³ Oddrun Sandrød,^5,6 Toril Skandsen,^2,7 Anne Vik,^2,4 Kent Gøran Moen^1,4,8,9

¹Department of Circulation and Medical Imaging, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ²Department of Neuromedicine and Movement Science, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ³Department of Radiology, Møre and Romsdal Hospital Trust, Molde Hospital, Molde, Norway, ⁴Department of Neurosurgery, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, ⁵Department of Anesthesiology and Intensive Care, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, ⁶Department of Public Health and Nursing, Faculty of Medicine an Health Science, Norwegian University of Science and Technology (NTNU), Trondheim, Norway, ⁷Clinic of Rehabilitation, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway, ⁸Department of Radiology, Vestre Viken Hospital Trust, Drammen Hospital, Drammen, Norway, ⁹Department of Radiology and Nuclear Medicine, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway

Introduction: We investigated correlation between cause of injury and lesion burden of traumatic axonal injury (TAI) and brain contusion on clinical MRI in patients with moderate and severe traumatic brain injury (TBI).

Methodology: We included 335 patients (8–70 years) admitted to St. Olav’s University Hospital (2004–2021) with known cause of injury, moderate (n=173) or severe TBI (n=162) and MRI within 6 weeks after injury (median 9, IQR 4–18 days). Cause of injury was registered as road traffic accident (RTA), fall or alpine skiing. Manual volume segmentations were performed on fluid-attenuated inversion recovery (FLAIR) and diffusion-weighted imaging (DWI) for TAI and on FLAIR for brain contusion. Number of microhemorrhages was counted on T2*-weighted gradient echo (T2*GRE) or susceptibility weighted imaging (SWI).

Results: TAI was found in 86% of patients injured in RTAs and 69% of patients injured in falls (p<0.001), while 64% of patients injured in RTAs and 89% of patients injured in falls had brain contusion (p<0.001). Largest TAI volumes and most microhemorrhages were found in RTAs (median volume FLAIR: 0.70 cm³, median volume DWI: 0.16 cm³, median number T2*GRE/SWI: 21) and alpine skiing accidents (median volume FLAIR: 0.58 cm³, median volume DWI: 0.54 cm³, median number T2*GRE/SWI: 23). In subgroups of RTAs, pedestrians had largest TAI volumes and most microhemorrhages (median volume FLAIR: 2.57 cm³, median volume DWI: 0.98 cm³, median number T2*GRE/SWI: 41). Fall injuries gave largest brain contusion volumes (median volume FLAIR: 20.55 cm³). The volume did not vary significantly by fall height (>own height: 21.72 cm³, stairs: 18.23 cm³, ≤own height: 27.03 cm³, p=0.60).

Conclusion: In this prospective TBI cohort, we found largest TAI volumes and most microhemorrhages in RTAs and alpine skiing accidents. Pedestrians had largest lesion burden of TAI. Largest volume of brain contusion was found in fall injuries regardless of the fall height.

O112 Alterations in Cerebral Metabolism Associated with Non-Fatal Strangulation in Females with Intimate Partner Violence

Alexander Lin,^1,2,3 Skyler McComas ,^2,3 Divya Jain,⁴ Katherine Breedlove,^2,3 Emily Carter,⁵ Andrew Cwiek,⁵ Katherine Dorman,⁴ Inga Koerte,^1,7,8,9 Amy Marshall,⁵ Adriana Méndez-Fernández,⁵ Emma Read,^9,10 Elizabeth Rebuck,⁵ Philine Rojczyk,^7,8 Carmen Velez,⁹ David Tate,^9,10 Frank Hillary,^5,6 Elisabeth Wilde,^9,10 Carrie Esopenko⁴

¹Psychiatry Neuroimaging Laboratory, Department of Psychiatry, Brigham and Women’s Hospital, Harvard Medical School, Somerville, USA, ²Department of Radiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA, ³Center for Clinical Spectroscopy, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA, ⁴Department of Rehabilitation and Human Performance, Icahn School of Medicine at Mount Sinai, New York City, NY, USA, ⁵Department of Psychology, Pennsylvania State University, University Park, Pennsylvania, USA, ⁶Social Life and Engineering Sciences Imaging Center, University Park, Pennsylvania, USA, ⁷cBRAIN, Department of Child and Adolescent Psychiatry, Psychosomatics, and Psychotherapy, Ludwig-Maximilians-University, Munich, Germany, ⁸Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA, ⁹Graduate School of Systemic Neurosciences, Ludwig-Maximilians-University, Munich, Germany, ¹⁰Traumatic Brain Injury and Concussion Center, University of Utah School of Medicine, Salt Lake City, UT, USA, ¹¹George E. Wahlen VA Salt Lake City Healthcare System, Salt Lake City, UT, USA

Introduction: Intimate partner violence (IPV) a significant public health concern, affects approximately 1 in 3 females globally. There is increasing awareness that a high proportion of females with exposure to IPV also experience some form of IPV-related head trauma (IPV-HT), including non-fatal strangulation (NFS). IPV-HT has been associated with impaired cognitive and psychological functioning, and alterations in brain structure and function. NFS, specifically, can cause hypoxic-ischemic brain injury (HIBI) and may further exacerbate the effects of IPV-HT. However, the impact of NFS on cerebral metabolism has not been examined. As part of an on-going study, females with exposure to severe physical IPV completed a comprehensive assessment that included magnetic resonance spectroscopy (MRS).

Methods: Proton MRS was acquired from 40 females (mean age = 40.2, 71.1% White) using point-resolved spectroscopy (TE=30ms) in the anterior (ACG) and posterior cingulate gyrus (PCG) across three sites using 3T MRI. MRS data was analyzed using Osprey processing software. NFS history was obtained from the Brain Injury Screening Questionnaire with the IPV module. Quality assurance was conducted on MRS data quality and voxel placement resulting in one spectrum removed per location. One-way ANOVA was performed using NFS status (with/without NFS exposure) to predict changes in neurometabolites (N-acetylaspartate, choline, creatine, glutamate/glutamine, myoinositol, and lactate), with age as a covariate and normalizing data across sites.

Results: Myoinositol levels were significantly decreased in the ACG in females with NFS, F(1,41)=4.392, p=0.0192. There were no significant changes in the PCG or for other metabolites including lactate in the ACG. Myoinositol is a marker of glial health and is shown to be reduced in post-cardiac arrest patients who suffered hypoxia.

Conclusion: This is the first study to examine neurometabolite levels in females with IPV-HIT and examining the overlap with NFS. Reduced myoinositol in NFS may reflect metabolic changes resulting from HIBI.

O113 Diagnostic and prognostic utility of acute phase proteins and candidate inflammatory cytokines in traumatic brain injury: a TRACK-TBI Study

Ava Puccio ,³ John Yue,¹ Sonia Jain,² Xiaoying Sun,² Catherine Demos,⁴ Nikhil Padmanabhan,⁴ Taron Gorham,⁴ George Sigal,⁴ Jacob Wohlstadter,⁴ Thomas van Essen,⁵ Romit Samanta,⁶ Patrick Belton,¹ Ester Yuh,¹ Lindsay Nelson,⁷ Mahmoud Elguindy,¹ Joye Tracey,¹ Shawn Eagle,³ Frederick Korley,⁸ Andrea Schneider,⁹ Pratik Mukherjee,¹ Raquel Gardner,¹ Amy Markowitz,¹ Firas Kobeissy,¹⁰ David Okonkwo,³ Ramon Diaz-Arrastia,⁹ Geoffrey Manley,¹ Kevin Wang,¹⁰ TRACK-TBIInvestigators¹

¹University of California, San Francisco, San Francisco, United States, ²University of California, San Diego, San Diego, United States, ³University of Pittsburgh, Pittsburgh, United States, ⁴MesoScale Discovery, Gaithersburgh, United States, ⁵Leiden University Medical Center, Leiden, The Netherlands, ⁶Addenbrookes Hospital, Cambridge, United Kingdom, ⁷Medical College of Wisconsin, Milwaukee, United States, ⁸University of Michigan, Ann Arbor, United States, ⁹University of Pennsylvania, Philadelphia, United States, ¹⁰Morehouse School of Medicine, Atlanta, United States

Introduction: Systemic and neuroinflammatory responses mediate secondary injuries after traumatic brain injury (TBI), often leading to worse outcomes. Diagnostic and prognostic properties of 26 inflammatory biomarkers were examined.

Methodology: The 18-center TRACK-TBI Study (2014–2018) enrolled patients receiving head computed tomography (CT) within 24-hours of TBI. This analysis comprised 394 TBI patients, 100 orthopedic trauma controls (OCs), and 67 healthy controls (HCs) with plasma biomarkers (MesoScale Diagnostics) and outcomes. Comparisons included CT-positive/negative TBI, Glasgow Coma Scale (GCS)=3–12/13–15, and 6-month unfavorable/favorable outcome (GOSE=1–4/5–8). Medians/quartiles were evaluated; differences between medians (fold-change) and discrimination (area under the curve (AUC)) were reported. Significance was reported at a threshold of p<0.002 (0.05÷26 biomarkers). Comparisons were reported at p<0.0001(*) unless denoted.

Results: Eleven biomarkers differentiated TBI/HC, TBI/OC, severity, and outcome. Fold-changes and AUCs were consistent across TBI severity comparisons (CT+/CT-, GCS=3–12/13–15, respectively): interleukin-10 (IL-10; 6.4-fold*/7.6-fold*; AUC=0.81/0.89), IL-6 (4.6*/3.7*; AUC=0.82/0.87), IL-2 (3.3*/4.1*; AUC=0.82/0.85), tumor necrosis factor-alpha (TNFa; 2.3*/3.0*; AUC=0.81/0.85), IL-15 (1.4*/1.5*; AUC=0.77/0.85), c-reactive protein (CRP; 8.8*/10.2*; AUC=0.79/0.81), serum amyloid A (SAA; 19.8*/18.6*; AUC=0.78/0.0.80), IL-1b (3.4*/4.3*; AUC=0.76/0.83), IL-4 (3.4*/3.9*; AUC=0.74/0.79), IL-17a (1.7*/2.0*; AUC=0.69/0.74), IL-12p70 (1.4*/1.6*; AUC=0.65/0.70). Notably, biomarker metrics for TBI severity were generally consistent for unfavorable outcome: IL-10 (5.6-fold*/AUC=0.83), IL-6 (1.5*/AUC=0.75), IL-2 (3.9*/AUC=0.81), TNFa (2.8*/AUC=0.77), IL-15 (1.6*/AUC=0.79), CRP (4.9*/AUC=0.71), SAA (4.1*/AUC=0.67), IL-1b (2.7*/AUC=0.74), IL-4 (2.5*/AUC=0.73), IL-17a (2.1*/AUC=0.73), IL-12p70 (1.7*/AUC=0.70), Bivariate correlations (Spearman’s ρ>0.7) emerged amongst IL-1b, IL-2, IL-4, TNFa (with several other markers), and between CRP/SAA.

Conclusions: In a large prospective cohort, we identified 11 priority blood-based inflammatory proteins with diagnostic and prognostic relevance to TBI. These biomarkers distinguished TBI severity and 6-month unfavorable outcome, with consistent magnitudes of association across endpoints.

O114 Neuropsychology Outcomes in Traumatic Spinal Injury: A Single Centre Study from Lusaka, Zambia

Swati Jain ¹ , Vivien Penda,² David Clark,¹ Jackson Zulu,² Brian Songkwe,² Rikin Trivedi¹

Introduction: Spinal cord injury (SCI) is physically and psychologically debilitating. In today’s era, 90% of the patients with SCI survive past the first year of the injury, and 50% beyond the 40 years of injury. SCI usually affects younger working population, especially in low and middle-income countries (LMICs). There are reports of significant psychological trauma in patients suffering from SCI. In this study, we aimed to assess the neuropsychological outcomes in patients with SCI.

Materials and Methods: Prospective data was collected from a single centre at University Teaching Hospital, Lusaka, Zambia for patients who presented with traumatic SCI. A data dictionary was developed for this purpose. All neuropsychology assessments were performed by a single neuropsychologist at discharge and on follow up. The measures of assessment included FIMFAM Score, SF-12, PHQ-9, Rosenberg Self Esteem Score, Hopkins Verbal Learning Test and Stroop test.

Results and Discussion: 303 patients with mean age of 33.6 ± 14 years were assessed in this study. 70 patients were employed at time of injury. ASIA grade on presentation was as follows: A (21.1%), B(5.6%), C(9.9%), D(7.6%), E(12.5%). 27.1% patients had a fall from height, 17.2% had a road traffic collision as a passenger in the car, and 11.6% had road traffic collision as a pedestrian. 51 patients had completed their FIMFAM assessment on discharge, 265 patients were followed up after discharge of which 87 had completed the assessment. Whilst there were improvements in activities of daily living, employability, adjustment to limitations, social integration remained low.

Conclusion: Re-integration of patients after TSI requires a holistic approach in intervention and assessment. Without a detailed neuropsychological assessment despite improvements in neurological deficits, it would be difficult to gauge the impact of TSI.

O115 Disruption of coupling connectivity and its predictive role in executive function post pediatric traumatic brain injury

Dr. Reut Raizman ,^1,2 Dr Tamar Silberg,^3,4 Mrs. Hadar Shapsa,¹ MA Yael Golan,^1,4 Mrs Moran Shechtman,^1,4 Msc Neta Erez,³ Dr. Jana Landa,^2,3 Dr Galia Tsarfaty,¹ Professor Chaim G. Pick,^2,5,6 Dr Abigail Livny^1,2,6

¹Division of Diagnostic Imaging, Sheba Medical Center, Tel-Hashomer, Ramat-Gan, Israel, ²Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel, ³Edmond and Lily Safra Children's Hospital, Sheba Medical Center, Ramat Gan, Israel, ⁴Department of Psychology, Bar Ilan University, Ramat Gan, Israel, ⁵Sagol School of Neuroscience, Tel Aviv University, Tel Aviv, Israel, ⁶The Dr. Miriam and Sheldon G. Adelson Center for the Biology of Addictive Diseases, Tel-Aviv University, Tel Aviv, Israel

Introduction: Pediatric traumatic brain injury (pTBI) is known as a disorder of brain connectivity. TBI induces substantial long lasting effects on the developing brain, resulting in alterations in both structural connectivity (SC) and functional connectivity (FC) which in turn are associated with cognitive deficits. While independent examinations of each connectivity modality have provided insights into the distinct impact of p-TBI on brain network organization and cognitive function, the precise pattern of structural-functional connectivity relationship post TBI and its impact on cognitive outcome, specifically executive functions, remain poorly understood.

Methods: pTBI patients aged 9–18 years, from all injury severities and healthy control subjects were examined in the acute and post-acute phases. Participants underwent an MRI protocol, including structural and functional connectivity scans, and a cognitive evaluation, focusing on executive functions. Structural and functional connectomes were generated, and graph theory was applied to each MRI modality to examine the global and local network architecture. The connectivity coupling was produced by Pearson correlation between the graph theory measures of SC and FC. Group differences were examined using analysis of covariance. The acute network architecture was used to predict the post-acute cognitive outcome with multiple regression models.

Results: Our findings demonstrated disrupted network topology in both SC and FC in pTBI compared to healthy controls, in the acute phase. Moreover, pTBI patients showed only lower working memory function in the acute phase, and lower function in all executive functions tasks in the post-acute phase. In addition, pTBI patients displayed lower strength, efficiency, and cluster coefficient coupling connectivity. Finally, the efficiency coupling connectivity of the acute phase predicted inhibition performance in the post-acute phase.

Conclusion: Our study provides insights into the complex relationship between structural and functional brain connectivity, and the potential of coupling connectivity as a valuable tool for predicting cognitive outcome post pTBI.

O116 Effects of Biological Sex and Brain Injury Severity on Sleep Architecture

Grant Mannino ,¹ Dr Tabitha Green, Dr Sean Murphy, Dr Mark Opp, Dr Rachel Rowe

¹University of Colorado Boulder, Boulder, United States

Introduction: Traumatic brain injury (TBI) disturbs sleep which delays recoveries and leads to long-term neurological morbidities in TBI survivors. Clinical data indicate that TBI alters acute sleep following mild and moderate brain injury. However, little preclinical research has employed injury severity as a parameter to investigate post-traumatic sleep after TBI. Efforts have also been made to include both male and female animals (e.g., rodents) in preclinical sleep research. Nevertheless, when sleep is used as a physiological outcome in research, few studies have disaggregated sleep data based on sex. In this study, we hypothesized that post-injury sleep disturbances would occur independent of injury severity, in a sex-dependent manner.

Methods: We exposed adult male and female mice to sham (n = 30), mild (n = 32), or moderate (n = 32) midline fluid percussion injury (mFPI). Physiological parameters were recorded using non-invasive piezoelectric cages to determine non-rapid eye movement (NREM) sleep, rapid eye movement (REM) sleep, and wakefulness (WAKE). To investigate sleep differences between sexes and injury severities we fit hierarchical generalized linear mixed models with nonlinear time effects.

Results: We found substantial sex differences in the sleep of C57BL/6J mice. Female mice slept less overall, specifically during the dark period, with less NREM sleep compared to males. TBI- regardless of injury severity- increased percent NREM sleep and total minutes of NREM sleep during the post-injury period for both male and female mice. Mice of either sex subjected to mild or moderate TBI also exhibited shorter WAKE bouts, predominately during the dark (active) period.

Conclusions: TBI substantially altered sleep after both mild and moderate TBI, in male and female mice. Thus, disturbed sleep architecture should be further explored as a bioindicator of brain injury and future studies should investigate the clinical utility of using disturbed sleep as a diagnostic tool for concussion.

O117 Paradoxical associations of biomarkers with patient reported outcomes in CT negative mild traumatic brain injury: a CENTER-TBI analysis

Dr Daniel Whitehouse ,¹ Professor Lindsay Wilson,² Dr Endre Czeiter,^3,4 Professor Andras Buki,⁵ Professor Kevin Wang,^6,7 Professor Nicole von Steinbüchel,⁸ Dr Marina Zeldovich,⁸ Professor Ewout Steyerberg,⁹ Professor Andrew Maas,^10,11 Professor David Menon,¹ Dr Virginia Newcombe¹

¹Perioperative, Acute, Critical Care and Emergency Medicine (PACE), Department of Medicine, University of Cambridge, Box 93, Addenbrooke’s Hospital, Hills Road, CAMBRIDGE, United Kingdom, ²Division of Psychology, University of Stirling, Stirling, UK, ³Department of Neurosurgery, Medical School, University of Pecs, Ret u. 2, H-7623, Pecs, Hungary, ⁴Neurotrauma Research Group, Szentagothai Research Centre, University of Pecs, Ifjusag utja 20, H-7624, Pecs, Hungary, ⁵Department of Neurosurgery Faculty of Medicine and Health Örebro University SE 70182, Örebro, Sweden, ⁶Department of Neurobiology, Center for Neurotrauma, Multiomics & Biomarkers (CNMB) Neuroscience institute, Morehouse School of Medicine (MSM), Atlanta, USA, ⁷Program for Neurotrauma, Neuroproteomics and Biomarker Research, Departments of Emergency Medicine, Psychiatry and Neuroscience, University of Florida, McKnight Brain Institute, Gainesville, US, ⁸Institute of Psychology, University of Innsbruck, Austria, ⁹Department of Biomedical Data Sciences, Leiden University Medical Center, Leiden, Netherlands, ¹⁰Department of Neurosurgery, Antwerp University Hospital, Edegem, Belgium, ¹¹Department of Translational Neuroscience, Faculty of Medicine and Health Science, University of Antwerp, Antwerp, Belgium

Introduction: There is seemingly contradictory evidence concerning relationships between day of injury biomarkers and outcome following mild traumatic brain injury (mTBI). To address this issue, we examined the association between a panel of biomarkers and multidimensional TBI outcomes.

Methodology: mTBI participants were separated into two groups by the presence or absence of traumatic intracranial abnormality. The correlation of outcomes (GOSE, QOLIBRI-OS, SF12v2 MCS and PCS, GAD-7, PHQ-9, PCL-5) was assessed in each group. Multivariable binary logistic regression analysis was used to assess the relation between biomarkers (GFAP, NFL, NSE, S100B, t-tau, UCH-L1) and outcomes with adjustment for age, sex, time to sampling, and degree of extracranial injury.

Results: Higher median log S100B (OR 1.78, CI 1.43–2.23) and log UCH-L1 (OR 1.16, CI 1.01–1.33) related to poorer outcome (GOSE <8) in CT negative participants. Similarly, all biomarkers aside from NSE, related to poorer outcome in CT positive participants. Less association was seen between the biomarkers and HRQoL or mental-health outcomes. In CT negative participants only, an inverse relationship was observed between log GFAP and impaired QOLIBRI-OS (OR 0.76, CI 0.66–0.88), SF12v2 MCS (OR 0.71, CI 0.61–0.82), SF12v2 PCS (OR 0.79, CI 0.68–0.91), GAD-7 (OR 0.80, CI 0.68–0.95), PHQ-9 (OR 0.80, CI 0.68–0.93), and PCL-5 (OR 0.80, CI 0.66–0.97).

Conclusions: Participants with higher median concentrations of biomarkers, in both CT positive and CT negative mTBI, had greater odds of impaired functional recovery. However, in CT negative participants, a higher day of injury GFAP concentration was associated with better HRQoL and less impaired mental health. These apparently paradoxical findings may indicate potentially discordant biological processes driving particular outcomes. This issue is important as evidence concerning the origin of symptoms reported after TBI could influence the management of patients and help to inform the design of clinical trials.

D001 Cerebral Autoregulation Alters the Relationship of Quantitative Electroencephalographic Activity with Intracranial Pressure after Pediatric Traumatic Brain Injury

Dr. Brian Appavu ,^1,2 Dr. Katie Rodriguez,¹ Dr. M’Hamed Temkit¹

¹Phoenix Children's, Phoenix, United States, ²University of Arizona College of Medicine - Phoenix, Phoenix, United States

Introduction: Intracranial hypertension (ICH) after traumatic brain injury (TBI) has been linked with poor outcomes. ICH etiologies vary, either coming from hyperemic cerebral blood flow (CBF) or extravascular sources that impede CBF. The alpha-delta power ratio (ADR) is a quantitative electroencephalographic biomarker that has been linked to CBF. We aimed to investigate the relationship of ADR to intracranial pressure (ICP) and arterial blood pressure (ABP) during plateau waves of ICH in pediatric TBI patients, and the impact of cerebral autoregulation (CA) on these relationship.

Methods: This is a retrospective analysis of pediatric TBI patients admitted to Phoenix Children’s undergoing ICP and electroencephalography monitoring. We collected ADR data during plateau waves of ICH, where ICP exceeded 20 mmHg. We fit linear mixed effects models with a random intercept for each subject and one fixed effect, and an auto-regressive order of 1 to model subject variation and correlation for within-subject observations. CA was assessed using the pressure reactivity index (PRx), and patients were differentiated from having inefficient CA (PRx (≥0.3) or efficient CA (<0.3).

Results: Thirty-three patients were identified with ICH plateau waves. Differences were identified in the relationship of ADR to ABP (p<0.001) and ICP (p=0.0314) based on CA. There was a positive relationship of ADR and ICP In patients with efficient CA (n=27, estimate 0.73, 95% confidence interval [CI] [-0.51–0.94], p<0.0001), and no relationship in patients with inefficient CA (n=6, estimate 0.15, 95% CI [-0.33, 0.63], p=0.5324). There was a positive relationship of ADR and ABP In patients with efficient CA (estimate 1.94, 95% CI [1.26, 2.62], p<0.0001), and a negative relationship in patients with inefficient CA (estimate -2.13, 95% CI [-3.69, -0.57], p=0.0074).

Conclusion: The relationship of ADR to ICP and ABP differ during plateau waves of ICH in pediatric TBI patients, and may help differentiate etiologies of ICH.

D002 STARSHIP Part 3: Physiological signal entropy for outcome prediction in pediatric traumatic brain injury: looking beyond the obvious

Stefan Yu Bögli ,¹ Claudia Ann Smith,¹ Peter Hutchinson,¹ Shruti Agrawal,^1,2 Peter Smielewski¹

¹Department of Clinical Neurosciences, University of Cambridge, Cambridge, United Kingdom, ²Department of Paediatrics, Cambridge University, Cambridge, United Kingdom, ³Paediatric Intensive Care, Cambridge University Hospitals, Cambridge, United Kingdom

Introduction: Management and outcome prediction in Paediatric traumatic brain injury (TBI) relies heavily on the evaluation of instantaneous and easily interpretable monitoring values. Entropy quantifies the level of disorder within a system reflecting overall activity of sensitive closed-loop feedback homeostatic mechanisms. In addition, as complex physiological systems operate across different time scales, coarse graining (i.e. averaging increasing number of data points) can be applied to allow for frequency specific examination of entropy. The resulting combined metrics is termed multiscale entropy (MSE). Current understanding of MSE suggests that low entropy reflects increased rigidity of the various homeostatic control systems, reflecting underperformance of mechanisms such as cerebral autoregulation. In adult TBI patients low MSE predicts unfavorable outcome. MSE has yet to be explored in a Paediatric TBI population.

Methodology: 135 children with severe TBI were prospectively enrolled in 10 centers as part of the STARSHIP study. In the enrolled patients, high frequency physiological data was collected using the ICM+ software and curated to remove artefacts. Resulting 10s time trends of arterial blood pressure, heart rate, intracranial pressure and amplitude were extracted. MSE was calculated for each of these metrics for consecutive, overlapping 3h data samples resulting in patient-specific time-trends. Overall averages as well as dose dependent metrics were calculated and correlated to 12-month outcome. In addition to univariate analysis, multivariate methods were used to assess validity of the metrics when corrected for known predictors of unfavorable outcome (age, clinical and imaging based severity).

Results: The results of this study will be presented in detail alongside the other STARHIP abstracts forming the first results announcement package of this important multicentre study of high resolution monitoring in Paediatric population.

Conclusions: We present the first investigation of MSE of physiological monitoring biosignals assessed in patients enrolled to the largest multicenter cohort of pediatric TBI.

D003 The evolution of extensive cortical hypoxic-ischemic injury from TBI in the intensive care unit: the interplay of seizure, hemorrhage, hypercapnia, and cardiac dysfunction

Tawny Stinson,¹ Inori Kawauchiya,¹ Frances Rodriguez Lara,¹ Dr. Alexandra Hochstetler,^1,2 Benjamin Baskin,¹ Aarush Patnala,¹ Krystal Qiao,¹ Mariana Gonzalez Rodriguez,¹ Dr. Declan McGuone,³ Ann-Christine Duhaime,^1,4 Dr. Kevin Staley,^1,4 Dr. Beth Costine-Bartell ^1,4

¹Massachusetts General Hospital, Boston, United States, ²Boston Children's Hospital, Boston, United States, ³Yale School of Medicine, New Haven, United States, ⁴Harvard Medical School, Boston, United States

Severe traumatic brain injury (TBI) can encompass a single pathoanatomic lesion, but more commonly, it demonstrates multiple pathoanatomic lesions including subdural/subarachnoid hemorrhage and contusion, and can present with seizures, apnea, hypoventilation, and hypotension, which might exacerbate the pathophysiology. In infants and toddlers, this combination of multi-pathoanatomic contributors can eventually result in hypoxic-ischemic tissue damage evolving over hours to days in either a patchy or holo-hemispheric pattern and is termed “hemispheric hypodensity”. The distinguishing feature from pure anoxic injuries is the sparing of deep gray brain regions, patchy damage, or sparing of an entire hemisphere indicating that hypoxic-ischemic injury can occur from mechanisms other than pure asphyxia. Worse brain injuries result in more seizures, but it is not known if seizures drive the evolving pathophysiology or is a mere consequence. Here we describe the threshold of subarachnoid hemorrhage required to direct damage to the cortex and the threshold of seizure length required to cause extensive hypoxic ischemic injury 24h post-injury in piglets of similar developmental stages as human infants and toddlers. Young children experiencing TBI with seizures are typically treated with anti-seizure medications (ASMs). The effect of GABA agonists after severe TBI in the immature brain when GABA is still depolarizing are unknown. We hypothesized that ASMs administered 1-hour post-induced seizure would reduce the amount of tissue damage. To date, we have not observed a reduction of tissue damage, but rather a potential increase in cardiac dysfunction. Ten minutes of hypercapnia after seizure reduced the length of seizure by 80% and relocated the seizure from the initiated hemisphere to the contralateral hemisphere or became generalized. Work is ongoing to determine if the negative effects of ASMs outweigh the benefits of stopping the seizures and whether mild and/or brief hypercapnia might demonstrate clinical utility for its anti-seizure effect in severe TBI.

D004 TBI results in sex-dependent pubertal disruption, hypopituitarism, and sleep disturbances in juvenile rats

Dr Tabitha Green ,^1,2 Grant Mannino,² Nicole Couillard,² Dana Ritterbusch,² Dr Sean Murphy,³ Dr Mark Opp,² Dr Rachel Rowe²

¹The University of Glasgow, Glasgow, United Kingdom, ²The University of Colorado Boulder, Boulder, United States, ³Cumberland Biological and Ecological Researchers, Longmont, United States

Limited research exists on TBI-induced pubertal disruption and hypopituitarism. Male and female peripubertal rats (post-natal day 35) received diffuse TBI or sham surgery (n=56). We investigated growth, puberty, and function of the hypothalamic-gonadotropin-axis. We also investigated post-traumatic sleep because hypopituitarism results in sleep-wake disturbances that further complicate puberty. Brains were collected at 7 days post-injury (DPI) to immunohistologically assess hypothalamic neuronal populations (GnRH, orexin, kisspeptin) and gliosis. We analyzed data using generalized linear mixed models; statistical and biological significance were determined based on p < 0.05 and standardized effect sizes (d) ≥0.5, respectively. Righting reflex times were similar between male and female rats subjected to TBI, but females had longer apnea and higher mortality probability. In both males and females, TBI caused biologically significant reductions in gonad weights and higher levels of circulating kisspeptin (plays a vital role in the onset of puberty) compared to respective shams. In females, TBI increased follicle-stimulating hormone, whereas in males, TBI impacted growth, with lower growth hormone and IGF-1 levels and lower terminal body weights compared to shams. Independent of TBI, significant sex differences existed in follicle-stimulating hormone, IGF-1, testosterone, and IL6. TBI increased sleep in both male and females at 1DPI but this increase was sustained to 5DPI in males only. A novel colocalization of kisspeptin and orexin neurons was identified in the hypothalamus and neuronal histology and analysis of gliosis is ongoing. Together, these results provide an improved understanding of hypopituitarism and pubertal disruptions following TBI. Funding: R21NS120022

D005 Comparison of Intracranial Pressure Dose Estimation from Low-Resolution Versus High-Resolution Intracranial Pressure Recording in Paediatric Traumatic Brain Injury

Dr Nalaayeni Kanesan ,¹ Miss Carly Tooke,¹ Miss Tracey Rowberry,¹ Dr Hari Krishnan Kanthimathinathan¹

¹Birmingham Children's Hospital, Birmingham, West Midlands

Introduction: Management of intracranial pressure (ICP) and maintenance of cerebral perfusion pressure (CPP) are management aspects in severe paediatric TBI (pTBI). Historically, studies analysing ICP in pTBI have typically used 1 data point per hour, usually the end of the hour ICP reading. It is unclear whether such low-resolution ICP data recording can reflect severity of intracranial hypertension events.

Methodology: This sub-study is part of prospective observational study of multimodal neuromonitoring in severe pTBI (November 2016 - October 2019) in PICU, Birmingham Children’s Hospital. High-resolution data was collected using ICM+. We analysed the overall hourly ICP dose (all ICP values, including <20mmHg), overall ICP dose per patient, as well as overall intracranial hypertension dose per patient (events with ICP > 20 mmHg only). We considered a difference of ±200mmHg of ICP dose per hour or ± 5% ICP or intracranial hypertension dose per patient as acceptable.

Results: 2179 hours of recording in 26 patients were analysed. Whole population mean ICP doses were similar with either method. However, 34% (731 of 2179 hours) of hourly estimated ICP burden values were outside of acceptable range. Bland-Altman plot showed that 95% of low-resolution hourly ICP dose values were within ± 600mmHg. At a per-patient level, only 2 patients were outside of a ± 5% level (range of -17% and +11%) when whole ICP dose was analysed. However, only 4/21 patients with intracranial hypertension were within the acceptable ±5% level, with 14/21 having a difference more than 20% intracranial hypertension dose (range: -204% to +100%)

Conclusion: Low resolution hourly ICP method was observed to provide an acceptable overall summary measure of ICP per patient when compared with minute-to-minute values, but had unacceptable accuracy for more in-depth hourly ICP-dose or intracranial hypertension dose based data analysis.

D007 Traumatic Brain Injury Biomarker Utility Varies by Age: A prospective analysis of 425 children

Professor Andrew Reisner ,^1,2 Laura S Blackwell,^1,2 Jacob R Lepard,^1,2 Ali Alawieh,^1,2 Meena S Verma,^1,2 Anna Trofimova,^1,2 Joshua J Chern,^1,2 Makda G Mulugeta,^1,2 Mahwish Javed,^1,2 Alvin Onyewuenyi,¹ Amanda J Pierzchala,^1,3 Scott Batchelor,^1,3 Guangzheng Cai,⁴ Daniel Rawad Arja,⁴ Firas Kobeissy,⁴ Kevin K. Wang⁴

¹Children's Healthcare of Atlanta, Atlanta, USA, ²Emory University School of Medicine, Atlanta, USA, ³Pediatric Emergency Medicine Associates, Atlanta, USA, ⁴Center for Neurotrauma, MultiOmics & Biomarkers, Department of Neurobiology, Morehouse School of Medicine, Atlanta, USA

Introduction: Contemporary management of traumatic brain injury (TBI) relies primarily on presenting clinico-radiographic findings. Emerging evidence from adult TBI populations suggest that biomarkers have potential to refine TBI care. Similar data in children is lacking.

Objective: Examine the relationship between TBI blood biomarkers and patient age.

Patients and Methods: Prospective cohort study of 425 children presenting to a Level I trauma center following a TBI with Glasgow Coma Scale (GCS) scores of 3 to 15. The primary outcome measure was intracranial lesion on CT scan. Blood samples were obtained on all patients within six hours of injury and measured with Quanterix™ Simoa platform for GFAP and UCH-L1, and ELISA assay for Osteopontin (OPN). Area under the receiver operating characteristic curve (AUC) with 95% confidence intervals was used to predict intracranial lesion.

Results: Age-related differences were found such that GFAP was better at detecting adolescents (12+ years) with traumatic intracranial lesions on CT, compared to younger cohorts (<5 years, 5–12 years); AUC for <5 years = .583 (95%CI = .480 to .686), AUC for 5–12 years = .770 (95%CI = .696 to .844, AUC for 12+ years = .825 (95%CI = .763 to .886). UCH-L1 and OPN also showed age related differences for CT positivity: UCH-L1: AUC for <5 years = .585 (95%CI = .481 to .688), AUC for 5–12 years = .702 (95%CI = .616 to .788, AUC for 12+ years = .580 (95%CI = .488 to .671); OPN: AUC for <5 years = .681 (95%CI = .589 to .774), AUC for 5–12 years = .532 (95%CI = .435 to .629, AUC for 12+ years = .495 (95%CI = .408 to .583).

Conclusions: TBI biomarkers known to be predictive of intracranial injury in adults are less accurate in younger children. Further studies to define age-related normal values are warranted.

D008 Comparison of Blood GFAP and UCH-L1 biomarker Measurements in Pediatric TBI from Two Assays: The Abbott i-STAT™ and (Quanterix™) Simoa platform

Professor Andrew Reisner ,^1,2 Laura S. Blackwell,^1,2 Jacob R Lepard,^1,2 Ali Alawieh,^1,2 Meena S Verma,^1,2 Joshua J Chern,^1,2 David W Wrubel,^1,2 Barunashish Brahma,^1,2 Michael S Sawvell,^1,2 Amanda J Pierzchala,^1,3 Scott Batchelor,^1,3 Mahwish Javed,^1,2 Makda G. Mulugeta,^1,2 Alvin Onyewuenyi,¹ Guangzheng Cai,⁴ Daniel Rawad Arja,⁴ Firas Kobeissy,⁴ Kevin K. Wang⁴

¹Children’s Healthcare of Atlanta, Atlanta, United States of America, ²Emory University School of Medicine, Atlanta, United States of America, ³Pediatric Emergency Medicine Associates, Atlanta, United States of America, ⁴Center for Neurotrauma, MultiOmics & Biomarkers, Department of Neurobiology, Morehouse School of Medicine, Atlanta, United States

Introduction: Incorporation of biomarker levels to traditional clinico-radiographic classifications may enhance traumatic brain injury (TBI) management. Point of care (POC) biomarker testing at initial triage is desirable to guide immediate care. This is especially true in children, given the deleterious risks of ionizing radiation associated with head CT.

Objective: To compare GFAP and UCH-L1 values measured with a POC device compared with a core laboratory platform in pediatric TBI patients.

Patients and Methods: GFAP and UCH-L1 were measured in 40 children who sustained TBI (initial GCS 13–15). Frozen plasma was analyzed on both (Quanterix™) Simoa platform and Abbott i-STAT™ handheld device (TBI plasma test). We correlated the values found on each platform and examined the predictive value of both biomarkers on intracranial injury. Descriptive statistics (median, interquartile range (IQR), and Pearson correlations were used as well as area under the receiver operating characteristic curve (AUC) with 95% confidence intervals.

Results: GFAP and UCH-L1 values from the two platforms were strongly correlated (p = 0.940 and 0.866, respectively). GFAP values measured by Simoa platform (median 1162.7 [IQR]: 442.2–4383.5] pg/mL) were higher than values measured by i-STAT (median 232.0 [IQR: 89–899.25] pg/mL). UCH-L1 values measured by Simoa platform (median 116.38 [IQR: 39.49–214.34] pg/mL) were lower than values measured by i-STAT (median 317.00 [IQR: 232.5–654.5] pg/mL). When using iStat to detect traumatic intracranial lesions on CT, there were no differences found: GFAP AUC = .445 (95%CI = .307 to .583), UCHL1 AUC= .412 (95%CI = .275 to .548). When using Simoa platform to detect traumatic intracranial lesions on CT, there moderate findings GFAP AUC = .707 (95%CI = .657 to .757), UCHL1 AUC= .605 (95%CI = .549 to .660).

Conclusion: In this pilot study, Abbott i-STAT® POC device is comparable to Simoa analyzer. Further investigation into TBI biomarker expression in children is warranted.

D009 Long-Term Neurometabolite Concentrations in the PCG Relate to Professional Football Experience and Persistent Symptomology

Dr. Katherine Breedlove ,^1,2 Dr Rachel Grashow,^3,4 Sai Merugumala,^1,2 Ona Wu,⁵ Meagan Wasfy,⁶ Jacob Dodelson,⁵ Annelise Kulpanowski,⁵ Ross Zafonte,⁷ Aaron Baggish,^4,8 Alexander Lin^1,2

¹Center for Clinical Spectroscopy, Brigham And Women's Hospital, Boston, United States, ²Department of Radiology, Harvard Medical School, Boston, USA, ³Harvard T.H. Chan School of Public Health, Boston, USA, ⁴Football Players Health Study at Harvard University, Boston, USA, ⁵Athinoula A Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Boston, USA, ⁶Division of Cardiology, Department of Medicine, Massachusetts General Hospital, Boston, USA, ⁷Department of Physical Medicine and Rehabilitation, Harvard Medical School, and Spaulding Rehabilitation Hospital, Boston, USA, ⁸Institute for Sport Science and Department of Cardiology, Lausanne University Hospital, Lausanne, Switzerland

Introduction: Understanding the long-term implications of participating in contact sports is crucial for optimizing patient outcomes. Magnetic resonance spectroscopy (MRS), a non-invasive technique, is able to assess long-term changes in brain neurochemistry. Specifically, this analysis focused on N-acetyl-aspartate (NAA), a putative marker of viable neurons. Reduced NAA levels are associated with acute brain injury, dementia, and stroke, and NAA levels are known to decline with age. In previous studies, NAA was found to decrease with the number of cumulative head impacts in post-career American-style football players, but years of play and symptoms at the time of injury were not measured directly. This study examines these measures specifically to provide a more direct assessment of player health.

Methods: Demographic and 3T MRS data were acquired from 90 male former American football players (age = 48.49± 7.65 years, pro career length = 5.72±3.5) utilizing point-resolved spectroscopy (TR=2000, TE=30ms, voxel size=20x20x20mm3, 128 averages) in the posterior cingulate gyrus. MRS datasets were processed using OpenMRSLab and quantified with LCModel. The relationship between NAA and years of professional football play, concussion signs and symptoms at the time of injury, age, and body mass index was investigated using multiple linear regression.

NAA=β_0+β_1 Number professional seasons+β_2 Concussion Symptom Score+β_3 Age+β_4 BMI+ϵ

Results: The overall regression was statistically significant (R2 = 0.133, F = 3.272, p = 0.015) with a positive relationship between NAA, years of professional football play, and concussion symptom score.

Conclusion: While the unexpected positive model result challenges previous findings, several plausible explanations exist. The observed rise in NAA levels among subjects may be attributed to neuroplasticity resulting from chronic injury. Furthermore, considering that NAA has a moderating effect against neuroinflammation, the increase in NAA could be a consequence of this protective mechanism. However, further investigation is warranted to fully understand these results.

D010 Position specific alterations in default mode network functional connectivity after a season of collegiate American football

Owen Griffith ¹

¹Penn State University, STATE COLLEGE, United States

Introduction: Repetitive head impact (RHI) exposure in sports like American football is a topic of growing concern in clinical medicine, especially in association with neurodegenerative processes. Few objective measures of brain alterations after acute exposure to RHIs exist, however functional connectivity, measured using functional magnetic resonance imaging (fMRI), may provide a sensitive method for detecting minute changes in brain integrity. The primary goal was to examine alterations in functional connectivity of the DMN in collegiate American football players after a season of RHI exposure.

Methodology: Collegiate football players from five years (2015, 2017, 2019, 2021, 2022) were scanned on 3T magnetic resonance imaging prior to the start of the collegiate football season and one week after the season. fMRI data was analyzed using a seed-based functional connectivity method to determine temporal correlation between the DMN and other brain regions. Paired sample t-tests were performed to identify significant changes in functional connectivity from pre- to post-season, by player position (speed, non-speed), and by previous concussion history (yes, no).

Results: 67 players were included (mean age 20.8 years; 100% male; 32 (48%) non-speed (i.e. linemen) position players). Speed players demonstrated a significant decrease in functional connectivity between the left and right hemisphere posterior cingulate cortices of DMN and the left hemisphere precuneus of the Frontoparietal Network. There was no significant change across the season in functional connectivity by history of previous concussion.

Conclusions: These findings imply that there are position specific differences in DMN functional connectivity after a season of collegiate American football. Given that positions in football vary in their exposure rate and magnitude of RHIs, fewer hits at higher magnitude may induce greater alterations in DMN functional connectivity.

D011 Associations between instrumented mouthguard-measured head acceleration events and post-match biomarkers of astroglial and axonal injury in male amateur Australian football players

Dr William O’Brien ,¹ Miss Lauren Evans,¹ Dr Gerson Spitz,¹ Miss Becca Xie,¹ Professor Terence O’Brien,^1,2 Professor Sandy Shultz,^1,2 Dr Stuart McDonald^1,2

¹Monash University, Melbourne, Australia, ²Alfred Health, Melbourne, Australia

Introduction: Mounting concern surrounds the potential neuropathological consequences of non-concussive head impacts in collision sports. Advances in instrumented mouthguards (iMGs) allow for accurate quantification of both single high-acceleration head impacts and cumulative head acceleration exposure within a match. However, relationships between these measures and risk of brain cell injury remain unclear. This study aimed to quantify measures of non-concussive head impact exposure and assess their association with post-match blood biomarkers of astroglial (GFAP) and axonal (NfL, p-tau-181) injury in amateur male Australian football players.

Methodology: Thirty-two athletes underwent in-season (24h post-match) and post-season (>5 weeks after final match) blood collections and/or wore HITIQ Nexus A9 iMGs measuring peak linear (PLA) and rotational (PRA) acceleration. Match footage was used to verify and code impacts. Simoa quantified serum NfL and GFAP, and plasma p-tau-181.

Results: In-season versus post-season elevations were found for GFAP (mean dif.=8.07 pg/mL, 95%CI=0.28–15.86), NfL (mean dif.=1.43, 95%CI=0.58–2.28) and p-tau-181 (mean dif.=10.44, 95%CI=6.81–14.08). Post-match GFAP was associated with maximum single impact PLA (β=0.18 pg/mL per g, 95%CI=0.02–0.33) and PRA (β=1.85 pg/mL per krad/s2, 95%CI=0.04–3.66), and cumulative PLA (β=0.07 pg/mL per g, 95%CI=0.02–0.12) and PRA (β=0.70 pg/mL per krad/s2, 95%CI=0.20–1.20), and impact number (β=1.9 pg/mL per impact, 95%CI=0.34–3.47) within a single match. Change in NfL levels between two-matches correlated with cumulative PLA (r=0.73, 95%CI=0.23–0.92), PRA (r=0.66, 95%CI=0.10–0.90), and impact number (r=0.67, 95%CI=0.12–0.91).

Conclusions: Maximum and cumulative head accelerations in Australian football, measured by iMGs, were associated with elevated blood biomarkers of brain cell injury, highlighting the potential of both technologies for enhanced head impact management in collision sports.

D012 Brain Health Concerns in Retired Rugby Players: Clinical and Cognitive Phenotypes

Dr Thomas Parker ,^1,2,3 Jessica Hain,^1,2 Erin Rooney,^1,2,4 Dr Karl Zimmerman,^1,2 Dr Ying Lee,^1,2,4 Martina Del Giovane,^1,2 Dr Neil Graham,^1,2,5 Dr Maneesh Patel,⁶ Professor Mathew Wilson, Professor Adam Hampshire,¹ Dr Daniel Friedland, Professor David Sharp,^1,2,5 Dr Richard Sylvester^4,7

¹Department of Brain Sciences, Division of Medicine, Imperial College London, London, UK, ²UK Dementia Research Institute, Centre for Care, Research and Technology, London, UK, ³Department of Neurodegenerative Disease, The Dementia Research Centre, UCL Queen Square Institute of Neurology, London, UK, ⁴Institute of Sport, Exercise and Health, University College London, London, ⁵Centre for Injury Studies, Imperial College London, London, UK, ⁶Imaging Department, Imperial College Healthcare NHS Trust, Charing Cross Hospital, UK, ⁷Acute Stroke and Brain Injury Unit, National Hospital for Neurology and Neurosurgery, Queen Square, London, UK

Introduction: There is concern that elite rugby participation increases the risk of neurodegenerative disease, most notably chronic traumatic encephalopathy (CTE). Detailed clinical data from retired elite rugby players with brain health concerns is lacking. In particular, the rates of traumatic encephalopathy syndrome (TES), the clinical correlate of CTE, and the relationship between clinical outcomes and repetitive head impacts are unknown.

Methodology: 200 retired elite rugby players with brain health concerns and 33 matched healthy controls were assessed. Self-reported concussion history, career duration, player position, self-rated scales of depression, anxiety, sleep quality, post-concussion symptoms and quality of life, as well as self and informant ratings of neuropsychiatric symptoms and executive function behaviours, were obtained. Formal cognitive testing, TES classification and 3T MRI were performed.

Results: Elite rugby players (median age=44 years, 90.5% male, median self-reported career concussions=7, median career length=10.5 years, 63% forwards/37% backs) had elevated scores compared to controls on all symptom scales except sleep quality. Despite frequent subjective memory complaints, performance on cognitive testing did not significantly differ from controls. 24 retired players fulfilled criteria for TES (21 “suggestive of CTE”, 3 “possible”, 0 “probable/definite”). No players fulfilled criteria for dementia. Symptom burden was higher in those with higher self-reported concussions, but was unrelated to years of play or position played. Cavum septum pellucidum was visible on structural imaging in 24% of players (12% in controls, p=0.19). Microhaemorrhages on susceptibility weighted imaging were visible in 3% of players (0% of controls, p=0.82).

Conclusions: Retired elite rugby players in mid-life had significant neuropsychiatric symptoms, especially those self-reporting more concussions. Objective cognitive impairments and TES were uncommon and there was no evidence of dementia. These results suggest symptom burden in retired elite rugby players is generally not entirely explained by CTE and their relationship to head impact exposure is complex.

D013 Head impacts and changes in subcortical brain morphology in Australian football players: a single-season longitudinal study

Dr Spencer Roberts ,¹ Dr Juan Dominguez Duque,² Dr Stuart McDonald,³ Associate Professor David Wright,³ Dr Elise Facer-Childs,⁴ Dr Adam McKay,⁴ Dr Dominique condo,¹ Dr Aaron Fox,¹ Professor Brad Aisbett,¹ Professor Karen Caeyenberghs²

¹Institute for Physical Activity and Nutrition, School of Exercise and Nutrition Sciences, Deakin University, Burwood, Australia, ²Cognitive Neuroscience Unit, School of Psychology, Deakin University, Burwood, Australia, ³Department of Neuroscience, Monash University, Clayton, Australia, ⁴Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash University, Clayton, Australia

Introduction: Sport-related concussive and non-concussive head acceleration events (HAE) may alter subcortical brain structures.1,2 However, subcortical morphology had not previously been explored longitudinally in Australian footballers using instrumented mouthguards that quantify HAE exposure.

Methodology: Twelve male Australian footballers (mean ± SD age, 24 ± 4y) wore instrumented mouthguards (HITIQ, Melbourne) across a season. Mouthguards recorded the number of HAE per game, and the peak linear (g-forces) and angular (rad·s2) acceleration per HAE. Players had MRI scans (T1 mprage, voxel size 0.8mm3) taken at pre- and post-season timepoints. Volumes of subcortical regions (hippocampus, amygdala, cingulate cortex, thalamus, caudate, putamen) were computed using Freesurfer (v7.4.1). Changes in pre- vs post season region volumes were analysed using paired student t-tests, and relationships between changes in region volumes and season aggregated mouthguard metrics were explored using Spearman’s correlations.

Results: There we no statistical differences pre- vs post-season in amygdala, hippocampus, cingulate cortex, thalamus, or left putamen volumes. Relative to pre-season, post-season volumes were lower for the left (mean ± SD Δ, -1.7 ± 2.0%, p = .010) and right (mean ± SD Δ, -2.4 ± 2.4%, p = .006) caudate, and the right putamen (mean ± SD Δ, -2.7 ± 3.6%, p = .022). There was a statistically significant correlation between within-participant change (%) in left caudate volume and cumulative HAE count (r = -.74, p = .046). There were no other statistically significant correlations.

Conclusion: This pilot study of longitudinal changes in subcortical brain morphology in Australian footballers identified that, relative to pre-season, volumes of some subcortical structures (putamen, caudate) were reduced post-season. Evidence that within participant change in left caudate volume strongly correlated with cumulative number of HAE across the season should encourage further investigation of these relationships with larger samples and including HAEs recorded during training.

1. Schultz V., et al. Age at first exposure to repetitive head impacts is associated with Smaller thalamic volumes in former professional American football players. J Neurotrauma. 2018; 35: 278–285.

2. Wojtowicz., et al. Cortical thickness and subcortical brain volumes in professional rugby league players. NeuroImage. 2018; 18: 377–381.

D014 Demographics and symptomatic burden in athletes attending a specialist sports concussion clinic

Mr Andrew Stevens ,^1,2 Dr Kamal Yakoub,¹ Mr Sheikh Momin,^1,2 Ms Emily Paton,² Ms Holly Davies,² Dr David Hacker,² Mr Phil O’Halloran,² Mr David Davies,^1,2 Prof. Antonio Belli^1,2

¹University Of Birmingham, Birmingham, United Kingdom, ²University Hospitals Birmingham, Birmingham, United Kingdom

Introduction: Sports-related concussion is a considerable cause of morbidity, predominantly amongst people of working age [1]. Complications, such as post-concussion syndrome or managing risk from recurrent concussion, require specialist management which remains an uncommon clinical service in the UK [2]. In support of their wider provision, we sought to present the demand for one such service: Birmingham Sports Concussion Clinic (BSCC).

Methods: Records of attendances were examined from a sample of 138 athletes attending BSCC between June 2021 and May 2022. As part of standard clinical assessment, all players underwent SCAT-5 symptom evaluation; mBESS balance testing; and vestibular oculomotor screening (VOMS) assessment.

Results: Median time from concussion to attendance was 45 days (IQR=24–87). Sports were predominantly rugby (75%) and football (10%), with 1:4.4 female to male ratio. 24% had sustained ≥2 concussions in 3m and/or ≥3 concussions in 12m. 60% of players were symptomatic on attendance with a SCAT-5 symptom severity score of ≥5, and 29% reported a symptom severity score of ≥20. Headache was the most commonly reported symptom (32%). 42% had vestibular/oculomotor dysfunction based on VOMS scoring, and 9% had abnormal balance on mBESS testing. 14% reported depressive symptoms based on PHQ-9 scores (4% moderate/severe). 10% reported anxiety symptoms based on GAD-7 score (2% moderate/severe). 34/136 were attending ≥3m post-injury: among this group, average symptom severity score was 29 (range 0–80), and 44% had severe provocation during VOMS testing. MRI was performed in 46%, medication was prescribed to 36% of attendees, and vestibular/oculomotor physiotherapy advised for 35%.

Conclusion: Post-concussion syndrome is a common cause of morbidity in athletes attending BSCC, often resulting in a complex cluster of debilitating symptoms. Appropriate management requires a multi-disciplinary approach [2], and the high numbers attending with severe symptoms persisting ≥3m post-injury suggests an urgent need for improved availability of specialist services nationally.

D015 Phase-contrast magnetic resonance imaging reveals changes in cerebrospinal fluid dynamics associated with subarachnoid space occlusion following traumatic thoracic spinal cord injury in the pig

Ms Madeleine Bessen ,^1,2 Ms Christine D Gayen,^2,3 Dr Ryan O’Hare Doig,^1,4,5 Mr Ryan M Dorrian,³ Dr Ryan D Quarrington,^1,2 Mr Adnan Mulaibrahimovic,^1,2 Professor Vartan Kurtcuoglu,⁶ Mrs Angela V Walls,⁵ Dr Anna V Leonard,³ Associate Professor Claire F Jones⁸

¹The University of Adelaide, Adelaide Medical School, Adelaide, Australia, ²The University of Adelaide, Centre for Orthopaedics & Trauma Research, Adelaide, Australia, ³The University of Adelaide, School of Biomedicine, Adelaide, Australia, ⁴South Australian Health and Medical Research Institute (SAHMRI); Neil Sachse Centre for Spinal Cord Research, Lifelong Health Theme, Adelaide, Australia, ⁵National Imaging Facility, South Australian Node, SAHMRI, Adelaide, Australia, ⁶SAHMRI, Clinical and Research Imaging Centre, Adelaide, Australia, ⁷University of Zurich, Institute of Physiology, Zurich, Switzerland, ⁸The University of Adelaide, School of Electrical & Mechanical Engineering, Adelaide, Australia

Introduction: Pulsatile dynamics of cerebrospinal fluid (CSF) may be altered by traumatic spinal cord injury (SCI) due to spinal cord swelling and occlusion of the subarachnoid space (SAS). However, the magnitude and temporal profile of such potential changes remain unknown and were therefore investigated over 14 days post-SCI in pigs.

Methodology: A T10 contusion SCI was induced in female domestic pigs using a weight-drop apparatus (N=5, 10 cm; N=5, 20 cm). Magnetic resonance imaging (MRI) was performed pre-SCI and at 3, 7 and 14 days post-SCI. SAS occlusion length (normalised to T8/T9−T11/T12 length), and injury-site SAS area, were measured using T2-weighted MRI. CSF dynamics, specifically peak mean velocity in the cranial/caudal direction and the corresponding time-to-peak (% cardiac cycle), were measured using phase-contrast MRI (PC-MRI) at C2/C3, T8/T9, T11/T12 and L1/L2. Linear-mixed effects models (a=0.05) were developed to assess effect of: (1) injury group and time-point on SAS occlusion; and (2) time-point and spinal level (adjusted for injury group) on CSF dynamics.

Results: SAS occlusion length decreased from 3 to 7 days (10cm: 11.4%; 20cm: 11.4%), and 7 to 14 days (10 cm: 6.9%; 20cm: 14.4%) post-SCI. SAS area decreased at day 3 (10cm: -22.79mm2; 20cm: -27.78mm2) and increased from 7 to 14 days (10cm: 10.45mm2; 20cm: 6.31mm2). At all spinal levels, peak mean cranial/caudal velocity (cranial: -0.33cm/s: caudal: -0.35cm/s) and time-to-peak mean caudal velocity (-11.10%) decreased at day 3 post-SCI, and peak mean caudal velocity (0.32cm/s) and time-to-peak mean caudal velocity (8.20%) increased towards baseline values, from 3 to 14 days post-SCI.

Conclusions: Changes in CSF dynamics, concurrently with SAS occlusion, occurred following SCI in this pre-clinical model. These suggest altered CSF pulsatility and craniospinal compliance. CSF dynamics derived from PC-MRI may help detect functional changes in the spinal intrathecal space, which could have implications for post-SCI management decisions.

D016 31P magnetic resonance spectroscopic measurement of alkalosis and high phosphocreatine/ATP ratio in the acutely traumatised brain: relationship to clinical outcome

Dr Matthew Stovell,¹ Dr Marius Mada,¹ Dr Christopher Wickens,¹ Dr Richard Ansorge,¹ Dr Tommi Korhonen,^1,2 Dr Thomas van Essen,^1,3 Professor Adel Helmy,¹ Professor David Menon,¹ Professor Adrian Carpenter,¹ Professor Peter Hutchinson,¹ Dr Keri Carpenter ¹

¹University of Cambridge, Cambridge, United Kingdom, ²Oulu University Hospital and University of Oulu, Oulu, Finland, ³Leiden University Medical Center, Haaglanden Medical Center, Leiden and The Hague, The Netherlands

Introduction: Although changes in brain glucose metabolism and NADH/NAD+ ratio (represented by lactate/pyruvate ratio) are known in the critical early post-injury period of TBI, it was previously unknown whether these translated to downstream changes in ATP metabolism and intracellular pH. We studied changes in brain energetics and brain pH in TBI patients during the acute phase by 31P MRS in-vivo and compared results with clinical outcome 6 months later.

Methodology: Oblique 2-dimensional single-slice chemical shift MRS was performed using a custom-built (PulseTeq) birdcage 31P head-coil on Siemens 3T MRI scanners. Data analysis focused on 8 central 2.5x2.5x2.5 cm3 voxels per subject. The PCr/gATP ratio (energy status), and intracellular pH (calculated from chemical shift difference between PCr and inorganic phosphate) were measured in 13 sedated, ventilated TBI patients and 10 age-group-matched healthy controls (HC). Clinical outcomes 6 months after 31P MRS were dichotomised: unfavourable (6 patients GOS-E 1–3) and favourable (7 patients; GOS-E ≥4).

Results: PCr/gATP ratio was significantly higher in TBI patients (median 1.09) than HC (median 0.93) (p<0.0001). PCr/gATP ratio was not significantly different between TBI outcomes. Brain pH was higher in TBI patients (median 7.04) than HC (median 7.00) (p=0.04), residing in patients with unfavourable outcome (median 7.07) (p<0.0001). These changes in PCr/gATP ratio and pH were still apparent when voxels with >5% radiologically visible injury were excluded. Machine learning algorithms using 31P MRS performed well above chance level, with the support vector machine (SVM) giving the best balanced accuracy (81.5%) for dichotomised outcome prediction.

Conclusions: 31P MRS revealed widespread biochemical injury, invisible by conventional 1H imaging. The degree of brain alkalosis (pH) demonstrated a difference between patients with unfavourable and favourable outcome. 31P MRS, combined with machine learning, has potential in brain injury, for non-invasive diagnosis, prognosis, and assessing therapy. Extension to mild TBI patients is ongoing.

D017 Brain Point of Care Ultrasound for Post-hemicraniectomy TBI Patients

Dr. Jefferson Chen ,¹ Dr. Walter Valesky,¹ Dr. Patrick Chen,¹ Dr. Isidora Beach,¹ Ms Cassie Poole¹

¹University Of California, Irvine, Orange, United States

Introduction: Brain point-of-care-ultrasound (B-POCUS) is a feasible technique with which to assess intracranial pathology in the setting of hemicraniectomy patients.

Methodology: Images were obtained with a Mindray M9 ultrasound machine using a 1.4–5.1-megahertz curvilinear transducer. A standard technique was developed for B-POCUS in two planes.

1. Patient is positioned to access the hemicraniectomy site.

2. The transducer is positioned in the axial plane at the level of the coronal suture, the probe is moved in the superior to inferior planes imaging the basal ganglia, internal capsule, midbrain, ventricles, and cortex.

3. The probe is placed in the para-coronal plane to visualize the falx cerebri and lateral ventricles, moving anteriorly to posteriorly.

4. Finally, the probe is repositioned in the axial plane at the level of the midbrain and color flow doppler is used to visualize the circle of Willis.

Results: Three post-hemicraniectomy patients were piloted. Visualization of the anatomy was identified with consistent patterns noted.

1. Qualitative differences between infarcted and non-infarcted tissue were notable with the ultrasound. Loss of the normal cerebral architecture was demonstrated as hyperechoic signal in the infarcted tissue.

2. Qualitative and quantitative differences of vascular blood flow could be assessed by color doppler and power doppler, respectively.

3. Hematoma size was reliably measured using ultrasound in comparison with computed tomography scans.

4. Global cerebral edema and brainstem compression could be assessed by visualization of the basal cisterns.

5. Ventricular catheter placement was identifiable within the ventricles

Conclusions: Cerebral imaging of post hemicraniectomy patients is feasible with B-POCUS to evaluate anatomic structures and clinically significant pathology. This imaging modality is useful in patients who cannot be transported for radiologic imaging due to hemodynamic and cerebral instability. Serial ultrasound imaging may readily be accomplished and may be a means of tracking the cerebral injury.

D018 The fate of bifrontal contusions in good coma score patients and a novel bleed-to-brain ratio

Lieutenant Colonel Sushila Chauhan,² Professor Sasidharan Gopalakrishnan ¹

¹Jawaharlal Institute of Post-graduate Medical Education and Research (JIPMER), Pondicherry, India, ²The Indian Army, Command hospital, Chandimandir, India

Introduction: Patients with frontal basal contusions (BFCs) tend to have unexpectedly bad outcomes even when the initial neurological conditions were good. We wanted to apply a ratio we had developed earlier called the bleed-to-brain ratio (BBR), to see how many of the decisions (to operate or conservatively manage a patient) could have changed if the ratio was used.

Methods: We did a retrospective cohort study of a consecutive series of adult, closed, traumatic brain injury patients with BFCs with a motor score of obeying commands or localising pain and studied their in-hospital mortality. We compared the mortality rate in the ‘congruent group’ (whether the BBR-based decision and the eventual decision to operate or not, matched) with that in the non-congruent group.

Results: In 103 out of the 159 (64.8%) patients with BFC, BBR exceeded the threshold in the first scan itself. 78 of these patients (75.7%) were eventually operated on. When BBR exceeded the threshold in the first or subsequent scans and the operating team did not operate - which we called the incongruent conservatively managed group - there was a 16% mortality (4 out of 25 patients. There were 6 out of 78 patients (7.7%) in the congruent operated group (where the BBR exceeded the threshold and the treating team operated on the patients). These differences, however, were not significant (p = 0.268)

The treating surgeons had also operated on a large number of patients ( 35,30.7 %), where the BBR never exceeded the threshold (0.0535). The mortality in this group was 3 (8.6%).

Conclusions: BBR calculated using Hounsfield bracketing is a promising tool to aid in decision-making in BFC and has a strong negative predictive value for surgery. Further prospective studies are needed to validate this variable and determine the cut-off value for clinical relevance.

D019 Does poor adherence to the National Institute for Health and Care Excellence (NICE) criteria for emergency CT scans in TBI patients lead to excessive radiation exposure? A cohort study

Mr Zahid Sayeed Iqbal,¹ Dr M S Gopalakrishnan ¹

¹Jawaharlal Institute Of Post Graduate Medical Education And Research, Puducherry, India, Puducherry, India

Introduction: NICE guidelines cover the early management of traumatic brain injury (TBI), including imaging in children and adults. In our emergency room, adherence may be suboptimal. We wanted to find the adherence rate and the consequential excess radiation exposure.

Methodology: We extracted the list of TBI patients who underwent emergency CT scans in October 2023 and analyzed their records to see if patients matched any of the NICE criteria. We assumed that the proportion of patients with no risk factors according to NICE guidelines had been unnecessarily exposed to CT scan radiation.. We noted the positive radiological findings in both indicated and unindicated scans and checked if they changed patient management. We converted the dose length product to an effective dose using an online calculator to quantify the radiation exposure. We recorded the duration of patients’ hospital stays and determined the relationship between unindicated scans and early discharge using Chi Square test.

Results: We found 254 TBI patients (92.5% adults, 7.5% children) who had undergone emergency CT scans. 118 patients (46.4%) had been scanned without a NICE criteria match. Each patient who underwent an unindicated scan was exposed to a median radiation dose of 1.7 mSv. Interestingly, 27.1% of the unindicated scans showed positive radiological findings, the most common being a simple skull fracture (50%) followed by a contusion (21.8%). These findings affected the management of the patient in 53.1% of the cases with the most common change being the addition of an anticonvulsant (53%). There was a statistically significant relation (p=0.002) between unindicated scans and shorter duration of stay.

Conclusion: Almost 50% of the scans were unnecessary, exposing our patients to radiation, but it speeded up discharge. It remains questionable whether strict implementation of these guidelines would result in better clinical outcomes in a congested emergency department like ours.

D020 Sex differences in repetitive head impact exposure effects on white matter neurometabolism

Julie M. Joyce ,^1,2,3 Luisa M Schuhmacher,^2,4,5 Alberto Villagran,^2,4 Anja K Betz,^2,4,5 Dr. Zhou Lan,³ Dr. Alexander P Lin,³ Dr. Inga K Koerte^2,4,5

¹Cumming School of Medicine, University Of Calgary, Calgary, Canada, ²cBRAIN, Department of Child and Adolescent Psychiatry, Psychosomatics, and Psychotherapy, Ludwig-Maximilian University, Munich, Germany, ³Center for Clinical Spectroscopy, Department of Radiology, Mass General Brigham, Harvard Medical School, Boston, United States, ⁴NICUM (NeuroImaging Core Unit Munich), Munich, Germany, ⁵Psychiatry Neuroimaging Laboratory, Department of Psychiatry, Boston, United States

Introduction: The consequences of repetitive head impacts (RHI) on brain health in young soccer players remain poorly understood. There is initial evidence of sex-specific differences in the effects of RHI. The neurometabolite N-acetyl-aspartate (tNAA) serves as a sensitive indicator of brain health and can be measured with magnetic resonance spectroscopic imaging (MRSI). This study examined the relationship between RHI exposure and neurometabolites, and specifically, sex-specific effects.

Methodology: Fifty-two soccer players (33M,19F) and fifty-two control athletes (22M,30F) age 20–25y completed a MRSI scan on a 3T Siemens PRISMA MRI. Control athletes had <5y of cumulative soccer exposure. A T1-weighted scan was acquired for MRSI slice placement (centrum semiovale) and tissue segmentation to gray and white matter. MRSI were summed for white matter and gray matter, respectively. Neurometabolites were calculated as ratios to an internal reference (total creatine, tCr). A two-way ANCOVA was conducted to examine the relationship between neurometabolites and predictor variables (i.e., RHI exposure and sex) with age as a covariate and assessed for interaction effects between RHI exposure and sex with post-hoc pairwise comparisons, adjusted using Tukey's correction in R.

Results: Duration of participation in soccer ranged from 5.0–21.0y (mean=13.5±4.2y). A significant interaction effect emerged between years of RHI exposure and white matter tNAA/tCr levels among male and female participants, F(4,99)=6.75, p=0.011. Male soccer players exhibited significantly lower levels of white matter tNAA/tCr compared to male controls (mean difference=-0.065, p=0.013). This difference was not observed in female athletes (mean difference=0.045, p=0.146), despite comparable years of play across sexes.

Conclusion: Lower levels of tNAA/tCr in male soccer players relative to controls reflect altered neurometabolism and may indicate a subtle white matter microstructural injury incurred over time. Further exploration is warranted regarding the absence of tNAA/tCr differences in female athletes, potentially indicating variations in exposure, resilience, or compensatory mechanisms.

D021 Alterations in resting-state functional connectivity of cerebellar networks following photobiomodulation treatment in adults with a history of repetitive head acceleration events

Finian Keleher ,^1,2 PhD Carrie Esopenko,^1,4 PhD Hannah M. Lindsey,^1,2 PhD Mary R. Newsome,^1,2,3 PhD Paula K. Johnson,^1,2,5 PhD Divya Jain,⁴ Elizabeth S. Hovenden,¹ BS Dayna Thayn,^1,5 Courtney McCabe,¹ MS Hilary A. Russel,¹ PhD Christine M. Mullen,¹ MA Carmen Velez,^1,2 BA Emma N. Read,¹ PhD Lance E. Davidson,^1,5 PhD Michael J. Larson,⁵ PhD Spencer W. Liebel,^1,2 PhD David Tate,^1,2 PhD Lawrence S. Carr,^1,5 PhD Elisabeth A. Wilde^1,2,3

¹University Of Utah, Sandy, United States, ²George E. Wahlen VA Medical Center, Salt Lake City, United States, ³Baylor College of Medicine, Houston, United States, ⁴Icahn School of Medicine at Mount Sinai, New York City, United States, ⁵Brigham Young University, Provo, United States

Introduction: There is growing concern that repetitive head acceleration events (RHAE) may have long-term consequences on brain health, including alterations in the resting-state functional connectivity (rsFC) of large-scale brain networks. The current study examined whether a non-invasive treatment called transcranial photobiomodulation (tPBM), which is thought to improve mitochondrial function, enhance circulation, and reduce inflammation, is associated with alterations in rsFC of cerebellar networks in individuals with a history of RHAE. We hypothesized that the use of tPBM would alter rsFC between cerebellar seed regions of the default mode (DMN), frontoparietal (FPN), and salience (SN) networks and the rest of the brain.

Method: Thirty participants with a history of RHAE (M age= 46.3 ± 15.2 years) used a tPBM headset for 20 minutes 3–5 times/week for 8–10 weeks and completed resting-state functional magnetic resonance imaging before and after treatment. Seed-to-voxel based rsFC was examined in 10 cerebellar seed regions within the DMN, FPN, and SN pre- and post-treatment.

Results: TPBM treatment seemed to enhance rsFC in the SN by decreasing between-network rsFC (superior frontal gyrus (SFG)) and increasing within-network rsFC (thalamic white matter). Similarly, tPBM seemed to have an ameliorative effect on the FPN by increasing within-network rsFC (superior parietal lobule) and decreasing between-network rsFC (medial frontal gyrus), although between-network rsFC also increased in the SFG after treatment. TPBM had varied and inconsistent effects on the DMN, as bidirectional changes were found in rsFC both within and between networks after treatment. Additionally, rsFC increased between regions of the DMN and FPN, two networks often reported to be active at different times.

Conclusion: Preliminary results suggest that tPBM may have some ameliorative effects in normalizing rsFC between the cerebellum and large-scale functional networks. However, further research is needed to understand the full impact of tPBM on rsFC, including increases in between-network rsFC.

D022 Enhanced recovery by negative allosteric modulation of the metabotropic glutamate receptor 5 by 3-((2-Methyl-4-thiazolyl)ethynyl)pyridine following experimental traumatic brain injury

Dr. Jakob Hakon,¹ Dr. Miriana Quattromani,¹ Prof. Niklas Marklund,¹ Prof. Tadeusz Wieloch,¹ Dr. Karsten Ruscher ¹

¹Lund University, Lund, Sweden

Introduction: A high percentage of patients with acquired traumatic brain injury suffer from insufficient recovery of lost brain functions and experience life-long disability. So far, therapeutic and rehabilitation approaches do not provide sufficient success to improve recovery for various reasons. We recently reported that negative allosteric modulation of the metabotropic receptors type 5 (mGluR5) improves to stroke outcome. The aim of the present study, therefore, was to investigate if treatment with 3-((2-Methyl-4-thiazolyl)ethynyl)pyridine (MTEP) enhances recovery of lost neurological function following experimental traumatic brain injury.

Methodology: Experimental traumatic brain injury was induced in male Sprague Dawley rats by controlled cortical impact (CCI). On day 2, functional deficits were assessed by the paw placement test and a composite Neuroscore. Only rats showing a significant neurological deficit were randomized into treatment groups receiving either MTEP (5 mg/kg) (n = 5) or saline (n = 5) for a total 12 days. Functional outcome was subsequently assessed using a composite Neuroscore and the paw placement test at 7 and 14 days.

Results: Treatment with MTEP significantly improved performance in the composite neuroscore as early as four days after the insult. At later time points, test scores in the MTEP-treated group persisted while the performance of rats in the vehicle-treated groups improved to levels found in MTEP-treated animals due to spontaneous recovery. In addition, fore- and hindlimb function was significantly improved in the paw placement test in the MTEP-treated group at 14 days after CCI. Better functional outcome was not related to differences in lesion volumes.

Conclusions: Our data support the idea that negative allosteric modulators of the mGluR5 can accelerate the recovery process following experimental traumatic injury. Inhibition of mGluR5 might be involved in modulating synaptic function strengthening processes of adaptive plasticity after acquired brain injury.

D023 Hormesis, Post-Conditioning and Posttraumatic Recovery of Function: Underlying Molecular Mechanisms

Bogdan Stoica ,¹ James Barrett,¹ Alan Faden

¹University Of Maryland School Of Medicine, Baltimore, United States

Introduction: Hormesis is a term applied to integrative adaptive processes in response to external stressors or toxins. Hormetic effects from lower stimulation stressors can provide protection against subsequent, more severe stressors or insults; the degree of protection has been remarkably consistent across many models and species, suggesting that the extent of the protective response may reflect the limits of biological resilience. Pre-conditioning and post-conditioning protective interventions reflect this broader, evolutionarily developed process of hormesis, and the timing, frequency, and duration of the stimuli. Post-conditioning using a variety of stressors, can provide protection and improve recovery from many disorders, including central nervous system trauma and dementia.

Methodology: Using high-throughput gene expression analysis and associated molecular pathway evaluation, we examined the effects of different post-conditioning exercise interventions to address potential molecular mechanisms associated with the hormetic response. Paradigm differences included timing of treatment initiation, exercise duration, and time of follow-up measurements. We employed unsupervised clustering methods and pathway activation Z-scores to provide a more global way of assessing significant molecular pathway changes related to treatment.

Results and Conclusions: Delayed post-conditioning exercise interventions significantly reduced chronic posttraumatic neuroinflammation and related cognitive changes. Posttraumatic molecular pathway changes show different patterns related to exercise intervention in the hippocampus as compared to the cortex. Exercise comprehensively reversed injury effects in the hippocampus across activated (primarily inflammatory) and inhibited (in part neuronal) pathways, supporting a return toward the non-injured homeostatic state; in contrast, the cortex showed more limited attenuation of inflammatory pathway activation and greater inhibition of selected non-inflammatory pathways triggered by injury. Earlier exercise intervention beginning two weeks after injury was less effective than that started at four weeks, and four-week exercise duration was superior to 2-week duration. More delayed exercise, begun at ten weeks after trauma and continued for eight weeks was not as protective.

D024 Gut-brain axis after traumatic brain injury: morphometric parameters of gut dysfunction correlate with neurological deficits

Francesca Tribuzio ,¹ Dr Francesca Pischiutta,¹ Francesca Buffelli,² Costanza Bertani,¹ Luther Loose,¹ Dr Federico Moro,¹ Dr Annamaria Vezzani,² Teresa Ravizza,² Dr Elisa Zanier¹

¹Laboratory of Traumatic Brain Injury and Neuroprotection, Department of Acute Brain and Cardiovascular Injury, Istituto di Ricerche Farmacologiche Mario Negri, Milan, Italy, ²Laboratory of Epilepsy and Therapeutic Strategies, Department of Acute Brain and Cardiovascular Injury, Istituto di Ricerche Farmacologiche Mario Negri, Milan, Italy

Introduction: Emerging data suggest that traumatic brain injury (TBI) is associated with gut inflammation, increased permeability, and changes in microbiota, leading to intestinal dysfunction. However, whether and how these changes may influence injury progression and neurological sequelae remains poorly understood. Here we investigate whether TBI induces chronic gut alterations and if they correlate with neurological impairments.

Methodology: Adult male C57BL/6J sham or severe TBI mice (n=10/group) were subjected to longitudinal evaluation of sensorimotor function (SNAP and Neuroscore, NS) up to 2-months post-injury. Mice were then histologically evaluated for gut morphological changes.

Results: Two-months after TBI, a significant reduction of gut length was observed (sham:46.8±1.8; TBI:45.0±1.1, p<0.05) with major involvement of small intestine (p<0.05). Histomorphometric analyses on duodena in TBI vs sham mice revealed: a shortened villus height (-21%, p<0.001), an increased crypt depth (+7%, p<0.05), a decreased villus height/crypt depth ratio (-26%, p<0.001) and reduction of Goblet cells/villus (-35%, p<0.001). Notably, when relating gut morphometric parameters with neurological outcomes in TBI mice, we found: i) a direct correlation between the shortening of small intestine and the worsening in the sensorimotor performance at 1w (SNAP, p<0.01, r=0.9) and 5w (SNAP, p<0.05, r=0.7); ii) an inverse correlation between the number of the Goblet cells/villus and sensorimotor recovery rate (SNAP 7w-1w, p<0.05; NS 7w-1w, p<0.05).

Conclusions: Our study shows a close relationship between TBI and gut morphometric alterations, with shortening of gut length and reduction of Goblet cells number associated with worse TBI outcome. Ongoing longitudinal analyses on blood, urine, and feces are aimed at identifying early biomarkers that could serve as prognostic indicators for TBI-related gut-brain axis alterations. Additionally, we are investigating the potential impact of gender on this response. These data will help in the design of gut-based interventions aimed at mitigating gut dysfunction with possible beneficial effects on TBI-associated neuropathology.

D025 Characterization of acute cellular responses in an in vitro model of traumatic brain injury

Noemi Wachtler ,^1,2 Doctor Xiangyu Gong,³ Doctor Zerin Mahzabin Khan,³ Professor Michael Mak,³ Professor Barbara Ehrlich,¹ Professor Declan McGuone¹

¹Yale School Of Medicine, New Haven, United States, ²TUM School of Medicine and Health, Munich, Germany, ³Yale School of Engineering & Applied Science, New Haven, United States

Introduction: Traumatic brain injury (TBI) is a global healthcare concern, and a leading cause of death and disability. Despite decades of research, there are currently no effective therapies to promote brain recovery after TBI, underscoring the necessity for improved understanding of early cellular responses in TBI. Existing in vitro models of non-penetrating TBI include blast, stretch, compression, and shear stress models. Our model focuses on molecular and cellular changes induced by mechanical compression, simulating the initial impact and subsequent compressive forces employed through neuronal swelling.

Methodology: Human neuroblastoma cell lines (SK-N-AS) were subjected to 100 Pascal compressive stress for one hour. Calcium imaging performed before, during, and after removal of compression forces was correlated with alterations in cell structure and immunophenotype to identify early changes in cellular phenotype in response to injury.

Results: Neuroblastoma cells showed marked disruption in calcium homeostasis after compression. Live imaging revealed striking morphological responses in compressed cells such as swelling and membrane blebbing, underscoring the interplay between biomechanical forces and cell signalling pathways. Intracellular calcium dysregulation is known to trigger various signalling cascades, including protein expression and apoptosis. Accordingly, we found a significant (p<0.001) elevation in the expression of proteins associated with cellular stress, including amyloid precursor protein (APP), TAR DNA-binding protein 43 (TDP-43), inositol 1,4,5-trisphosphate receptor type 3 (ITPR-3), activating transcription factor 4 (ATF4), calpain, and sterile alpha and TIR motif-containing protein 1 (Sarm1) in compressed cells. We are actively investigating apoptosis after compression in this model.

Conclusion: This in vitro model of compression helps validate prior cellular models of TBI and helps expand our knowledge of the biochemical and morphological events regulating acute responses to TBI. The increased expression of cell stress proteins early after injury has several different implications for acute and long-term pathophysiology and therapeutic interventions for preservation of neuronal function.

D027 Modulating microRNA-15a/16–1 to improve long-term functional recovery after traumatic brain injury

Dr. Chao Zhou,¹ Dr. Milton H. Hamblin,² Dr. Kejie Yin ^1,3

¹University of Pittsburgh School of Medicine, ²University of California Riverside School of Medicine, ³Veterans Affairs Pittsburgh Healthcare System

Abstract not published

D026 Interleukin-4: a promising new agent for treating traumatic brain injury?

Dr. Johannes Walter ,¹ Samuel Hutagalung,¹ Jannis Mende,¹ Noah Kruck,¹ Olga Kovalenko,¹ Dr. Obada T. Alhalabi,¹ Dr. Christa Maurer,² Prof. Joachim Kirsch,² Prof. Thomas Skutella,² Prof. Sandro M. Krieg,¹ Prof. Andreas W. Unterberg,¹ Prof. Klaus Zweckberger,¹ PD Dr. Alexander Younsi¹

¹Department for Neurosurgery, Heidelberg University Hospital, Heidelberg, Germany, ²Institute for Anatomy and Cell Biology, University of Heidelberg, Heidelberg, Germany

Introduction: New options for the treatment of traumatic brain injury (TBI) are desperately needed. The anti-inflammatory Interleukin-4 (IL-4) has been shown to be neuroprotective in several preclinical models of acute neuronal injury; therefore, we assessed its value in treating experimental TBI.

Methodology: A total of 178 C57/Bl6 wildtype and 20 C57BL/6-Il4tm1Nnt/J IL-4 knockout mice were subjected to controlled cortical impact (CCI). IL-4 was administered subcutaneously to wildtype mice as a single dose 15 minutes after CCI or as daily doses (all 5mg/kgBW) for the first 7 days post injury (dpi) while control animals received phosphate buffered saline. Neurological function (hole board, video open field and CatWalk XT® gait analysis), contusion volume (Nissl staining) and the local (immunofluorescent staining) as well as the systemic (bead-based multiplex immunoassay) immune response were assessed up to 28 dpi.

Results: Within the first 7 dpi, lesion volumes were significantly higher (e.g. 13.45 ± 0.88 mm³ vs. 9.50 ± 0.97 mm³, p=0.02 1 dpi) and neurological function significantly more impaired (e.g., difference in left forepaw area: -0.01± 0.01cm² vs. 0.05 ± 0.01cm², p=0.02 7 dpi) in IL-4 knockout mice compared to wildtype mice. The single dose administration of IL-4 significantly attenuated the local inflammatory response in the acute phase after CCI (e.g., 1.79 ± 0.15 Iba-1+/CD86+ cells/sROI vs. 1.06 ± 0.21 Iba-1/CD86+ cells/sROI, p = 0.03 in the penumbra 3 dpi) and reduced lesion volumes up to 14 dpi (6.82 ± 0.65 mm3 vs. 10.85 ± 1.19 mm3, p= 0.02 for IL-4 and control 7 dpi). Surprisingly, repetitive IL-4 administrations did not reduce lesion volumes within the first 28 dpi and had no effect on the systemic inflammatory response.

Conclusions: IL-4 is an emerging experimental therapeutic option for TBI; however, its optimal treatment regimen needs to be established on order to exploit its full neuroprotective potential.

D028 Enhancing neuronal regeneration in traumatic brain injury via Ascl1-mediated astrocyte-to-neuron reprogramming

Johannes Tödt,¹ Xunlei Zhou,² Guoli Zheng,² Anna Schnaubelt,² Maryam Hatami,² Thomas Skutella,² Andreas Unterberg,¹ Pd Dr. Med. Alexander Younsi ¹

¹Department of Neurosurgery Heidelberg University Hospital, Heidelberg, Germany, ²Department of Neuroanatomy Heidelberg University, Heidelberg, Germany

Introduction: Traumatic brain injury (TBI) often leads to significant neuronal loss, accompanied by astrogliosis and scar formation, which collectively impair the regenerative capabilities of the central nervous system. Our study explores the potential of reprogramming astrocytes to neurons using the transcription factor Ascl1 to mitigate astrogliosis effects and foster neuronal regeneration in a TBI rat model.

Methodology: Cortical astrocytes from rats were cultured and subjected to lentiviral transduction to induce overexpression of the Ascl1 gene. The reprogramming was traced using GFAP-Cre recombinase for astrocytes and Synapsin1-mScarlet for neurons. Cellular changes were analyzed at day 7 and 14 post-transduction through immunocytochemistry (ICC). For in vivo application, adeno-associated viral (AAV) vectors were developed, purified, and quantified. A controlled cortical impact (CCI) or sham surgery was induced in 36 male Wistar rats, and vectors or placebo were injected into the perilesional area 7 days thereafter. Behavioral analyses, including Rotarod, Open Field, and CatWalk, evaluated functional recovery over 38 days. Lineage tracing and immunohistochemistry (IHC) verified neuronal conversion and assessed neuroinflammation.

Results: Significant in vitro astrocyte-to-neuron (AtN) reprogramming was observed, with a 60% induction of mScarlet-positive neurons in the Ascl1 group versus 5% in controls at 14 days, indicating successful conversion (p=0.03). Enhanced neuronal marker expression was noted, with Synapsin1 and Beta-Tubulin 3 showing increases of 70% and 75%, respectively, compared to controls (p=0.04 and p=0.01). In vivo, Ascl1-overexpressing AAV-treated rats demonstrated superior functional outcomes (e.g., 33s longer run duration in the Rotarod test or 0.5 cm2 wider print width in the CatWalk gait analysis) compared with controls, correlating with effective AtN reprogramming and reduced neuroinflammation.

Conclusions: An in vitro lentiviral approach for AtN was successfully established and translated into an in vivo AAV-based rat TBI model. The strategy showed promising functional improvements, suggesting its therapeutic potential in TBI treatment.

D029 Active CSF exchange lowers complications and allows superior drug delivery compared to traditional EVDs

Dr Nicholas Brandmeir ¹

¹West Virginia University, Morgantown, United States

Introduction: External ventricular drains (EVD) are a staple in the management of brain injury. While they are lifesaving, EVDs have a high rate of complications including occlusion and infection. Further, the blood brain barrier (BBB) limits the effectiveness of many pharmacotherapies. Intrathecal (IT) medication delivery can overcome this limitation by directly crossing the BBB for drug delivery. An Active Cerebrospinal fluid (CSF) Exchange (ACE) system used in lieu of an EVD could allow delivery of IT medications as well as potentially avoid common complications of EVDs.

Methodology: We conducted a multi-institutional, international retrospective cohort trial comparing standard EVDs with an active CSF exchange catheter (IRRAflow, IRRAS Inc., San Diego, USA). Electronic medical records at 4 large academic hospitals in the US and Europe were reviewed and data abstracted from patients treated with EVD or ACE. Complications were compared between the 2 groups as well determining the rates of IT medication use.

Results: 118 patients received ACE and 410 received EVD. 114 ACE patients received IT medications. Common diagnoses included IVH, SAH, and ventriculitis. Total complications were lower in the ACE group (Odds Ratio (OR) .29, p<.0001). The ACE group also had fewer occlusions (OR .21, p=.0004), infections (OR .2, p<.0001), shunt dependence (OR .28, p<.0001), and early replacements (OR .4, p<.036). Hemorrhage, CSF leak, and inadvertent removal were not different.

Conclusions: ACE therapy with an IRRAflow catheter appears to be safer than treatment with a standard EVD, while also allowing continuous IT medication treatment. Further studies, especially prospective registries and randomized trials are needed to confirm these results. In the meantime, ACE therapy should be considered in patients requiring ventricular access or CSF diversion.

D030 Innovations in the Implementation of Cerebral Microdialysis at a Level I Trauma Center: an 8-year Experience

Dr. Jefferson Chen ,¹ Dr. Gianna Fote,¹ Dr. Isidora Beach,¹ Ms Carmenita Ong,¹ Dr. Patrick Chen,¹ Mr. Marlon Sy¹

¹University Of California, Irvine, Orange, United States

Introduction: Cerebral microdialysis (CMD) provides hourly monitoring of cerebral metabolism in patients with acute neurotrauma or cerebrovascular (CV) injuries. It is an important component of multimodal brain monitoring; nonetheless, it is labor intensive and requires many institutional resources, thus many are reluctant to adopt it. We describe the 8-year evolution of our CMD program and the experience at our Level 1 trauma and comprehensive stroke center that have made its use seamless.

Methodology: We reviewed current and past policies and procedures for nursing staff and laboratory staff relating to CMD taking note of any changes and improvements. We also performed specific reviews(2021-present) of our recent CMD experience.

Results: We identified key components and methodology that insured the continued success at our institute. These include: 1. Centralization of the CMD analysis in hospital maintained labs 2. Adoption and validation of the pneumatic tube delivery system to transport the hourly microdialysis samples 3. Training of different ICU (ie. NeuroICU, SICU, MICU) nurses to do CMD 4. Direct upload of the results from the CMD analyzer to the electronic medical record(EMR) and 5. Storage of residual samples at -80 C in the central laboratory for possible repeat analysis/research. Analysis of EMR/clinical lab records from 7/2021 to the present demonstrated 31 patients that had CMD with 5017 samples. Specific analysis of 2021–2022 as a representative year revealed 13 patients. 92% were male, median age was 39.3(range 19–82), 70% were TBI and 30% CV related. Total hours with CMD monitoring was 1509 with an average of 116 hours/patient. The average samples/hour/catheter was 0.96. Preliminary analysis of experience from other years suggests similar trends.

Conclusions: Institutional infrastructure is important in making CMD a seamless and reliable component of brain multimodal monitoring. As more centers explore the use of CMD, our experience may help guide its establishment.

D031 Hemodynamic measures in relation to fluid management in traumatic brain injured patients in the intensive care unit

MD Saliha Ergezen ,^1,2 Bsc Şebnem Yıldırım,¹ Dr. ir. David van Klaveren,³ Bsc Laurene Brobbel,¹ Prof. dr. Wilco Peul,⁴ MD Iain Haitsma,² Prof. dr. Diederik Gommers,¹ Dr. Jasper van Bommel,¹ Dr. Eva Klijn,¹ Dr. Mathieu van der Jagt¹

¹Department of Intensive Care Adults, Erasmus MC, Rotterdam, Netherlands, ²Department of Neurosurgery, Erasmus MC, Rotterdam, Netherlands, ³Department of Public Health, Center for Medical Decision Making, Erasmus MC, Rotterdam, Netherlands, ⁴Department of Neurosurgery, LUMC, Leiden, Netherlands

Introduction: Adequate hemodynamic and fluid management is essential in traumatic brain injured (TBI) patients to maintain cerebral perfusion and oxygenation. Higher positive fluid balance during ICU stay is associated with worse outcome in TBI patients (1). However, insights into underlying hemodynamic mechanisms are scarce. We aimed to explore the association of fluid balance and input with cardiac index (CI) and central venous pressure (CVP) in TBI patients in the ICU.

Methodology: We performed a prospective observational study, including patients aged ≥18 years admitted to the ICU with TBI and estimated ICU stay >48 hours. Continuous CO monitoring was performed during the first week after admission with non-invasive (bioreactance) or invasive (pulse contour analysis) techniques. Data on fluid input and balance, intracranial pressure (ICP), CVP and CI were collected and associations analyzed with linear mixed effects models.

Results: We included 38 TBI patients; mean age 49 years (SD 19), 15 female (40%) and median GCS at trauma site 5 (IQR 3 – 7). Mean daily fluid balance was 0.79 L (SD 1.75), CI 3.64 L/min/m2 (SD 0.92), CVP 7.21 mmHg (SD 4.67) and ICP 9.23 mmHg (SD 6.67). Lower CI was observed with higher fluid balance, β = -0.12 (95% CI -0.20; -0.05), and fluid input, β = -0.10 (95% CI -0.20; -0.01). CVP increased non-significantly with higher fluid balance, β = 0.10 (95% CI -0.58; 0.70), and fluid input, β = 0.17 (95% CI -0.39; 0.69). ICP increased non-significantly with fluid balance, β = 0.22 (95% CI -0.11; 0.56), and fluid input, β = 0.21 (95% CI -0.43; 0.83).

Conclusion: More positive fluid balance and input are associated with decreased CI, which could be explained by increased CVP. CVP and ICP tended to increase with higher fluid balance and input, possibly explaining the link with poor outcome.

D032 Event-Related Potentials and Psychosocial Trends in a Population of Individuals with Neuropathic Pain and a Spinal Cord Injury

Mr. Scott Frank ,^1,2 Dr. Roberta Vastano,^1,2,3 Dr. Eva Widerstrom-Noga^1,2,3

¹University of Miami - Miller School of Medicine, Miami, United States, ²The Miami Project to Cure Paralysis, Miami, United States, ³Department of Neurological Surgery - University of Miami, Miami, United States

Nearly 60% of people with a spinal cord injury experience neuropathic pain within their first year of their injury. However, there is a gap in knowledge comparing injury groups within this population. This study investigates body representation correlates from an EEG recording comparing cervical injuries to thoracic injuries. Correlations will be drawn from psychological questionnaires and pain assessments.

EEG recordings were conducted using a 64-electrode system while participants engaged in a laterality judgment task involving body-related (hands and feet) and non-body related stimuli (a duck). This task assessed performance (reaction times and accuracy) and event-related potentials associated with mentally rotating body parts, specifically the activation at the brain level of one's body schema. Psychological assessments included the Beck Depression Index, Cambridge Depersonalization Scale, Multidimensional Pain Inventory, and Psychological General Well Being Index. Pain assessments utilized the Neuropathic Pain Symptom Inventory, Basic Pain Dataset, and Spinal Cord Injury Pain Instrument.

Analysis of task behavior revealed more accurate responses for easier rotation angles, with faster reaction times (RTs) for right stimuli compared to left stimuli and for easier angles compared to harder angles. Body-related stimuli elicited slower RTs than non-body related stimuli, irrespective of angle difficulty. There was a significant main effect of the group between the cervical and thoracic groups in the early processing correlates: frontal-central P100 and parietal-occipital N100. The cervical group demonstrated greater access to the motor cortex, as indicated by the parietal-occipital P200. Positive correlations were observed between depersonalization symptoms, depression, and neuropathic pain intensity, while well-being showed negative correlations with life interference and depression.

The EEG results indicate that body representation disruptions that are topographically organized may be occurring at earlier points than previously thought. Correlations with the Cambridge Depersonalization Scale reveal that depersonalization symptoms positively correlate with important pain outcomes, including neuropathic pain intensity.

D033 Diagnosis of cerebral vasospasm in subarachnoid hemorrhage patients – critical assessment of neurocritical care monitoring strategies

Dr. Iftakher Hossain ,^1,2,3 Dr. Elin Abrahamsson,³ Prof. Niklas Marklund³

¹Deapartment of Neurosurgery, Turku University Hospital and University Of Turku, Turku, Finland, ²Division of Neurosurgery, Addenbrooke's Hospital and University of Cambridge, Cambridge, UK, ³Department of Clinical Sciences Lund, Neurosurgery, Lund University, Department of Neurosurgery, Skåne University Hospital, Lund, Sweden

Introduction: The most feared complication of subarachnoid hemorrhage (SAH) remains development of cerebral vasospasm (CVS) and delayed cerebral ischemia (DCI). The aim of this study was to evaluate the diagnostic tools used clinically to detect CVS, primarily focusing on clinical symptoms, neuroimaging and transcranial doppler ultrasonography (TCD).

Methods: The study was retrospectively conducted by assessing the medical records of all SAH-patients admitted to the neurointensive care unit at Skåne University hospital during a three-year period (2018–2020). 167 patients were eligible for inclusion.

Results: Clinical symptoms of CVS were seen in 57 % of patients. TCD was found to have high specificity but low sensitivity in detection of CVS. No significant difference was found in TCD-velocities depending on grade of arterial narrowing found radiographically. No significant difference was found in TCD-velocities or grade of arterial narrowing in patients expressing clinical symptoms. Computed tomography angiography (CTA) was the most common diagnostic tool used for evaluation and diagnosis and arterial narrowing was found in most CVS patients evaluated.

Conclusion: To enable detection and diagnosis of CVS a high clinical suspicion is needed. TCD remains of value as a diagnostic monitoring tool, but due to its low sensitivity it should be used in combination with other methods such as close monitoring of clinical symptoms, cerebral microdialysis and neuroimaging. A combination of these monitoring tools is needed in order not to miss significant CVS that may lead to the development of infarctions following SAH.

D034 Changes in the Direct Cortical Response during Spreading Depolarisations in the injured human brain

Dr Sharon Jewell ,^1,4 Dr Tomas Watanabe,² Dr Jose-Pedro Lavrador,^1,3 Dr Ahilan Kailaya-Vasan,³ Dr Prajwal Ghimire,³ Dr Ahmed Raslan,³ Dr Sascha Freigang,³ Dr Christos Tolias,^1,3 Dr Eleni Maratos,³ Dr Sinan Barazi,³ Dr Aminul Ahmed,^1,3 Dr Clemens Pahl,³ Professor Martyn Boutelle,⁴ Professor Anthony Strong¹

¹King's College London, London, United Kingdom, ²Vagalume LLC, Palo Alto, USA, ³King's College Hospital, London, United Kingdom, ⁴Imperial College London, London, United Kingdom

Introduction: Spreading depolarisations (SDs) disrupt neuronal processing and are a known mediator of secondary injury. In patients, evaluation is restricted to a subjectively determined event count on large invasive electrodes. Directly translational measures of function are lacking, and quantification of critical SD properties – duration, burden – has proved methodologically elusive. These factors have hindered widespread adoption of monitoring and development of effective interventions. We have developed a reliable index of neuronal functional integrity by measuring changes in the Direct Cortical Response (DCR): a ubiquitous cross-species reaction of the cortex to electrical stimuli, contingent in humans upon placement of only a miniature electrode.

Methodology: Electrocorticographic electrodes were placed in 15 craniotomy patients and used for evoking and recording DCRs and the hallmark features of SD: DC-shift & spreading depression. DCRs were elicited via biphasic charge-balanced pulses of <40µC/cm2/pulse at 0.2Hz. Recordings lasted >6min<5hrs. Features of the DCR were measured semi-automatically and analysed using custom MATLAB scripts.

Results: Principal findings were: 1) DCRs were highly reproducible and could be evoked against a spontaneously isoelectric background; 2) DCR amplitude fell dramatically during SDs (105) and recovery was unexpectedly bi-phasic; 3) Between SDs, DCRs were remarkably stable facilitating objective measurement of SD number, duration and burden (determined by AUC); 4) DCR amplitude fell a median of 2min 51s and up to 10min 38s before the DC-shift.

Conclusions: 1) DCR provides a continuous, reliable index of functional integrity with unparalleled translatability; 2) In humans, evoked as well spontaneous activity is interrupted by SD with complete recovery likely reliant upon divergent metabolic pathways and/or substrates; 3) A loss of neuronal excitability is measurable minutes before catastrophic collapse of ion gradients; 4) DCR is an ideal method to reliably quantify critical SD properties and may pave the way for development and testing of novel therapies.

D035 National U.S. trends in brain tissue oxygen pressure monitoring after severe TBI

Dr Amelia Maiga ,¹ Rebecca Irlmeier,¹ Fei Ye,¹ Dr Areg Grigorian,² Dr Mayur Patel¹

¹Vanderbilt University Medical Center, Nashville, United States, ²University of California Irvine, Irvine, United States

Introduction: In severe traumatic brain injury (TBI) patients with intracranial pressure (ICP) monitors, the OXY-TC multicenter randomized controlled trial from France reported no benefit with brain tissue oxygen pressure monitoring (PbtO2). We hypothesized that PbtO2 monitoring does not improve short-term outcomes in severe TBI patients as compared with standard ICP-guided neuromonitoring. We further hypothesized that PbtO2 usage is driven by hospital rather than patient characteristics.

Methodology: This national retrospective cohort study included adults with ICP monitors placed for severe TBI from the U.S. Trauma Quality Improvement Program (2017–2021). We compared patients with and without PbtO2 monitoring, adjusting for demographics, injury severity, and hospital characteristics (teaching status; number of beds). Logistic regression was used for inpatient mortality and discharge disposition. Linear regression was used for ICU length of stay.

Results: Of 27,336 ICP-monitored patients with a median Glasgow Coma Score 6 [interquartile range 3,7], 1699 (6.2%) had PbtO2 monitoring. The proportion of PbtO2-monitored patients increased from 5.4% in 2017 to 6.9% in 2020, dropping to 6.2% in 2021. PbtO2 patients had comparable demographics and injury severity to non-PbtO2 patients, but were more often at teaching facilities (63% vs. 45% of non-PbtO2 patients) and at medium-sized hospitals (46% vs. 31%) rather than at the largest hospitals (28% vs. 41% of non-PbtO2 patients). PbtO2 monitoring was associated with lower inpatient mortality (odds ratio, OR 0.86, p=0.042), 4 days longer ICU stay (p<0.001), and no difference in discharge disposition (OR 1.0, p=0.997).

Conclusions: In the U.S., PbtO2 monitoring is used in less than 1 in 10 severe TBI patients with ICP monitors, but disproportionately in teaching hospitals and medium-sized hospitals. PbtO2 monitoring may prolong length of stay and offer a survival benefit, without a functional impact at discharge. Ongoing international trials (BOOST3, BONANZA) will determine the utility of PbtO2 monitoring in severe TBI management.

D036 Noninvasive assessment of intracranial pressure: Deformability index as an adjunct to optic nerve sheath diameter to increase diagnostic ability

Dag Ferner Netteland ,^1,2 Mads Aarhus,¹ Else Charlotte Sandset,^3,4 Llewellyn Padayachy,⁵ Eirik Helseth,^1,2 Reidar Brekken⁶

¹Department of Neurosurgery, Oslo University Hospital, Oslo, Norway, ²Faculty of Medicine, University of Oslo, Oslo, Norway, ³Department of Neurology, Oslo University Hospital, Oslo, Norway, ⁴The Norwegian Air Ambulance Foundation, Oslo, Norway, ⁵Department of Neurosurgery, School of Medicine, Faculty of Health Sciences, University of Pretoria, Steve Biko Academic Hospital, Pretoria, South Africa, ⁶Department of Health Research, Medical Technology, SINTEF, Trondheim, Norway

Introduction: Today, invasive intracranial pressure (ICP) measurement is standard, but its demands on resources limits availability. Here, we evaluated a novel ultrasound-based optic nerve sheath parameter called Deformability Index (DI), and its ability to assess ICP non-invasively. Furthermore, we asked whether combining DI with optic nerve sheath diameter (ONSD), a more established parameter, resulted in increased diagnostic ability compared to using ONSD alone.

Methodology: We prospectively included adult traumatic brain injury patients with invasive ICP monitoring, which served as the reference measurement. Ultrasound images and videos of the optic nerve sheath were acquired. ONSD was measured bedside, while DI was calculated by semi-automated postprocessing of ultrasound videos. Correlations of ONSD and DI to ICP were explored and a linear regression model combining ONSD and DI was compared to a linear regression model using ONSD alone. Ability of the non-invasive parameters to distinguish dichotomized ICP was evaluated using Receiver Operating Characteristics curves and a logistic regression model combining ONSD and DI was compared to a logistic regression model using ONSD alone.

Results: Forty-four ultrasound examinations were performed in 26 patients. Both DI (R=-0.28; 95%CI R<-0.03; p=0.03) and ONSD (R=0.45; 95%CI R>0.23; p<0.01) correlated with ICP. When including both parameters in a combined model, the estimated correlation coefficient increased (R=0.51; 95%CI R>0.30; p<0.01) compared to using ONSD alone, but the model improvement did not reach statistical significance (p=0.09). Both DI (AUC=0.69, 95%CI 0.53–0.83) and ONSD (AUC=0.72, 95%CI 0.56–0.86) displayed ability to distinguish ICP dichotomized at ICP≥15mmHg. When using both parameters in a combined model, AUC increased (0.80, 95%CI 0.63–0.90), and the model improvement was statistically significant (p=0.02).

Conclusions: Combining ONSD with DI holds the potential of increasing the ability of optic nerve sheath parameters in the non-invasive assessment of ICP, compared to using ONSD alone, and further study of DI is warranted.

D037 Characterising cerebral microdialysis sampling region and retromicrodialysis delivery in-vitro and in-vivo with gadolinium contrast agents and MRI

Dr Matthew Stovell ,^1,2,3 Mr Chisomo Zimphango,¹ Dr Pascal Ruetten,¹ Dr Daniel Tozer,¹ Dr Marius Mada,¹ Dr Stephen Sawiak,¹ Dr Erik Thelin,^1,4 Dr Adrian Carpenter,¹ Professor Peter Hutchinson,¹ Dr Keri Carpenter¹

¹University of Cambridge, Cambridge, United Kingdom, ²The Walton Centre, Liverpool, United Kingdom, ³Royal Brisbane & Women's Hospital, Brisbane, Australia, ⁴Karolinska Institutet, Stockholm, Sweden

Introduction: Cerebral microdialysis catheters inserted into patients’ cerebral interstitium allow the continuous monitoring of their cerebral metabolism after a severe traumatic brain injury (TBI). Catheters consist of a semi-permeable membrane, across which perfusion fluid can equalize with the surrounding cerebral extracellular environment before being recovered through a central non-porous channel. However, it is unclear how far recovered fluid and suspended metabolites have diffused from within the brain, and therefore what volume or region of brain tissue the analyses of metabolism represent.

Methodology: Using MRI, we characterised the diffusion of gadolinium contrast agents from cerebral microdialysis catheters as a surrogate for other small-molecule study substrates in 4 agarose brain phantoms (5–10mmol/L Gadopentetic acid/Magnevist®) and 6 human TBI patients (10mmol/L Gadobutrol/Gadavist®).

Diffusion distance and pattern was assessed at 1mm isotropic resolution in phantoms at 4.7T using T1 weighted imaging, and in human cerebral white matter at 3T using T1 maps.

Results: Contrast agents diffused from cerebral microdialysis catheters in agarose brain phantoms and patients’ human cerebral white matter in a uniform spheroidal (ellipsoid of revolution) pattern.

Evidence of contrast agent diffusion was found up to 18.5±1.7mm from catheters in phantoms, and 13.4±0.5mm from catheters in human cerebral white matter after 24hours perfusion (mean±s.d.).

In TBI patients, a steep concentration drop off was found so that ≤50% of maximum concentration was achieved at ≈4 mm, and ≤10% of maximum was found beyond ≈7mm from the catheters. There was little variation between subjects. The grey matter-white matter boundary did not obstruct diffusion.

Conclusions: It is now clearer what the size and shape of the region of brain that surrounds a microdialysis catheter its analysis represents. This study also provides a platform for future development of new catheters optimally designed to deliver small molecule therapies to focal pathologies of the human brain.

D038 Brain Tissue Oxygen Monitoring in Traumatic Brain Injury – PbtO2 in Relation to Global Cerebral Physiology and Outcome

Dr Teodor Svedung Wettervik ,¹ Dr Erta Beqiri,² Dr Anders Hånell,¹ Dr Stefan Yu Bögli,² Dr Michal Placek,² Dr Mathew Guilfoyle,² Dr Adel Helmy,² Dr Andrea Lavinio,² Dr Ronan O’Leary,² Professor Peter Hutchinson,² Dr Peter Smielewski²

¹Uppsala University, Uppsala, Sweden, ²University of Cambridge, Cambridge, United Kingdom

Background: The primary aim was to explore the association of pbtO2 vs. global cerebral physiology (ICP, PRx, CPP, and CPPopt) and long-term outcome (GOS).

Methods: 425 patients with ICP and pbtO2 monitoring were included for the analysis of pbtO2 vs global cerebral physiology and a subset of those 239 TBI patients who also had available GOS data were included in the outcome analysis. All patients had been treated at the neurocritical care unit, Addenbrooke’s Hospital, Cambridge, UK, between 2002 and 2022. Outcome was dichotomized into favourable/unfavourable (GOS 4–5/1–3). PbtO2 was studied over the entire monitoring period.

Results: PbtO2 below 20 mmHg occurred in median during 17% of the monitoring time and in less than 5% in combination with ICP > 20 mmHg, PRx > 0.30, CPP < 60 mmHg, or ΔCPPopt < -5 mmHg. In GAM analyses, pbtO2 remained around 25 mmHg over a large range of ICP and PRx, but deteriorated below 20 mmHg for CPP below 30 mmHg and ΔCPPopt below -30 mmHg. In linear mixed effect models, ICP, CPP, PRx, and ΔCPPopt were significantly associated with pbtO2, but the fixed effects could only explain a small extent of the pbtO2 variation.

In outcome analysis, pbtO2 below 15 mmHg was independently associated with unfavourable outcome. In combined heatmaps, it was evident that isolated episodes of low pbtO2 was only weakly associated with outcome, but it was particularly correlated with poor outcome if it occurred together with high ICP, high PRx, low CPP, or negative ΔCPPopt.

Conclusions: PbtO2 below 20 mmHg was frequent and often occurred in the absence of disturbances in ICP, PRx, CPP, and ΔCPPopt. Thus, low pbtO2 is often a complex and independent pathophysiological event. However, low pbtO2 in combination with impairment in global cerebral physiology rather than isolated pbtO2-insults carried most prognostic value.

D039 Validation of the Loke Microdialysis in an Experimental Pig Model: Are We Ready for Continuous Monitoring of Brain Energy Metabolism?

Dr Teodor Svedung Wettervik ¹

¹Uppsala University, Uppsala, Sweden

Objective: Brain energy metabolism is often disturbed after acute brain injuries. Current neuromonitoring methods with microdialysis (MD) are based on intermittent measurements (1–4/hour), but such a low frequency could miss transient, but important events. The solution may be the recently developed Loke MD, which provides high-frequency data of glucose and lactate. Before clinical implementation, the reliability and stability of Loke remains to be determined in vivo. The purpose of this study was to validate Loke in relation to the standard intermittent MD method.

Methods: Four pigs aged 2–3 months were included. They received two adjacent cerebral MD catheters, one for standard intermittent and one for continuous (Loke) assessments of glucose and lactate. The standard MD was measured every 15 minutes. The continuous Loke MD was sampled every 2–3 second and was averaged over corresponding 15-minute intervals for the statistical comparisons with standard MD. Intravenous glucose injections and intracranial hypertension by inflation of an intracranial epidural balloon were performed to induce variations in intracranial pressure (ICP), cerebral perfusion pressure (CPP), and systemic and cerebral glucose and lactate.

Results: In a linear mixed effect model of standard MD glucose there was a fixed effect value (±standard error [SE]) at 0.94 ± 0.07 (p < 0.001) for Loke MD glucose with an intercept at -0.19 ± 0.15 (p=0.20). The model showed a conditional R2 at 0.81 and marginal R2 at 0.72. In a linear mixed effect model of standard MD lactate, the fixed effect value (±SE) at 0.41 ± 0.16 (p = 0.01) for Loke MD lactate with an intercept at 0.33 ± 0.21 (p=0.25). The model showed a conditional R2 at 0.47 and marginal R2 at 0.17.

Conclusions: The same established MD glucose thresholds for Loke can be used as for standard MD, while this should be done with caution for lactate.

D040 relationship between autonomic nervous system activity and ICP pulse morphology in traumatic brain injury patients

Dr Agnieszka Uryga ,¹ PhD Agnieszka Kazimierska,¹ Msc Eng Cyprian Mataczyński,¹ Prof Marek Czosnyka,² Prof Magdalena Kasprowicz,¹ CENTER-TBI High-Resolution ICU (HR ICU) Sub-Study Participants and Investigators

¹Wroclaw University Of Science And Technology, Wrocław, Poland, ²Division of Neurosurgery, Department of Clinical Neurosciences, Addenbrooke’s Hospital, University of Cambridge, Cambridge, UK

Introduction: The morphology of intracranial pressure (ICP) pulse waveform is determined by pulsatile cerebral arterial blood inflow, cerebral venous outflow, and the state of pressure-volume compensation. As ICP rises, pressure-volume reserve decreases. Furthermore, the activity of the autonomic nervous system, which regulates arterial inflow, may also be affected by increasing ICP. To understand the mechanisms underlying the relationship between the ICP pulse waveform and autonomic activity, we investigated periods of ICP elevation in traumatic brain injury (TBI) patients.

Methodology: ICP and arterial blood pressure recordings from 186 TBI patients from the CENTER-TBI high-resolution sub-study were analysed retrospectively. Morphology of ICP pulses was assessed by AI-based metric known as the pulse shape index (PSI). A neural network model was employed to automatically categorize ICP pulses into four classes from 1 (normal pulse waveform) to 4 (pathologically altered waveform), with PSI calculated as the weighted sum of class numbers. PSI was compared with autonomic indices: heart rate (HR) and baroreflex sensitivity (BRS) during periods of ICP below and above 22 mm Hg (lasting at least 5 min) within the first seven days after injury.

Results: During episodes of elevated ICP observed in 151 patients, PSI (2.5±1.2) and HR (80±23 bpm) were significantly higher compared to periods of ICP below 22 mm Hg (2.2± 1.2 and 75±19 bpm, respectively; p<0.001 for both). No differences were found in BRS for low ICP. However, when ICP was elevated above 22 mm Hg, PSI significantly correlated with BRS (rS=0.29; p<0.001) and HR (rS=-0.39; p<0.001).

Conclusions: Our findings suggest a potential influence of elevated intracranial pressure (ICP) on autonomic activity. Increased baroreceptor activity during episodes of ICP elevation may further alter the shape of the ICP pulse. This study was supported by the National Science Center (UMO-2019/35/B/ST7/00500, UMO-2022/47/D/ST7/00229).

D041 APOE4 and age influence entorhinal cortex thickness and volume and blood arachidonic acid and docosahexaenoic acid lipid levels after mTBI

Dr. Laila Abdullah ¹

¹Roskamp Institute, SARASOTA, United States, ²James A. Haley VA Hospital, Tampa, USA

Introduction: Alterations in blood arachidonic acid (AA) and docosahexaenoic acid (DHA) levels are associated with mild Traumatic brain injury (mTBI) and Alzheimer’s disease (AD) among apolipoprotein E (APOE) ε4 carriers. A loss of the brain entorhinal cortex (EC) thickness and volume is associated with these lipids and AD risk for ε4 carriers. We, therefore, examined the influence of APOE ε4 on the relationship between AA- and DHA-containing lipids and the EC thickness and volume after mTBI among Veterans to determine their role as potential biomarkers of chronic consequences of TBI.

Methods: 3D T1-Weighted Magnetic Resonance Imaging (MRI) measurements of EC thickness and volume from the Long-Term Impact of Military-Relevant Brain Injury Consortium/Chronic Effects of Neurotrauma Consortium (LIMBIC-CENC) cohort (n = 631) were examined. Plasma (n =181) were subjected to lipidomic analyses for identifying and quantifying AA- and DHA-containing phosphatidylcholine (PC), cholesterol esters (CE), phosphatidylethanolamine (PE), diglycerides (DG) and ethanolamides (EA).

Results: Among ε4 carriers 2+ repetitive mTBI (r-mTBI) was associated with lowered thickness in the right EC. Among ε4 carriers with mTBI, increasing age and mTBI chronicity together were associated with increased White Matter Volume (WMV) within the right and left EC. Increasing age and number of blast mTBI together among ε4 carriers were associated with lowered right EC Grey Matter Volume (GMV) and thickness. The presence of ε4 together with chronicity of mTBI and r-mTBI and altered AA to DHA ratios were associated with low GMV and WMV in the left EC.

Conclusion: Further investigation of the brain EC in relation to mTBI and these biomarkers could help with developing an easily accessible blood-based prognostic tool for providing care and clinical management of mTBI patients.

D042 Serial Measurements of Blood Based Biomarkers in Pediatric Traumatic Brain Injury: Utility of Temporal Profiling to Predict Long-Term Outcome

Laura Blackwell ,^1,2 Makda Mulugeta,² Guangzheng Cai,³ Daniel Rawad Arja,³ Firas Kobaissy,³ Kevin Wang,³ Dr. Andrew Reisner^1,2

¹Emory University School Of Medicine, Atlanta, United States, ²Children's Healthcare of Atlanta, Atlanta, United States, ³Center for Neurotrauma, MultiOmics & Biomarkers, Department of Neurobiology, Morehouse School of Medcine, Atlanta, United States

Introduction: Heterogeneity in outcomes in traumatic brain injury (TBI) presents challenges for clinical care and management, particularly in children. Serial testing and temporal profiling of blood-based biomarkers may provide additional information to support management and prognosis.

Objectives: Examine the time course of blood biomarkers within 72 hours of TBI and correlate with functional outcomes.

Methods: Prospective study including 85 children with TBI presenting to a Level I trauma center following a TBI with Glasgow Coma Scale (GCS) scores of 3 to 15. The primary outcome measure was Glasgow Outcome Scale Extended, pediatric version (GOSE-peds), categorized into poor (GOSE</=4) and good (GOSE>/=5) outcomes. Blood samples were obtained on all patients within six hours of injury and serially for 3 days and measured with Quanterix Simoa platform for GFAP, UCH-L1, and ELISA platform for S100B. Area under the receiver operating characteristic curve (AUC) with 95% confidence intervals was used to predict poor outcomes.

Results: Patients average age was 8.56 ± 5.3 years, 64.7% male, median lowest GCS=8.5. Temporal differences were found across biomarkers, such that second time point (24hr) was the highest and best prdictor of outcomes. GFAP AUC for admission= .646 (95%CI = .414 to .878); 24hr AUC= .946 (95%CI = .871 to 1.0); 48hr AUC= .911 (95%CI = .809 to 1.0); 72hr AUC=.869 (95%CI = .724 to 1.0). S100B AUC for admission= .508 (95%CI = .285 to .731); 24hr AUC= .883 (95%CI = .751 to 1.0); 48hr AUC= .850 (95%CI = .707 to .993); 72hr AUC=.835 (95%CI = .667 to 1.0). No significant findings were observed with UCH-L1.

Conclusions: The optimal model for assessing outcomes and long-term planning is likely to involve multiple components, including several blood biomarkers, along with clinical and radiographic findings, at different time points. Additional investigation of the trajectory of known biomarkers in children is warranted.

D043 A high-dimensional proteomic approach identifies acute TBI-specific derangements in plasma expression of neurodegenerative, inflammatory and cell metabolism proteins that relate to injury patterns

Dr Lucia Li ,^1,2 Dr Eleftheria Kodosaki,^3,4 Dr Amanda Heselgrave,^3,4 Prof Henrik Zetterberg,^4,5,6 Dr Neil Graham,^1,2 Dr Karl Zimmerman,^1,2 Dr Eyal Soreq,^1,2 Dr Thomas Parker,^1,2 Dr Elena Garbero,⁷ Dr Federico Moro,⁷ Dr Sandra Magnoni,⁸ Dr Guido Bertolini,⁷ Prof David Loane,⁹ Prof David Sharp^1,2

¹Imperial College London, London, United Kingdom, ²UKDRI Centre for Care Research & Technology, London, United Kingdom, ³UCL UKDRI, London, United Kingdom, ⁴UCL Institute of Neurology, London, United Kingdom, ⁵University of Gothenberg, Gothenberg, Sweden, ⁶Sahlgrenska University Hospital, Mölndal, Sweden, ⁷Mario Negri Institute, Ranica, Italy, ⁸Santa Chiara Hospital, Trento, Italy, ⁹Trinity College Dublin, Dublin, Ireland

Background: Classification of TBI that reflects TBI pathophysiology is needed to improve prognostication and treatment development. Blood biomarkers are a promising way to identify specific TBI pathophysiology. We used a multiplex proteomic assay (NULISA™) to investigate plasma protein expression following acute moderate-severe TBI.

Methods: We used differential expression analysis to compare protein expression in samples taken within 10 days of TBI from 88 participants from the BIO-AX-TBI cohort (n=38 TBI, n=22 non-TBI trauma (NTT), n=28 non-injured controls (CON)) on the novel high-dimensional (120 proteins) Alamar NULISA™ CNS Diseases panel, the OLINK® Target 96 Inflammation panel, Simoa® (neurofilament light, GFAP, total tau and UCHL1) and Millipore (S100B). A subset (n=29 TBI, n=24 CON) also had subacute 3T MRI injury measures (fractional anisotropy [FA], lesion volumes), scanned 10 days to 6 weeks after injury.

Results: 16 proteins with TBI-specific significantly different plasma expression were identified. These included neuronal markers (calbindin2, UCHL1, visinin-like protein1), astroglial markers (S100B, GFAP), neurodegenerative disease proteins (total tau, pTau231, PSEN1, amyloid beta42, 14–3-3g), cytokines (IL16, CCL2, ficolin2), cell metabolism (MDH1) and autophagy (sequestome1) proteins. Acute plasma levels of UCHL1, PSEN1, total tau and pTau231 correlated with subacute lesion volume, while sequestome1 correlated with subacute whole white matter skeleton FA and CCL2 inversely correlated with corpus callosum FA. Cluster analysis of acute plasma protein expression identified 3 TBI subgroups which had different patterns of injury but did not differ in age or outcome. Most proteins that overlapped on two platforms had excellent (r >0.8) correlations between values.

Conclusions: We found TBI-specific changes in acute plasma levels of neuronal, astroglial and neurodegenerative proteins, cytokines and cell metabolism proteins. These changes were related to later patterns of injury. Our study identifies amyloid processing, inflammatory and cell metabolism proteins previously not known to be impacted by TBI.

D044 GFAP and UCH-L1 biomarker levels in suspected TBI patients can distinguish acute intracranial lesions from chronic CT findings

Dr Ksenia Musaelyan ,¹ Dr Raj Chandran,² Dr Jaime Marino,² Paula Bernander,² Dr Gangamani Beligere,² Dr Saul A. Datwyler,² Prof Peter Biberthaler,³ Dr Beth McQuiston²

¹Abbott Laboratories, Dublin 22, Ireland, ²Abbott Diagnostics, Lake County, United States, ³Technical University of Munich, Munich, Germany

Introduction: Glial fibrillary acidic protein (GFAP) and ubiquitin C-terminal hydrolase L1 (UCH-L1) are being implemented in clinical care of traumatic brain injury (TBI) patients for rule out of acute intracranial lesions visible on a computed tomography (CT) scan. This study aimed to elucidate whether imaging findings indicative of brain injury having occurred prior to the suspected acute TBI being assessed (chronic CT findings) would affect clinical performance of GFAP and UCH-L1 for CT rule out of acute TBI.

Methods: GFAP and UCH-L1 levels were analysed in a cohort of 109 patients with suspected acute TBI. Blood draws were done within 6 hours from injury. ARCHITECT GFAP and UCH-L1 prototype assays were used. All patients having undergone a head CT scan, and based on results, were assigned to CT-negative (no intracranial lesions, n=43), CT-positive (TBI-related acute intracranial lesions, n=24) and chronic CT findings (n=42) groups.

Results: GFAP levels were significantly higher in the CT-positive group compared to CT-negative and chronic CT findings groups. GFAP levels could distinguish patients with CT-positive lesions from patients with chronic CT findings with an AUC of 0.80. Similarly, in samples collected within 2 hours from injury UCH-L1 was significantly higher in the CT-positive group compared to the chronic CT findings group, and distinguished CT-positive patients from patients in the chronic CT findings groups with an AUC of 0.75.

Conclusions: This data suggests that GFAP and UCH-L1 can distinguish between acute intracranial lesions and chronic CT findings. Future studies providing medical histories of patients with the chronic CT findings and acute TBI will be able to affirm GFAP and UCH-L1 performance.

The study was funded by Abbott Laboratories and in collaboration with the US Army Research and Materiel Command under Collaborative Research and Development Agreement No. 20-1266-CRA

D045 Evaluating Traumatic Brain Injury Classification through Targeted Biomarker Integration

Peter Ngum ,^1,2,3 Harri Merisaari,^4,5 Jussi Posti,^1,6 Olli Tenovuo^1,2

¹Turku Brain Injury Center, Turku University Hospital, Turku, Finland, ²Department of Clinical Neurosciences, University of Turku, Turku, Finland, ³Johns Hopkins University, Carey Business School, Baltimore, USA, ⁴Turku Brain and Mind Center, University of Turku, Turku, Finland, ⁵Department of Diagnostic Radiology, University of Turku, Turku, Finland, ⁶Neurocenter, Department of Neurosurgery, Turku University Hospital, Turku, Finland

Introduction: Traumatic Brain Injury (TBI) represents a complex medical challenge with highly variable outcomes. Advances in biomarker research offer new avenues to enhance the classification and treatment of TBI. This study assesses the integration of FDA-approved biomarkers Glial Fibrillary Acidic Protein (GFAP) and Ubiquitin Carboxy-terminal Hydrolase L1 (UCHL1) with basic clinical assessments, exploring their impact on the clustering of TBI patients.

Methodology: Data from 181 TBI patients of all severities in the EU-funded Finnish TBIcare dataset were analyzed, integrating clinical assessments (age, motor GCS, pupil reactivity) with GFAP and UCHL1 acute biomarker levels. Hierarchical clustering grouped data based on similarities, with cluster validity assessed via the Cophenetic Correlation Coefficient (CCC). This metric ranges from -1 to 1, where values closer to 1 indicate better preservation of original data distances. Preprocessing included imputation for missing values, outlier handling, and normalization.

Results: In this study (69.61% male, median age 51), the majority (70.72%) had GCS scores of 13–15, 19.34% had GCS 3–8, and 10.94% were GCS 9–12. Hierarchical clustering with basic clinical assessments achieved a Cophenetic Correlation Coefficient (CCC) of 0.89, indicating robust data structure preservation. Integrating GFAP slightly reduced this clarity (CCC = 0.80). Adding UCHL1, highly correlated with GFAP at r = 0.972513, further lowered the CCC to 0.69. The substantial overlap in biomarker information suggests that their concurrent utility in clustering may be limited by redundancy, complicating delineation of clinically meaningful patient groups.

Conclusions: While GFAP and UCHL1 offer critical insights and could substantially reduce unnecessary CT scans by accurately ruling out intracranial injuries, their combined use in clustering analyses provides limited additional benefit due to their high correlation. This redundancy can mask clinically meaningful distinctions among patient groups. Future research should refine the TBI biomarker panel and consider sophisticated clustering algorithms that can effectively manage highly correlated data.

D046 Initial experience with Abbott Alinity-i TBI blood test as an aid-in-diagnosis of traumatic brain injury in emergently evacuated combat casualties

Dr. Avital Perry,^1,2 Mrs. Yael Rosen Lang ,¹ Anton Peled,² Alisa Sheldon,¹ Keren Asraf,³ Yoram Klein,⁴ Avi Epstein,⁵ Avinoah Irony,⁵ Geoffrey T. Manley,⁶ Raquel C. Gardner¹

¹Joseph Sagol Neuroscience Center, Ramat Gan, Israel, ²Department of Neurosurgery, Sheba Medical Center, Israel, Israel, ³Clinical Laboratory Division, Sheba Medical Center, Israel, ⁴Department of Trauma Surgery, Sheba Medical Center, Israel, ⁵Department of Emergency Medicine, Sheba Medical Center, Israel, ⁶Department of Neurosurgery, University of California San Francisco, USA

Introduction: Blood-based diagnosis of traumatic brain injury (TBI) has potential to streamline combat casualty care. We aimed to investigate the Abbott Alinity-i TBI blood test as an aid-in-diagnosis of TBI in combat casualties emergently evacuated to our trauma center.

Methodology: In this ongoing registry study, we leveraged data collected under an IRB-approved laboratory validation protocol of the Abbott Alinity-i TBI blood test, a semi-quantitative test of serum UCHL1 (threshold 400pg/mL) and GFAP (threshold 35pg/mL) that is reported as positive if either marker is above threshold. Inclusion criteria were: emergently evacuated combat casualty, received TBI blood test 24h post-injury, had documentation of TBI symptom screening. TBI was phenotyped across 4 dimensions: 1.blood (positive/negative TBI test), 2.CT-hemorrhage (positive/negative intracranial hemorrhage), 3.CT-fracture (positive/negative facial/skull/cervical fracture), 4.clinical (traumatically-induced loss/alteration of consciousness or amnesia).

Results: N=29 combat casualties met inclusion criteria from 1/1/2024–4/3/24. Among N=25 with exact time of injury recorded, time to blood collection was 6h for 23/25 and 3h for 15/25. 29/29 were male, median age was 24y (range 19–37y), and 26/29 sustained blast injury. Arrival GCS was 15 in 25/29.

Among the 13/29 blood-negatives, all were CT-hemorrhage/fracture-negative (N=7) or CT not indicated (N=6), and N=2 were clinical-positive only.

Among the 16/29 blood-positives, N=4 were negative across all other dimensions, while N=12 were positive in other dimension: CT-hemorrhage only (N=1), CT-fracture only (N=2), clinical only (N=4), clinical plus CT-hemorrhage/fracture (N=5).

Conclusions: Preliminary experience with this TBI blood test in 29 emergently evacuated combat casualties has found 100% sensitivity (and 100% negative predictive value among those undergoing CT) for acute intracranial hemorrhage and acute facial/skull/cervical fracture. Additional research is needed in larger cohorts to establish added value of this test to aid-in-diagnosis of CT-negative TBI, and particularly for discrimination of TBI symptoms from acute stress reaction in the context of combat casualties.

D047 Novel antibodies to neurofilament NF-L, NF-M and NF-H allow the visualization of primary, secondary and Wallerian axonal degeneration and the study of neurofilament degradation

Professor/ceo Gerry Shaw ,¹ Ms. Irina Madorsky,¹ Ms. Marda Jorgensen,¹ Ms. Anna Fusco,² Dr. Ying Li,¹ Dr. YongSheng Wang,¹ Dr. Sabhya Rana,² Professor David Fuller²

¹University of Florida/EnCor Biotechnology Inc., Gainesville, United States, ²University of Florida, Gainesville, United States

Introduction: Antibody based detection of neurofilament NF-L in blood and CSF has become the most widely used and useful blood and CSF biomarker to monitor axonal injury and degeneration. We wished to characterize the antibodies utilized in current NF-L assays as a first step in understanding the exact form of NF-L being detected.

Methodology: The epitopes for key NF-L antibodies were mapped using appropriate recombinant and peptide constructs and their utility as reagents for immunocytochemistry on control and various damage and disease states were assessed. We then raised a panel of antibodies to the epitopic and flanking regions of NF-L and homologous regions of NF-M and NF-H. We then characterized all neurofilament antibodies on calpain derived fragments of the three proteins identifying proteolytically resistant fragments of each.

Results: Key NF-L assay antibodies bind to a proteolytically resistant region in the center of the “Coil II” region of NF-L. All novel antibodies to this and flanking regions of NF-L and also of NF-M and NF-H bind epitopes hidden in healthy axons but released proteolytically on degeneration. These antibodies spdifically revealed degenerating and degenerated axonally derived profiles in models of spinal cord injury, traumatic brain injury, stroke and appropriate transgenic mice. We also characterized NF-L, NF-M and NF-H calpain fragments which likely represent the forms of these molecules released into blood and CSF.

Conclusions: Antibodies useful for detecting neurofilament proteins in blood and CSF bind epitopes hidden in normal neurons but exposed on degeneration. These antibodies can be used not only to measure the levels of the relevant proteins in blood and CSF but also to visualize primary, secondary and Wallerian axonal degeneration. Finally we have identified calpain derived fragments of NF-L, NF-M and NF-H which are likely to be useful targets for further biomarker assay development.

D048 Synaptic neurotransmission protein genotypes are associated with post-traumatic epilepsy and long-term outcomes after severe TBI

Dr. Sarah Svirsky ,¹ Mr. Matthew Dain,¹ Dr. Yuefang Chang,¹ Dr. James Castellano,² Dr. David Okonkwo,¹ Dr. Yvette Conley,³ Dr. Shaun Carlson,¹ Dr. Ava Puccio¹

¹Dept of Neurological Surgery, University Of Pittsburgh, Pittsburgh, United States, ²Dept of Neurology, University Of Pittsburgh, Pittsburgh, United States, ³School of Nursing, University Of Pittsburgh, Pittsburgh, United States

Introduction: Post-traumatic epilepsy (PTE) occurs in up to one third of severe TBI patients (sTBI). PTE is independently associated with impoverished long-term neurolgic outcomes. Epileptogenesis, the process by which brain tissue becomes prone to seizures, is associated with aberrant synaptic regulation. Alterations in synaptic neurotransmission have been specifically linked to the development of PTE, however underlying genetic etiologies remains relatively unknown. We hypothesize that single nucleotide polymorphisms (SNPs) of synaptic genes are associated with PTE as well as poor functional outcomes in patients with sTBI.

Methods: Participants were prospectively enrolled under an approved IRB, between 2002–2013, with corresponding PTE status extracted from the electronic medical record. SNPs from AP2M1, CLTA, CLTC, and SYT1 were genotyped using the Human Core Exome (Illumina). Outcome up to 24 months post-injury was measured using Glasgow Outcome Scale (GOS) and Disability Rating Scale (DRS).

Results: 206 participants had DNA extracted and PTE analyzed (Mean Age: 38.57±16.83, 80.63% male, Median GCS: 6). 28% (n=57) of patients had confirmed PTE. 38 SNPs across 4 genes were identified. Multivariate logistic regression, controlling for sex age, and initial severity found AP2M1 rs8478 and rs2231224 minor allele variants significantly increased the odds of PTE (p<0.05, OR=2.21 and 2.19, respectively). CLTA rs4879960 major allele variant significantly reduced odds of poor outcome on 6–24mo DRS (p<0.05, OR<0.6). 6 Syt1 minor allele variants significantly reduced odds of poor outcome on 12mo GOS (rs1405499, rs1918193, rs1918191, rs1245804, rs1245824 and rs2272500, p<0.05, OR<0.5).

Conclusions: We found an association between SNPs in three synaptic neurotransmission genes (AP2M1, CLTA and Syt1) and PTE or long-term outcomes in a sTBI population. Identification of these genetic variants may improve early identification of patients at high-risk for PTE. Further work is needed to determine the implication of AP2M1, CLTA and Syt1 polymorphisms on protein function and pathological mechanisms.

D049 Vascular Endothelial Growth Factor-A (VEGF-A) levels are associated with increased severity and unfavorable outcomes in traumatic brain injury (TBI): A TRACK-TBI Study

Dr. Rachel Thomas ,¹ Ms. Catherine Demos,² Dr. Nikil Padmanabhan,² Dr. Teron Gorham,² Dr. George Sigal,² Dr. Jacob Wohlstadter,² Dr. Sonia Jain,³ Dr. Xiaoying Sun,³ Dr. Joseph Giacino,⁴ Dr. Michael McCrea,⁵ Dr. David Okonkwo,⁶ Dr. Claudia Robertson,⁷ Dr. Nancy Temkin,⁸ Dr. Pratik Mukherjee,⁹ Dr. Kevin Wang,¹⁰ Dr. Ava Puccio,⁶ Dr. Andrea Schneider,¹ Dr. Danielle Sandsmark,¹ Dr. Geoff Manley,⁹ Dr. Ramon Diaz-Arrastia¹

¹University of Pennsylvania, Philadelphia, United States, ²MesoScale Diagnostics, Gaitherburg, United States, ³University of California, San Diego, San Diego, United States, ⁴Spaulding Rehabilitation Hospital/Harvard Medical School, Boston, United States, ⁵Medical College of Wisconsin, Milwaukee, United States, ⁶University of Pittsburgh, Pittsburgh, United States, ⁷Baylor College of Medicine, Houston, United States, ⁸University of Washington, Seattle, United States, ⁹University of California, San Francisco, San Francisco, United States, ¹⁰Morehouse University, Atlanta, United States

Introduction: Plasma VEGF-A is associated with post-traumatic inflammation and vascular permeability. Here we describe the relationship of VEGF-A with TBI severity and 6-month outcomes.

Methodology: A subset of TRACK-TBI participants with TBI (n=367), orthopedic injuries (OI, n=86), and healthy controls (HC, n=69) was analyzed. Plasma was collected at 1 day (D1), 2 weeks (W2), and 6 months (M6) after injury, and VEGF-A levels were measured using Meso Scale Diagnostics V-PLEX assays. Group comparisons used the Wilcoxon Rank Sum test. Outcome was assessed at M6 using the Glasgow Outcome Scale-Extended (GOSE) and dichotomized as good recovery (GOSE >6) and unfavorable outcome (GOSE<5).

Results: The mean (SD) age of the TBI participants was 39.6 (16.1) years, 70.3% were male, and 48.8% had GCS 3–12. D1 VEGF-A levels (median [25th-75th percentiles], pg/mL) was higher in TBI 169.5 [110.2–267.8], compared to 133.6 [89.2–188.2] in HC and 136.5 [97.4–206.4] in OI (p<0.007 for both comparisons). On D1, levels were higher for GCS 3–12 (183.4 [128.3–269.1]) than in GCS 13–15 (158.3 [100.8–258.1], p=0.039). At W2, VEGF-A levels normalized for GCS 13–15 (135.4 [81.9–230.3] but increased for GCS 3–12 (231.5 [120.2–511.6], p<0.0001). VEGF-A levels were also higher in those with intracranial pathology on cranial computed tomography (p=0.0048 at D1, and p <0.0001 at W2). VEGF-A levels returned to control values by M6. D1 VEGF-A levels did not differ by M6 GOSE, but W2 levels were higher in those with disability (GOSE 1–6: 196.9 [113.5–400.2]; GOSE 7–8: 154.1 [89.2–284.6], p=0.04) and those with unfavorable outcome (GOSE 1–4: 298.1 [158.8–551.7]; GOSE 5–8: 151.9 [83.75–283.5], p<0.0001).

Conclusion: VEGF-A is increased after TBI, and in severely injured cases higher levels increase over 2 weeks. VEGF-A is a biomarker of traumatic microvascular injury and shows promise as a potential prognostic biomarker of a possibly treatable endophenotype.

D050 Severe Head Injury Brain Analysis: Characterizing TBI through analysis of fresh brain tissue in the acute phase

Professor Chris Uff ,¹ Miss Shumaila Hasan,¹ Miss Melisa Aria Cetin,² Miss Chloe Kam,² Professor Mark Pepys,³ Dr Ping Yip²

¹Royal London Hospital, Barts Health NHS Trust, London, United Kingdom, ²Centre for Neuroscience, Surgery & Trauma, Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, LONDON, United Kingdom, ³Wolfson Drug Discovery Unit, University College London, London, United Kingdom

Traumatic brain injury (TBI) is a heterogeneous and complex disease and current classification systems which grade TBI as mild, moderate, and severe fail to capture its complexity. Neuroimaging cannot resolve cellular and molecular changes due to lack of resolution, and post-mortem tissue examination may not represent acute disease.

With the aim of mirroring advances in neuro-oncology where tissue examination is the standard, brain tissue was sampled opportunistically if dura was breeched for therapeutic or diagnostic purposes, and blood, urine and CSF were collected for seven days.

Analysis of brain tissue was performed with various state-of-the-art techniques revealing novel insights into the pathobiology of severe TBI in the acute setting.

Immunohistochemistry allowed the injury to be graded according to various domains including neuronal damage (NeuN), dendritic damage (MAP2), vascular injury (VWf / Claudin5) and neuroinflammation (IBA1 / P2Y12) and a significant negative correlation was demonstrated between increasing injury grade and outcome measured by GOS-E (ref1).

The neuro-inflammatory biomarkers TMEM119 and Iba1 were detected in urine, and levels were correlated to outcome.

Significantly reduced staining for synaptophysin and parvalbumin in GABA-ergic interneurons was demonstrated in patients who died compared to survivors regardless of GOS-E grade.

Multiplex immunohistochemistry allowed simultaneous immunostaining of 12 immunostaining markers on a single tissue section including markers for neurones (NeuN, N52, SMI31, SMI32, MAP2, somatostatin), oligodendrocytes (CNPase), astrocytes (GFAP), microglia (Iba1, P2Y12), and vasculature (claudin5, vWF).

The normal plasma protein serum amyloid-P component was demonstrated in the acute phase of TBI, suggesting a mechanistic linkage between TBI and chronic traumatic encephalopathy and Alzhermer's disease (ref2).

Preservation of the cytoarchitecture permitted spatial transcriptomic and spatial proteomic evaluation, and immune electron microscopy.

These techniques demonstrate that tissue sampling in TBI is safe, and that significant information across multiple domains may be derived from analysis of fresh cerebral tissue in TBI

D051 Cervical spine immobilisation following blunt trauma in pre-hospital and emergency care: a systematic review

Prof Fiona Lecky ,^1,2 Mr Abdullah Pandor,¹ Ms Munira Essat,¹ Ms Anthea Sutton,¹ Dr Gordon Fuller,¹ Dr Stuart Reid,³ Prof Jason Smith,⁵ Ms Rachel Fothergill,⁴ Dr Dhushy Surendra Kumar,⁶ Mr Angelos Kolias,⁷ Prof Peter Hutchinson,⁷ Prof Gavin Perkins,⁸ Professor Mark Wilson⁹

¹University Of Sheffield, Sheffield, United Kingdom, ²Emergency Department, Salford Royal Hospital, NCA, Salford, United Kingdom, ³Emergency Department, Sheffield Teaching Hospitals NHS Foundation Trust, United Kingdom, ⁴London Ambulance Service NHS Foundation Trust, London, United Kingdom, ⁵University Hospitals Plymouth NHS Foundation Trust, Plymouth, UK, ⁶University Hospital of Coventry and Warwickshire NHS Foundation Trust, Coventry, UK, ⁷University of Cambridge Department of Clinical Neurosciences, Cambridge, United Kingdom, ⁸University of Warwick, Warwick, United Kingdom, ⁹Imperial College Biomedical Research Centre, London, United Kingdom

Objectives: To assess whether different cervical spine immobilisation strategies (full immobilisation, movement minimisation or no immobilisation), impact neurological and/or other outcomes for patients with suspected cervical spinal injury in the pre-hospital and emergency department setting.

Design: Systematic review following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines.

Data Sources: MEDLINE, EMBASE, CINAHL, Cochrane Library and two research registers were searched until September 2023.

Eligibility Criteria: All comparative studies (prospective or retrospective) that examined the potential benefits and/or harms of immobilisation practices during pre-hospital and emergency care of patients with a potential cervical spine injury (pre-imaging) following blunt trauma.

Data Extraction and Synthesis: Two authors independently selected and extracted data. Risk of bias was appraised using the Cochrane ROBINS-I tool for non-randomised studies. Data were synthesised without meta-analysis.

Results: Six observational studies met the inclusion criteria. The methodological quality was variable, with most studies having serious or critical risk of bias. The effect of cervical spine immobilisation practices such as full immobilisation or movement minimisation during pre-hospital and emergency care did not show clear evidence of benefit for the prevention of neurological deterioration, spinal injuries and death compared with no immobilisation. However, increased pain, discomfort and anatomical complications were associated with collar application during immobilisation.

Conclusions: Despite the limited evidence, weak designs and limited generalisability, the available data suggest that pre-hospital cervical spine immobilisation (full immobilisation or movement minimisation) was of uncertain value due to the lack of demonstrable benefit and may lead to potential complications and adverse outcomes. High-quality randomised comparative studies are required to address this important question.

D052 Randomised Evaluation of Early v Late cranioplasty: A Pilot Study

Dr Harry Mee ,^1,2 Mr A Kolias,^1,2 Mr A Helmy,^1,2 Mr I Timofeev,² Mr E Viaroli,^1,2 Dr F Anwar,² Ms G Whiting,¹ Professor P Hutchinson^1,2

¹University Of Cambridge, Cambridge, United Kingdom, ²Cambridge University Hospital NHS Foundation Trust, Cambridge, United Kingdom

Introduction: Patients with significant brain swelling and/or raised intracranial pressure after TBI or stroke can undergo a decompressive craniectomy (DC) to help in management. Those who survive often require a second operation, known as cranioplasty (CP). This aims to restore the integrity of the skull, giving a degree of mechanical protection with the timing of the procedure possibly impacting neurological outcomes.

Methodology: A single-centre, pilot randomised study comparing two distinct time intervals for cranioplasty: early (within 3 months after DC) vs standard of care (more than 6 months after DC) collecting functional outcomes at 6- and 12-months following DC. Adult patients (>16) who have undergone a DC for either TBI or stroke and who were clinically assessed for cranioplasty. Functional outcome was measured using GOSE (dichotomised: unfavourable (1–3) or favourable (4–8)), and functional independence was measured using functional independence measure (FIM) with FIM efficiency (change per day).

Results: 14 patients randomised. At 6 months, the number of patients with a GOSE favourable outcome in the early group was higher compared to the late group (5 compared to 4). Improvements in functional independence following CP were seen in both groups, with mean FIM total change in the early group (27.7 (SD 26.94)) compared with the late group (33.92 (SD 33.54)); however, FIM efficiency was greatest in the early group (0.21) compared to late (0.13).

Conclusions: Although no definitive findings can be drawn, there is evidence of an increase in functional independence following cranioplasty independent of timing. Although FIM efficiency was greatest in the early group, a further substantial study is required to evaluate further before any definitive conclusions can be drawn.

D053 A Comparison of Brain Oxygen Optimization Clinical Trial Design and Enrollment: BOOST-3 and BONANZA

Ava Puccio ,¹ Lori Shutter,¹ Andrew Udy,² Jamie Cooper,² Carol Moore,³ Camila Battistuzzo,² Toby Jeffcote,² William Barsan,⁴ Sharon Yeatts,⁵ Ramon Diaz-Arrastia³

¹University of Pittsburgh, Pittsburgh, United States, ²Alfred Health, Monash University, Melbourne, Australia, ³University of Pennsylvania, Philadelphia, United States, ⁴University of Michigan, Ann Arbor, United States, ⁵Medical University of South Carolina, Charleston, United States

Introduction: A relationship between reduced brain tissue oxygenation (PbtO2) and poor outcome following severe traumatic brain injury (sTBI) has been reported in several trials, including a Phase II randomized trial.

Methodology: Large single-blind, randomized, controlled, phase 3, multi-center parallel studies examining brain oxygenation optimization are occurring internationally (BOOST-3: North America; BONANZA: Australia, New Zealand, and Europe). These trials will assess for benefit of multimodal ICP plus PbtO2 directed management on functional outcome in sTBI. The target population for both studies are patients presenting with acute sTBI requiring intracranial pressure monitoring and randomized within 12 hours from injury, using an exception from informed consent where allowable. Patients are randomized to receive tiered management interventions aimed at treating either ICP and PbtO2 or ICP only. Primary outcome measure in both studies is a blinded 6-month post-injury GOS-E, with BOOST-3 utilizing a sliding dichotomy analysis approach based on the IMPACT Core Model and BONANZA using a fixed dichotomy where GOS-E>4 is a favorable outcome.

Results: Target enrollment is statistically powered within each study with current enrollment of BOOST-3: 608/1094 at 44 sites and BONANZA: 121/860 at 15 sites. Demographics in each study are comparable with average age 40 years, 39 years; 81%, 79% male and mean initial GCS score 6 and 5, BOOST-3 and BONANZA, respectively.

Conclusion: Unique in the planning processes of trial design was foresight to allow for a future IPDMA (individual patient data meta-analysis), which requires the harmonization of many data elements across studies. Thus far, this is a targeted achievement.

D054 MAST Trials – strengthening the evidence behind the use of antiepileptic drugs in TBI patients

Mr Edoardo Viaroli , Dr Harry W. Mee, Mrs Carole Turner, Mr Angelos Kolias, Prof Peter J.A. Hutchinson

¹University of Cambridge, Department of Clinical Neurosciences, Division of Neurosurgery, Cambridge, United Kingdom, Cambridge, United Kingdom

Objectives: This NIHR HTA-funded project aims to define the best practice in the use of AEDs for TBI patients by conducting two studies run in parallel but independent of each other.

Background: Post-traumatic seizures (PTS) are classified as early (within 7 days post-TBI) or late (after 7 days). The incidence of early PTS following severe TBI is as high as 14% and their prevention can limit impairments in brain autoregulation and may prevent the development of late PTS. There is no high-quality evidence regarding the optimal duration of treatment for patients started on an AED for acute PTS or regarding their prevention.

Methods: MAST Duration: A multi-centre, pragmatic, randomised trial, aiming to recruit 312 patients, to compare the clinical effectiveness (absolute difference in rate of late PTS 24 months post-TBI) of a longer course of AED (>6 months) versus a shorter course (up to 3 months) for TBI patients with early PTS.

MAST Prophylaxis: A multi-centre, pragmatic, three-arm, randomised trial, aiming to enrol 960 patients, to compare the clinical effectiveness (absolute difference in the rate of PTS within the first 2 weeks post-TBI) of a 7-day course of prophylactic phenytoin or levetiracetam versus no AED.

Results: Both studies are now in their third year of recruitment. 676 patients have been enrolled in MAST Prophylaxis, while 115 patients were recruited in MAST Duration.

Conclusions: Our trials are still open to recruitment and we are confident they will significantly contribute to strengthening the evidence surrounding AEDs in TBI. We are planning to open new Sites across the UK and abroad. Please contact us for further information at ev349@cam.ac.uk.

D055 A prehospital triage system to detect traumatic intracranial hemorrhage using machine learning algorithms

Dr Daisu Abe ,¹ Dr Motoki Inaji,¹ Mr Takeshi Hase,¹ Dr Yoji Tanaka,¹ Dr Yasuhiro Otomo,² Dr Taketoshi Maehara¹

¹Tokyo Medical And Dental University, Bukyo-ku, Japan, ²National Disaster Medical Center, Tachikawa, Japan

Background: An adequate prehospital triaging system for head trauma is essential for improving the prognosis of these patients. In this study, we developed a prehospital triage system to stratify head trauma patients according to trauma severity by using several machine learning techniques and to evaluate the predictive accuracy of these techniques.

Patients and Methods: This was a single-center retrospective cohort study. We reviewed the electronic medical charts of consecutive patients who were transported to Tokyo Medical and Dental University Hospital from April 2018 to March 2021 for head trauma. Patients younger than 16 years, with cardiopulmonary arrest on arrival or with a significant amount of missing data were excluded. We constructed machine learning-based predictive models to detect the presence of traumatic intracranial hemorrhage (tICH).

Results: A total of 2,123 head trauma patients were enrolled in this study (72% male; mean [SD] age of 57.6 [19.8] years), and tICH was detected in 258 of these patients. Among several machine learning algorithms, eXtreme Gradient Boosting (XGBoost) achieved the highest area under the receiver operating characteristics (AUROC) of 0.78±0.02 and area under precision recall curve (AUPRC) of 0.46 in cross validation studies. In the test set, the AUROC was 0.80, the sensitivity was 74%, and the specificity was 74.9%. The prediction using the NICE guidelines, which was calculated after consultation with medical doctors, had a sensitivity of 72.0% and a specificity of 73.3%. The McNemar test revealed no statistically significant differences between the XGBoost and the National Institute for Health and Care Excellence (NICE) guidelines for the sensitivity and specificity (P = 0.80 and 0.55,respectively).

Conclusion: In this cohort study, our predictive model achieved a comparatively accurate performance in detecting tICH using only pretransportation information. Further validation with a prospective multicenter dataset is needed in the future.

D056 Cambridge brain physics clinical neuromonitoring high-resolution physiology database

Dr Erta Beqiri ,¹ Ihsane Olakorede,¹ Xuhang Chen,¹ Dr Stefan Bogli,¹ Tommaso Rochat,¹ Claudia Smith,¹ Dr Marina Sandra Cherchi,¹ Dr Giada Cucciolini,^1,2 Virginia Motroni,¹ Dr Michal M Placek,¹ Dr Joseph Donnelly,^1,3 Dr Masumi Tanaka Gutiez,¹ Caroline Lambley,⁴ Mr Matthew R. Guilfoyle,⁵ Mr Adel Helmy,⁵ Dr Virginia Newcombe,⁴ Dr Ronan O’Leary,⁴ Prof Jonathan Coles,⁴ Prof Peter Hutchinson,⁵ Dr Ari Ercole,⁴ Dr Peter Smielewski¹

¹Brain Physics Laboratory, Department of Clinical Neuroscience, Division of Neurosurgery, University of Cambridge, Cambridge, United Kingdom, ²Department of Surgical, Medical, Molecular Pathology and Critical Care Medicine, University of Pisa, Pisa, Italy, ³Department of Medicine, University of Auckland, Auckland, New Zeland, ⁴Perioperative, Acute, Critical Care and Emergency Medicine (PACE) Section, Department of Medicine, University of Cambridge, Cambridge, United Kingdom, ⁵Neurosurgery Department, Cambridge University Hospitals NHS Foundation Trust, Cambridge, United Kingdom

Introduction: Patients admitted to intensive care with traumatic brain injury (TBI) receive advanced neuromonitoring. High resolution physiology (HRP) at waveform level can be recorded and HRP-derived metrics are known to be clinically relevant. The structure of such big data has been largely neglected from the worldwide attempts in providing a standardised database framework within ethically approved regulations. However, high-quality and reproducible research relies on such datasets. We aimed to implement a local framework that could enable the creation of a suitable database of HRP data for TBI patients.

Methods: We obtained ethical approval for the Brain Physics Database in 2023 (REC 23/YH/0085). The protocol defines procedures for data collection, deidentification (choosing a k-anonymity of 6), storage and project-based access for analyses. We retrospectively curated data collected between 2002 and June 2021, resulting in the TBI_v0.1 database (858 records). Pitfalls in the data structure (non-homogeneity, missing data) were identified and solutions proposed. From June 2021 we organised a neuromonitoring team that engaged in: ICM+ based standardised and structured (hardware, software) prospective data collection and weekly curation; weekly meetings for issues and physiological patterns discussion and cataloguing; engagement with the clinical staff. Data collected up to December 2023 were further curated and stored as TBI23_v0.0 database.

Results: TBI23_v0.0 contains records from 241 patients, featuring HRP time series (28 modalities, median(q1:q3) duration of 7(4:13) days), clinical descriptors (demographics, injury presentation and severity, relevant interventions), signals metadata relevant for waveform analysis, data quality metrics, compliance descriptors for common data elements, as well as pipelines and configurations for basic analyses. Harmonisation with the retrospective database resulted in a combined dataset of 1,099 records.

Conclusion: We implemented a sustainable and ethically-compliant framework for a high-quality database tailored to the HRP data structure. This will power analytical and artificial-intelligence-driven clinical research based on HRP in TBI patients.

D057 Casemix, management, and mortality of patients receiving emergency neurosurgery for traumatic brain injury in the Global Neurotrauma Outcomes Study: a prospective observational cohort study

Mr David Clark ,¹ Mr Amos Olufemi Adeleye, Mr Abdul Hafid Bajamal, Dr Tom Bashford, Mr Hagos Biluts, Mr Karol Budohoski, Dr Ari Ercole, Dr Rocío Fernández-Méndez, Mr Anthony Figaji, Dr Deepak Kumar Gupta, Dr Roger Härtl, Mr Tariq Khan, Mr Corrado Iaccarino, Mr Tsegazeab Laeke, Mr Andrés Rubiano, Mr Hamisi K Shabani, Mr Kachinga Sichizya, Mr Manoj Tewari, Mr Abenezer Tirsit, Mr Myat Thu, Mr Manjul Tripathi, Mr Rikin Trivedi, Dr Bhagavatula Indira Devi, Mr Franco Servadei, Dr David Menon, Mr Angelos Kolias, Professor Peter Hutchinson

¹University Of Cambridge, Cambridge, United Kingdom

Background: Neurosurgical interventions are an important aspect of care for patients with TBI, but there is little epidemiological data available on this patient population. We aimed to characterise differences in casemix, management, and mortality of patients receiving emergency neurosurgery for TBI across different levels of human development.

Methods: We did a prospective observational cohort study of consecutive patients with TBI undergoing emergency neurosurgery. The primary outcome was mortality at 14 days postoperatively. Countries were stratified according to their HDI. Mixed effects logistic regression was used to examine the effect of HDI on mortality.

Findings: Our study included 1635 records from 159 hospitals in 57 countries, collected between Nov 1, 2018, and Jan 31, 2020. The median age was 35 years (IQR 24–51), with the oldest patients in the very high HDI tier (median 54 years, IQR 34–69) and the youngest in the low HDI tier (median 28 years, IQR 20–38). The most common procedures were elevation of a depressed skull fracture in the low HDI tier (69 [45%]), evacuation of a supratentorial extradural haematoma in the medium HDI tier (189 [31%]) and high HDI tier (173 [32%]), and evacuation of a supratentorial acute subdural haematoma in the very high HDI tier (155 [47%]). After adjustment for casemix, the odds of mortality were greater in the medium HDI tier (odds ratio [OR] 2·84, 95% CI 1·55–5·2) and high HDI tier (2·26, 1·23–4·15), but not the low HDI tier (1·66, 0·61–4·46), relative to the very high HDI tier. There was significant between-hospital variation in mortality (median OR 2·04, 95% CI 1·17–2·49).

Interpretation: Patients differed considerably in their admission characteristics and management by HDI. Substantial opportunities to improve care globally were identified, including reducing delays to surgery.

D058 European Neurotrauma Organization (ENO): a focused summary of the neurotrauma outcome working group from 2022 to 2024

Thom van der Meer, MSc.,¹ Prof. Dr. sc. nat. Marina Zeldovich,^2,3 PD Dr. Katrin Rauen ,^4,5,6,7 ENO Neurotrauma Outcome Working Group

¹Department of Geriatric Psychiatry, Psychiatric Hospital Zurich, University of Zurich, Switzerland, ²Faculty of Psychotherapy Science, Sigmund Freud University, Vienna, Austria, ³Institute of Psychology, University of Innsbruck, Austria, ⁴Neurological Rehabilitation Center Godeshöhe, Bonn, Germany, ⁵Institute for Stroke and Dementia Research, University Hospital, LMU Munich, Germany, ⁶Department of Traumatology, University Hospital Zurich, Switzerland, ⁷Center for Psychiatric Research, Psychiatric University Hospital & Neuroscience Center Zurich, University of Zurich, Switzerland

Introduction: The ENO Neurotrauma Outcome Working Group was established during the INTS 2022 meeting in Berlin. It aims to advance clinical and scientific knowledge, to promote academic teaching, and to foster interaction among neurotrauma experts across Europe. This is particularly relevant as approximately 80 million people are affected annually by neurotrauma, including traumatic brain injury (TBI), spinal cord injury (SCI), and degenerative cervical myopathy (DCM). Here, we summarize the efforts and achievements of this neurotrauma consortium since its inception.

Methodology: Multiple experts in different (sub)fields of neurotrauma shared their insights during the 6-weekly meetings on the most recent clinical practice and academic advances since 2022. Structural efforts were made to group this information into status quo and need for attention across three predefined areas: (a) neurotrauma, (b) diagnostics and outcome assessment, and (c) rehabilitation, to summarize the status and focus on improving future research design, clinical practice, and academic teaching.

Results: Key findings from all presentations showed that 30–50% of TBI patients have persistent symptoms at or beyond six months post TBI. Up to 90% of TBI patients report ongoing symptoms in one out of the five outcome domains (daily life activities, physical, cognitive, speech, and psychological problems). TBI, SCI, and DCM diagnostics need more comprehensive, multidimensional, and standardized assessments, and particularly DCM and posttraumatic neurodegeneration remain often undiagnosed. Across Europe, there is still a lack of appropriate rehabilitation service with less than 30% of TBI patients receiving cognitive and/or psychological treatment to overcome their long-term burden after TBI.

Conclusions: This consortium work highlights key areas for improving clinical research and practice. Specifically, discrepancies between historical prognostic modeling and actual morbidity trajectories; and how improvements in multimodal imaging and neurotrauma assessment are presented as focal points to improve the wide-ranging demands for enhanced rehabilitation in patients with neurotrauma.

D059 Impact of Temporal Resolution on Autocorrelative Features of Cerebral Physiologic Data Streams in Acute Traumatic Neural Injury: A CAnadian High Resolution TBI (CAHR-TBI) Cohort Study

Mx Nuray Vakitbilir ,¹ Dr Donald Griesdale,² Dr Mypinder Sekhon,² Dr Francis Bernard,³ Dr Clare Gallagher,⁴ Dr Eric Thelin,^5,6 Dr Rahul Raj,⁷ Dr Marcel Aries,⁸ Dr Logan Froese,⁶ Mr Kevin Y. Stein,¹ Dr Andreas Kramer,⁴ Dr Frederick A. Zeiler^1,9

¹University of Manitoba, Winnipeg, Canada, ²University of British Columbia, Vancouver, Canada, ³University of Montreal, Montreal, Canada, ⁴University of Calgary, Calgary, Canada, ⁵Karolinska University Hospital, Stockholm, Sweden, ⁶Karolinska Institutet, Stockholm, Sweden, ⁷University of Helsinki and Helsinki University Hospital, Helsinki, Finland, ⁸University Maastricht, Maastricht, The Netherlands, ⁹Pan Am Clinic Foundation, Winnipeg, Canada

Introduction: Therapeutic interventions during the acute phase of care in acute traumatic neural injury, termed traumatic brain injury (TBI), rely on continuous multi-modal cerebral physiologic monitoring devices for detection and prevention of secondary brain injury. Such data streams pose significant challenges for treating clinicians, as they are often of high temporal resolution and difficult to integrate into contemporary temporally resolved prognostic models. Previous attempts to remedy these issues with large data streams have included data reduction techniques, such as moving average filters, which decrease the amount of information into more manageable lower resolution summary metrics. However, such methods may not address statistical autocorrelative features, or introduce new statistical features, which impact the utility and interpretation of any models which include such data streams.

Methodology: Using the CAHR-TBI high-frequency cerebral physiologic data set, we aimed to characterize the impact of different data resolutions on the Box-Jenkin's time-series statistical structure (i.e., autoregressive integrated moving average (ARIMA) model) for both raw and derived multi-modal cerebral physiologic measures. Data resolutions from 1 minute to 24 hours were evaluated, testing signal stationarity and optimal ARIMA model features at the individual signal and patient levels.

Results: 380 patient data sets were interrogated at different temporal resolutions. Regardless of cerebral physiologic measure, at least first-order differencing was required, for all data resolutions, given persistent trend structure. Furthermore, variation in the optimal time-series autoregressive structures was observed across individual signals and patients, highlighting the critical need for careful consideration of ARIMA structures in temporal modeling for precise interpretation of model significance and performance.

Conclusions: The findings reveal that regardless of the data resolution, both raw and derived cerebral physiologic indices consistently exhibit autoregressive features that are intrinsic and cannot be disregarded. Ignoring such features may impact the significance of any models developed using such data streams.

D060 Interaction of Obesity and NLRP3 Inflammasome Activation following Mild Traumatic Brain Injury

Dr Shawn Eagle ,¹ Mahesh Basantani,¹ Natalie Sherry,¹ Peyton McIntyre,¹ Erin Kershaw,¹ Ava Puccio,¹ David Okonkwo¹

¹University of Pittsburgh, Pittsburgh, United States of America

Introduction: The NOD-like receptor pyrin domain-containing protein 3 (NLRP3) inflammasome has been associated with worse outcomes from severe traumatic brain injury (TBI) and obesity. Little is known about the potential effect of mild TBI (mTBI) on the NLRP3 inflammasome and the extent to which modifying factors, such as obesity, may augment the inflammatory response. The purpose was to evaluate the association of NLRP3 inflammasome proteins with obese body mass index (BMI≥30) within 24 hours of mTBI.

Methodology: This is a secondary analysis of patients with mTBI who presented to the emergency department of one U.S. Level 1 trauma center (n=243). Participants completed Trail Making Test Part A and B (TMT-A and -B) at six- and twelve-months, Blood samples were obtained within 24 hours of mTBI and apoptotic speck-like protein (ASC), caspase-1, interleukin-18 (IL-18), and IL-1β were assayed. Regression models evaluated the association of NLRP3 proteins and BMI as well as the potential interaction effect of higher BMI with NLRP3 proteins (n=243) on total symptoms (Rivermead post-Concussion Questionnaire), quality of life (Quality of Life after Brain Injury), physical function, (Short Form-12 Physical) TMT-A and TMT-B at two-weeks, six-months, and twelve-months.

Results: There was an association between IL-18 (p=0.004) and IL-1β (p=0.01) in mTBI patients with obese BMI compared to non-obese BMI (<30). Moderation analyses revealed interaction effects between ASC, caspase-1, IL-18, IL-1β and obese BMI which worsened symptom burden, quality of life, and physical function at two-weeks and six-months (p<0.05). The interaction of severely obese BMI and IL-1β was associated with TMT-B at six- (p=0.049) and twelve-months (p=0.02). Interaction of ASC with severely obese BMI was associated with TMT-B at six- (p=0.02) and twelve-months (p=0.02).

Conclusions: Obese BMI was associated with higher NLRP3 inflammasome responses to mTBI and worsened short- and long-term outcomes. NLRP3 may represent a treatment target for mTBI.

D061 A clinical trial of an investigational medicinal product for patients with traumatic brain injury in the intensive care unit

Mrs Natalie Khalique ,¹ Mrs Dot Chatfield,¹ Mr Adam Safwat,¹ Mr Adel Helmy,¹ Mr Arun Gupta¹

¹Cambridge University Hospitals, CUH, United Kingdom

Introduction: Traumatic brain injury (TBI) is one of the leading causes of mortality and morbidity globally. The long-term impact of sustaining a TBI can be devastating on the individual, their families, the healthcare system, and world economy.

There has been considerable research investigating the mechanisms and effect of secondary brain injury, with current ICU management centred around controlling brain oedema and subsequent rise in ICP.

One such cause of secondary oedema is the expression of substance P, which is widely distributed throughout the nervous system and is associated with neurogenic inflammation. Although current treatments have proven to be effective to a point, this study redirects the focus of treating secondary brain injury to a cellular level by blocking the expression of substance P with the use of a neurokinin-1 receptor (NK-1R) antagonists.

Objectives: This clinical trial aims to assess the efficacy and safety of a NK-1R antagonist in TBI. In addition, we hope to demonstrate a reduction in cerebral oedema and ICP post drug administration.

Methods: The trial is currently being conducted in the UK and Australia. The open label pilot phase which aimed at assessing the safety of EU-C-001 is now complete.

The trial moves into a larger double-blind phase with the drug being given over a four-day period, with patients receiving a weight adjusted dose twice a day. Safety and efficacy continue to be evaluated.

Conclusion: To date, patients enrolled into the open label phase have experienced little or no treatment related adverse events. The drug has been shown to be safe to administer and its efficacy continues to be investigated. We hope to demonstrate that NK-1R antagonist will significantly reduce the secondary insults to the brain post TBI, thereby reducing mortality and increasing the chance of survival with a better quality of life.

D062 Microglial process convergence and spatially resolved transcriptional changes in human postmortem tissue from traumatic brain injury cases

Dr. Amanda Logan-Wesley,¹ Dr. Declan McGuone,^2,3 Radina Lilova,¹ Dr. Peter Hamilton,¹ Karen Gorse,¹ Dr. Audrey Lafrenaye ¹

¹Virginia Commonwealth University, Richmond, United States, ²Yale School of Medicine, New Haven, United States, ³State of Connecticut Office of the Chief Medical Examiner, Farmington, United States

Introduction: Traumatic brain injury (TBI) affects millions globally with a majority of TBIs being mild, in which diffuse pathologies prevail. While progress has been made investigating the breadth of TBI-induced microglial changes in rodents, the neuroinflammatory progression following brain injury in humans, particularly within the thalamus, is not well-understood. Our previous studies demonstrated that microglial processes converge onto injured axons within the pig, but not the rat, thalamus following diffuse TBI. In this study we sought to validate our pig findings in human postmortem tissue and to advance investigations of microglial changes following diffuse TBI in the human thalamus.

Methods: We utilized both biobanked samples from various brain regions from the Uniformed Services University (USU) that were curated into TBI cases with axonal injury and control cases. We also used postmortem thalamic tissue from a population-based sample of critically investigated cases of lethal TBI from the State of Connecticut Office of the Chief Medical Examiner (OCME). Using the USU tissue paired with multiplexed immunohistochemistry against Iba-1 for microglia, APP for injured axons, and MBP for myelinated fibers, we investigated the degree of microglial process convergence onto injured axonal swellings compared to myelinated fibers. To extend these studies we are actively investigating the whole transcriptome of microglia in the thalamus utilizing the OCME tissue paired with digital spatial profiling using the GeoMx platform.

Results: We found microglial process convergence occurs in human tissue, in which more microglial processes converged directly onto APP+ injured axons compared to MBP+ fibers in either injured or control samples (X2(3)=15.53, p=0.001). The digital spatial profiling studies are actively ongoing.

Conclusions: These studies validate our findings within our pig model of diffuse TBI and expand on our knowledge of transcriptional changes to microglial in the human thalamus following brain injury.

Funding: NINDS grants R21NS126611, R56NS128104, and R01NS128104

D063 Analysis of chemokine receptors and their ligands (CXCL11 and CXCL12) in severe TBI patients

Adaliana Mousessian, Wellingson Paiva ,¹ Sueli Oba-Shinjo, Angelos Kolias, Suely Marie

¹University of Sao Paulo, Sao Paulo, Brazil

Introduction: Traumatic brain injury (TBI) is a global health issue due to the neurological dysfunctions caused by subsequent primary and secondary injuries. Neuroinflammation plays a crucial role in containing cerebral damage and promoting recovery in the pericontusional area. Considering the role of chemokines in neuroinflammation, we conducted an analysis of chemokine receptors (CXCR4, CXCR7) and their ligands (CXCL11 and CXCL12), as well as transcriptome analysis using RNA-Seq in samples of TBI cerebral contusion compared to TBI-free cerebral tissue to understand the activated signaling pathways during the acute phase of TBI in humans.

Methods: Twenty-two pericontusional cerebral tissue samples from severe TBI patients requiring surgery were compared with 20 TBI-free autopsy control samples using real-time quantitative PCR (RQ-PCR) and immunohistochemistry. Subsequently, transcriptome analysis via RNA-Seq was conducted on 10 TBI samples compared to 10 autopsy control samples.

Results: Significant increases in gene expression of both receptors and their ligands were observed in TBI, demonstrating high sensitivity and specificity in discriminating TBI from control. The expression level of CXCR7 receptor correlated with ligand expressions. Elevated expression levels of these chemokines were associated with favorable clinical evolution up to 6 months of TBI follow-up. Pathway enrichment analyses by RNA-Seq identified increased expression of genes involved in inflammation, angiogenesis, and extracellular matrix (ECM) remodeling, along with decreased expression of genes related to ion transport and synaptic transmission in TBI samples compared to controls. A highly connected network was observed among positively regulated genes, and a positive correlation between CCL2 and SPHK1 was noted, associated with inflammation and cell proliferation, respectively.

Conclusion: Simultaneous activation of angiogenesis, inflammation, and ECM remodeling pathways suggests processes activated for damaged tissue recovery. Genes identified in the acute phase of human TBI represent potential candidates for clinical evolution monitoring and new therapeutic strategies.

D064 Investigating the role and heterogeneity of regulatory T-cells in post-Traumatic Brain Injury (TBI)

Mr Aaditya Prabhu ,¹ Dr Zoya Georgieva,^1,3 Dr Daniel Rainbow,¹ Dr Lorna Jarvis,¹ Dr Annabelle Curle,¹ Dr Valerie Coppard,¹ Dr Vincent Degos,⁵ Dr Alice Jacquens,⁵ Dr Adel Helmy,⁴ Dr David Menon,² Dr Edward Needham,^2,3 Dr Joanne Jones^1,3

¹Department of Clinical Neurosciences, University of Cambridge, Cambridge, United Kingdom, ²Department of Anaesthesia, University of Cambridge, Cambridge, United Kingdom, ³Department of Neurology, University of Cambridge, Cambridge, United Kingdom, ⁴Department of Neurosurgery, Cambridge, United Kingdom, ⁵NeuroAnesthesia and NeuroCritical Care Medicine - Pitié Salpêtrière Hospital-Sorbonne University, Paris, France

Introduction: There is a growing awareness of the role of inflammation in secondary injury and outcome diversity following traumatic brain injury (TBI), and recent preclinical studies have identified regulatory T-cells (T-regs) as potential therapeutic tools that could be harnessed to mitigate inflammation-induced secondary injury. T-regs are a specialised subset of T cells that constitutively express the transcription factor FOXP3 and suppress exaggerated immune responses, promoting homeostasis and self-tolerance. In this project, we analysed blood samples collected from patients with moderate to severe TBI (GCS <12 at presentation) to determine the nature of T-reg responses post-injury. We hypothesised that inadequate T-reg expansion (or function) is associated with poor neurological outcomes.

Methodology: Conventional flow cytometry was used to identify and characterise the T-reg response in TBI patients on Day 7 post-injury. To confirm that any identified FOXP3-expressing cells post-TBI are bonafide Tregs (and not activated effector cells transiently expressing FOXP3), we performed bisulfite next-generation sequencing of FACS sorted Tregs to assess the methylation status of the T-reg specific demethylated region (TSDR) of the FOXP3 gene. Finally, a 26-colour spectral flow cytometry panel was used to explore heterogeneity in the T-cell pool (including T-regs) at various time points post-injury.

Results: By flow cytometry, we observed an increase in the proportion of CD4+ T-regs following TBI. Their identity was confirmed by demonstrating that the sorted Treg population was fully demethylated at the TSDR - as expected. The spectral panel revealed that although there is a T-reg expansion at Day 7, several markers, such as PD-1, are highly expressed suggesting that these Tregs may be functionally exhausted. This remains to be confirmed.

Conclusions: We have demonstrated that there is a peripheral expansion of T-regs 7 days post-TBI, but these cells seem to have an “exhausted” phenotype which may impair their ability to suppress inflammation effectively.

D065 Selective Brain Cooling is possible via CSF Exchange with double lumen external ventricular drainage- proof of concept in porcine model

Behnam Rezai Jahromi ,¹ Jaakko Kauhanen, Niko Järveläinen, Jarkko Kuivalainen, Vladimir Zamotin,¹ Nicholas Brandmeir, Mika Niemelä, Seppo Ylä-Herttuala

¹Helsinki University Hospital, Helsinki, Finland

Hypothermia, considered neuroprotective post brain insult (BI), alters brain metabolism by reducing oxygen and glucose consumption in the acute phase. Prior complications primarily stemmed from systemic hypothermia, with selective brain cooling not proven in large mammals without core temperature changes.

To investigate selective brain cooling, we inserted double-lumen active external ventricular drainage (aEVD) into four pigs' lateral ventricles. We augmented cerebrospinal fluid (CSF) exchange with spinal drainage, infusing cooled NaCl (1 pig) or ringer acetate (3 pigs) solutions. CSF exchange rates ranged from 180ml-720ml/h, aiming to cool brain parenchyma while maintaining core temperature.

Results revealed significant temperature drops in contralateral brain hemispheres (2.2–3.1°C) while core temperature changed minimally (0.5°C). Ipsilateral temperatures decreased by 4.5–7.5°C to 29.9–33.8°C, with average core temperature at 37.7°C. Selective cooling duration varied (10 minutes to 1.5 hours) based on CSF exchange rate. One pig experienced arrhythmias as brain temperature approached 30.8°C, resolved by switching off NaCl solution; others using ringer acetate had no such events. Histological analysis indicated less edema and intact neural tissue in the cooler hemisphere post-stroke.

We are first to describe that selective brain cooling is possible via CSF exchange with double-lumen EVD to achieve a significant temperature difference between the body core and brain. Establishing optimal timing and rates of cooling is crucial, especially considering arrhythmias observed in some cases. Further research is warranted before human trials commence.

D066 Parsimonious inflammatory mediators define immune-response endotypes associated with global outcome in traumatic brain injury

Dr Romit Samanta ,¹ Ms Anne-Cécile Chiollaz,² Dr Edward Needham,¹ Dr John Yue,³ Dr Adel Helmy,¹ Professor Elisa Zanier,⁴ Professor Kevin Wang,⁵ Professor Firas Kobeissy,⁵ - - TRACK-TBI,⁶ - - CENTER-TBI,⁶ Dr Jussi Posti,⁷ Professor Charlotte Summers,¹ Professor Geoffrey Manley,³ Professor Andrew Maas,⁸ Professor Olli Tenovuo,⁷ Professor Jean-Charles Sanchez,² Professor David Menon¹

¹University Of Cambridge, ²University of Geneva, ³University of California San Francisco, ⁴Istituto di Ricerche Farmacologiche Mario Negri IRCCS, ⁵Morehouse School of Medicine, ⁶Participants and Investigators, ⁷University of Turku, ⁸University of Antwerp

Abstract not published

D067 Characterisation of complement activation in human severe traumatic brain injury

Ms Claudia Smith ,¹ Dr Adam Safwat,² Dr Alice Jacquens,³ Dr Vincent Degos,³ Professor David Menon,² Dr Edward J Needham,² Mr Adel Helmy¹

¹Department of Clinical Neurosciences, University Of Cambridge, Cambridge, United Kingdom, ²Department of Anaesthesia, University Of Cambridge, Cambridge, United Kingdom, ³Department of Anaesthesia and Intensive Care, APHP-Sorbonne University Hospital, Pitié-Salpêtrière, Paris, France

Introduction: Neuroinflammation is known to contribute to worse outcomes following severe traumatic brain injury (sTBI). Specifically, complement activation, as part of the innate immune response, has been explored in animal models and post-mortem human tissue, however it has not been well characterised in a substantial clinical cohort. We aimed to investigate acute (within the first 10 days) and persistent (day 42, and 365) complement activation in sTBI patients, and describe this temporal relationship in the context of clinical variables.

Methods: Plasma blood samples from 65 sTBI patients (over days 1–10, 42, and 365) and 17 single timepoint healthy controls were analysed for 10 complement analytes using single and multiplexing protein assays. Analytes included initiator (MBL, MASP2, ficolin3), effector (C4b, C2, FactorD, FactorI) and downstream (C5, C5a, C5b9) complement proteins. Clinical variables available in a subset of patients include injury severity scale, Glasgow coma scale motor score, pupil reactivity, and Glasgow outcome scale at 6 months post injury.

Results: Overall mean complement levels were significantly elevated in 8 out of the 10 proteins compared to healthy controls (all p values < 0.03). Ficolin3, an initiator, was particularly lower in sTBI versus controls (0.21 vs 1.76ng/ml, p < 0.001), and was related to injury severity. Initiators MBL and MASP2 peaked at day 5, with effectors’ peak following at day 6–7. Downstream analytes peaked at day 8, with C5b9, the terminal complement complex, significantly higher in sTBI (1738 versus 985ng/ml, p=0.03) indicating full complement activation.

Conclusion: This study characterised the complement cascade in human sTBI with high temporal resolution in initiator, effector, and downstream proteins. 9 out of the 10 proteins analysed were significantly different in sTBI versus healthy controls, with terminal complex showing complete complement activation in sTBI. Further investigation is required, with complement as a potential therapeutic target in sTBI.

D068 Teaching Targeted Management of Neurotrauma in Low and Middle Income Countries - A Pilot Course and Experience

Ms Swati Jain ,¹ Mr David Clark,¹ Ms Sara Venturini,¹ Dr Fahim Anwar,¹ Ms Nicola Owen,¹ Prof Peter Hutchinson,¹ A/Prof Rikin Trivedi¹

¹University Of Cambridge, Cambridge, United Kingdom

Traumatic brain injury (TBI) is increasingly contributing a substantial proportion to the global disease burden. Rapid assessment and neurotrauma directed management in the golden hour of trauma are prudent to improving outcomes after TBI. There has been significant investment in developing modified courses to teach a standardised approach and management of general trauma in resource limited low- and middle-income countries (LMICs). Management of neurotrauma forms a very small component of these courses and more often than not, overlooked, leaving a substantial gap in knowledge for the providers dealing with neurotrauma. With experiences from teaching courses such as Advanced Trauma Life Support (ATLS) and Primary Trauma Care (PTC), the authors have developed a targeted neurotrauma course that focusses on recognition, assessment, and initiation of neuroprotective measures in resource limited setting. A 2-day brain and spine trauma course has been developed that is suitable for first responders dealing with trauma. The course includes a mixture of didactic teaching, case based discussion, and hands on session with modifiable components depending on the experience and knowledge background of the audience. Day 1 would focus brain injury and day 2 on spine injury. Prior to starting the course, a multiple-choice questionnaire is undertaken by the participants to gauge their current knowledge in managing neurotrauma. After completion of the course, the participants attempt the questionnaire again to assess the knowledge improvement. The authors have conducted the courses in Cairo, Egypt and Lusaka, Zambia for 7 and 48 participants respectively. There was 33% and 28.4% increase in MCQ scores for Cairo and Lusaka group respectively. Detailed feedback was collected at the end of the course. Participants across both courses rated the focussed delivery of principles of management of neurotrauma high. The authors are planning future courses to assess its suitability and effectiveness in LMICs.

D069 The Global Epidemiology and Outcomes after Traumatic Brain Injury (GEO-TBI) registry: A Collaborative TBI Audit and Research Platform – Status Update and Ongoing Research

Mr Tommi Kalevi Korhonen ,¹ Mr Midhun Mohan,¹ Mr David Clark,¹ Mr Michael Martin,¹ Mr Angelos Kolias,¹ Prof Peter Hutchinson,¹ Mr Alexis Joannides¹

¹NIHR Global Health Research Group on Acquired Brain and Spine Injury, Cambridge, United Kingdom

Introduction: Traumatic brain injury (TBI) epidemiology is unclear, treatments are variable and outcomes differ significantly between healthcare settings. The factors underlying these differences remain unclear. The Global Epidemiology and Outcomes after Traumatic Brain Injury (GEO-TBI) registry was established to produce a more accurate global picture of the burden of TBI and intervention outcomes to facilitate service benchmarking against a global standard, highlight best practices and facilitate evidence-based prioritization of TBI in decision-making.

Methodology: The GEO-TBI provides a readily accessible, standardised platform for TBI data collection, audit and collaborative research, which was constructed via a consensus-based process. The first study to run on the platform, GEO-TBI: Incidence, is currently ongoing.

Results: The registry has been active for one year, and currently includes 753 patients from eight centres situated in Africa, Asia, Americas and Europe. Patients from low-to-middle income countries were younger than those from higher income countries, and differences in TBI mechanisms were evident.

Conclusions: We will present the one-year report of the international GEO-TBI dataset, which will provide epidemiological and outcome data from different healthcare settings. We will also highlight future studies to be run on the collaborative registry. Further information is available on https://geotbi.org.

D070 The Effects of COVID -19 Lockdown on Head Injury Patients in an Indian Tertiary Trauma Care Center

Dr Karthigeyan Madhivanan ,¹ Dr Dhandapani Sivashanmugam, Dr Pravin Salunke, Dr Rajesh Chhabra, Dr Sunil Gupta

¹Postgraduate Institute Of Medical Education & Research (pgimer), Chandigarh, India

Introduction: The COVID-19 lock down strategies has possibly altered neurotrauma patterns and care. We studied the impact of COVID-19 restrictions on head injury (HI) patients in a high-volume Indian referral trauma-center.

Methodology: The admission variables of patients with HI were retrospectively analyzed during the phases of lock-down and its sequential lifting, in comparison with that of a similar time-frame (190 days) just prior to COVID-19; the HI volume, clinico-epidemiological and radiological data were considered.

Results: A total of 3372 patients were studied. The head-injury admissions during the pre-pandemic period was 83 per week which dropped by 60% to 33 during lock phases and stabilized at 46 per week during unlock phases. The proportion of HI-grades remained unchanged. During lock-down, the incidence of vehicular accidents reduced however that of the falls and assault increased. Importantly, the injury to arrival time was higher during COVD-19. Between the pre-COVID and lock-down phases, the mortality for severe HI (47% vs 59%) was significantly different.

Conclusions: A decline in HI-admissions with change in etiology and prolonged time of hospital arrival has occurred during COVID-19. The lock down also impacted the mortality in severely head injured patients. Future COVID curtailing measures should take into consideration such unintended effects in the neurosurgical emergency services. Sufficient resources should be potentially mobilized towards neurotrauma patients, even during such unprecedented times.

D071 Fairway to Fractures: Income inequality and violent crime as the driving factors for Golf Club related Assaults - A Case Series of 21 compound skull fractures in South Africa

Dr. Ruan Grobler,¹ Dr Iain Walker ,¹ Prof Adriaan Johannes Vlok¹

¹Stellenbosch University, CAPE TOWN, South Africa

Background: Golf club related traumatic brain injuries are an uncommon occurrence in adults, with the use of golf clubs as a weapon of interpersonal assault resulting in compound skull fractures being rare. This study presents a case series of golf club related compound skull fractures in adults secondary to assault, representing the largest study of this entity to date.

Methods: A retrospective analysis was performed of a prospectively maintained database for patients admitted to Tygerberg Academic Hospital in Cape Town, South Africa, with golf club related compound skull fractures between 1st January 2018 to 31 December 2021. Data on demographic details, CTB findings, presenting Glasgow Coma Scale, surgical operative notes, septic complications, and outcomes at discharge were collected.

Results: A total of 21 patients were included, with the majority being male (95.2%) and the mean age being 32.6 years. Fractures were most commonly seen in the frontal bone (n=9), followed by parietal (n=8), temporal (n=3), and occipital (n=1) bones. Depressed skull fractures were the most common type of injury, and local pneumocephalus was present in the majority of patients. The mean presenting Glasgow Coma Scale was 14, and most patients had no focal neurological deficits. Surgical debridement was required in the majority of patients, with a high rate of septic complications (33.3%). However, most patients had good neurological outcomes at discharge, and the mean length of stay was 11.9 days. Discussion: South Africa has 300000 registered golfers and a high background prevalence of assault and income inequality with the highest Gini coefficient, a unique socio-economic background possibly contributing to golf club related assaults.

Conclusion: This study highlights the potential dangers of golf clubs as a weapon of interpersonal assault and the need for prompt and appropriate management of compound skull fractures to reduce the risk of complications.

D072 Early neuro-rehabilitation in traumatic brain injury: the need for an African perspective

Dr FRANKLIN CHU BUH ,¹ Prof. Peter JA Hutchinson,¹ Dr Fahim Anwar¹

¹University of Buea, Buea, Cameroon

Background: Traumatic brain injury (TBI) is a global public health challenge, affecting about 69 million individuals annually and being one of the leading causes of mortality. It has adverse consequences in terms of cognitive and physical functioning, which makes rehabilitation interventions an integral part of its management. Early neuro-rehabilitation guidelines for traumatic brain injury have not yet been developed and implemented in most of Africa especially Sub-Saharan Africa.

Body: We aimed with this Opinion to propose a collective reflection on the development and implementation of early neuro-rehabilitation guidelines as an integral part of the care in traumatic brain injury. The different aspects to be considered for reflection have been highlighted: Traumatic brain injury severity to be considered in early neuro-rehabilitation; who should be assessed and receive early neurorehabilitation, barriers to be considered for early neurorehabilitation; what early neurorehabilitation to be considered; the different phases involved in rehabilitation after mild, moderate, and severe TBI; and lastly, what perspective for the creation of neurorehabilitation teams. In conclusion, neuro-rehabilitation should start at the time of admission and should continue from the intensive care unit through the community for the moderate-to-severe traumatic brain injury population. However, mild TBI should also be considered for long-term follow-up in the community due to the fact that some mild traumatic brain injury patients might develop chronic cognitive problems or fatigue with time.

Conclusion: Neurorehabilitation should start at the time of admission and continue from the intensive care unit through the community for the moderate-to-severe traumatic brain injury population. There is a need to develop, agree on, and implement guidelines on early neuro-rehabilitation interventions for patients with moderate to severe traumatic brain injury in the African region, where disparities in care are common reality.

D073 The impacts of contusion expansion, bifrontal contusions and low platelet count on patient outcome following traumatic brain injury

Dr. Iftakher Hossain ,^1,2,3 Dr. Alice Andersson,² Prof. Niklas Marklund²

¹Department of Neurosurgery, Turku University Hospital and University Of Turku, Turku, Finland, ²Division of Neurosurgery, Addenbrooke's Hospital and University of Cambridge, Cambridge, UK, ³Department of Neurosurgery, Skåne University Hospital, Lund, Sweden, Lund, Sweden

Introduction: Cortical contusions are common in moderate-severe traumatic brain injury (TBI). Cortical contusions often expand, potentially causing neuro-worsening several hours to days post-trauma. While contusion expansion (CE) may affect outcome, potential clinical and radiological markers that can predict CE have been insufficiently explored. In the present single-center retrospective observational cohort study, we evaluated clinical outcome by the Glasgow Outcome Scale extended (GOSE) scale and evaluated risk factors for CE.

Methods: Adult TBI patients >18 years of age, and of all injury severities, were included. Main variables of interest were low platelet count, defined as <150x109/L, presence of bifrontal contusions and CE, defined as absolute contusion volume increase in cm³. Factors associated with CE and clinical outcome according to GOSE were analyzed.

Results: Between 2012–2022, 271 patients were included. Contusion size on admission correlated positively with CE, as did the Marshall and Rotterdam radiological classification scores. Bifrontal contusions were significantly larger at admission, experienced larger CE, and had a worse outcome than contusions in other locations. Patients with a platelet count <150x109/L experienced a greater volume CE and had a worse outcome when compared to patients with a normal platelet count. In a multivariate analysis, CE remained significantly associated with a poor outcome six months post- injury.

Conclusion: Contusion volume at admission, and Marshall- and Rotterdam scores, positively correlated to CE. Bifrontal contusions and a platelet count <150x109/L were associated with CE, and a poor clinical outcome. Large CE volumes were associated with a worse clinical outcome, and CE was per se associated with outcome in a multivariate analysis. Management of these risk factors for CE in the acute post-injury setting may be needed to attenuate contusion expansion and to improve clinical outcome in TBI patients suffering from cortical contusion injuries.

D074 Exposure to Repetitive Head Impacts and Traumatic Brain Injury is Associated with Tau Astrogliopathy

John Arena,¹ Professor William Stewart,² Professor Gabor Kovacs,³ Associate Professor Edward Lee,¹ John Robinson,¹ Professor Virginia Lee,¹ Professor John Trojanowski,¹ Assistant Professor Andrea Schneider,¹ Professor Douglas Smith,¹ Dr. Victoria Johnson ¹

¹University Of Pennsylvania, Philadelphia, United States, ²University of Glasgow, Glasgow, United Kingdom, ³University of Toronto, Toronto, Canada

Introduction: Exposure to traumatic brain injury (TBI) and/or repetitive head impacts (RHI) increases risk of a range of neurodegenerative pathologies, including chronic traumatic encephalopathy neuropathologic change (CTE-NC). Astrocytic tau pathology reminiscent of aging-related tau astrogliopathy (ARTAG) is a component feature of the pathognomonic lesion of CTE-NC in many cases. Nevertheless, the relationship between RHI/TBI exposure and tau astrogliopathy remains poorly characterized.

Methodology: Autopsy derived material from 556 individuals was selected to include: individuals with a history of single moderate or severe TBI (sTBI; survival >6 months; n=77); individuals with history of contact sports participation (n=45); or uninjured controls either with (n=397) or without (n=37) neuropathologically confirmed neurodegenerative disease (NDD). Representative tissue sections immunostained for hyperphosphorylated-tau (p-tau; PHF-1) were then assessed in accordance with harmonized evaluation criteria for ARTAG.

Results: Correcting for age and sex, p-tau immunoreactive thorn-shaped astrocytes (TSA), consistent with ARTAG in both morphology and distribution, were more frequently observed in sections from contact sports participants and individuals with sTBI history versus controls with (Sports p<0.001; sTBI p=0.006) or without NDD (Sports p<0.001; sTBI p=0.016). This was observed regardless of whether the pathognomonic lesion of CTE-NC was present in the section. Intriguingly, while subpial TSA at sulcal depths were observed in aged controls with and without NDD, this pathology was more commonly observed following RHI/TBI.

Conclusions: These findings indicate that exposure to RHI/TBI is associated with astroglial p-tau accumulation consistent with that typically encountered as ARTAG in normal aging or wider neurodegenerative disease. Possible common underlying mechanisms between trauma-related and age-related tau astrogliopathies will be important to explore. Further studies are required to determine the role of this glial pathology in TBI-related neurodegeneration (TReND) and its clinical association.

Funding: Supported by funding from DoD (TP220158) and NIH (R01NS094003; R01-NS123034; U54NS115322; P30AG072979; P01AG066597; U19AG062418; T32NS043126; K23NS123340).

D075 Subependymal injury of the anterior horn of lateral ventricles and rotational brain injuries

Dr Naohito Kuroda ,¹ Prof. Kazuki Harada,² Prof. Shirushi Takahashi³

¹Southern Tohoku General Hospital, Koriyama, Japan, ²Department of Forensic Medicine, Fukushima Medical University, Fukushima, Japan, ³Department of Forensic Medicine, Hirosaki University Graduate Shool of Medicine, Hirosaki, Japan

Subependymal injury of the anterior horn of the lateral ventricles (SEIAHLV) is not well recognized among patterns of fatal brain injuries. It is suspected that SEIAHLV might have some relationship with rotational acceleration to the head along the sagittal direction, based on our study of this traumatic lesion through autopsy analyses. It is well known that rotational head trauma causes diffuse axonal injury (DAI), but it is sometimes difficult to diagnose DAI properly by an autopsy, especially in cases without any indications of such particular movements. We have experienced more than 20 cases strongly suggesting rotational brain injuries, although the mechanism underlying SEIAHLV has not yet been clarified. The histological findings of SEIAHLV are characterized by a tear of the neuropil in the subependymal zone of the anterior horns of the lateral ventricles, minimal hemorrhaging, and ballooning or expansion of astrocytes. SEIAHLV also has several strange features; for example, it is observed bilaterally in almost all cases, it is formed in cases with a short survival time after injury, and it has not been detected in infants under one year old. It might be useful for diagnosing DAI with instant death, although it takes several hours for beta amyloid precursor protein to be detectable by immunohistochemistry, and it takes much longer for axonal retraction balls to form. The further accumulation of carefully performed case studies of SEIHLV until the lesion is regarded as further evidence for diagnosing DAI.

D076 Consensus Recommendations for Targeted Temperature Management in Severe Traumatic Brain Injury: Insights and Future Directions

Dr Masumi Tanaka,³ Ms Emily Sidlow,² Dr Andrea Lavinio ¹

¹University Of Cambridge, Cambridge, United Kingdom, ²Page & Page Healthcare, London, United Kingdom, ³King's College Hospital Foundation Trust, London, United Kingdom

Introduction: Traumatic brain injury (TBI) remains a leading cause of morbidity and mortality globally. The management of brain temperature post-TBI is critical for mitigating secondary brain injury. Recent consensus recommendations developed by a panel of international neuro-intensive care experts aim to standardize the approach to targeted temperature management (TTM) in severe TBI patients, addressing both the paucity of randomized controlled trials (RCTs) and the clinical heterogeneity in TTM application.

Methods: Utilizing a modified Delphi process, 18 experts reviewed systematic literature on TTM in severe TBI patients requiring intracranial pressure (ICP) management. Consensus thresholds were set at ≥88% for strong consensus and ≥78% for moderate consensus across multiple rounds of anonymous online surveys and discussions, culminating in a comprehensive set of recommendations.

Results: Strong consensus was achieved on the essential role of TTM in high-quality TBI care, emphasizing continuous temperature monitoring and prompt fever management. Controlled normothermia (36.0°C–37.5°C) was recommended within the first two tiers of the ICP management protocol, with individualization of temperature targets based on secondary brain injury risk assessment. The recommendations underscore the importance of understanding the physiological and pathological impacts of temperature variations on brain injury outcomes.

Conclusions: The expert consensus on TTM following severe TBI highlights critical areas for clinical practice and future research, including the necessity for individualized patient care and the potential benefits of controlled normothermia. This framework provides a foundation for improving TBI outcomes and underscores the need for further investigation into the optimal application and impact of TTM in diverse healthcare settings.

D077 Traumatic Cerebral Venous Sinus Thrombosis: incidence and outcomes at a UK major trauma centre

Mr Sheikh Muktadir Bin Momin ,^1,3 Mr Faheem Anwar,² Miss Georgina Shallard,² Mr Azam Baig,¹ Dr Yousra Rasool,¹ Dr Anam Fatima,¹ Mr Philip Ho,¹ Mr David J Davies,^1,3 Professor Ramesh Chelvarajah,^1,4 Professor Antonio Belli,^1,3 Mr Philip J O’Halloran^1,5

¹Department of Neurosurgery, Queen Elizabeth Hospital, Birmingham, United Kingdom, ²University of Birmingham Medical School, Birmingham, United Kingdom, ³Institute of Inflammation and Ageing, University of Birmingham, Birmingham, United Kingdom, ⁴Centre for Human Brain Health, University of Birmingham, Birmingham, United Kingdom, ⁵Department of Physiology & Medical Physics, Royal College of Surgeons of Ireland, Dublin, Ireland

Introduction: Traumatic Cerebral Venous Sinus Thrombosis (tCVST) is an increasingly recognised complication of TBI; treatment is nuanced by the risk of haemorrhagic and thrombotic complications, with a lack of evidence in the literature.

Methodology: A single-centre retrospective case series was performed from an institutional cranial trauma database. All patients >16 years old with radiologically-confirmed tCVST between 2021–2023 were included. All patients with devastating brain injury were excluded. Patient demographics, presenting neurological status, anticoagulation and surgical treatment strategy, haemorrhagic and/or thrombotic complications and recanalisation rate were recorded.

Results: 40 patients (28 male, mean age 41 (range 20–79)) had tCVST, 32.5% of which was occlusive. 45%, 22.5% and 32.5% of patients had severe (GCS ≦8), moderate (GCS 9–12) and mild TBI (GCS ≧13) respectively. 70% had isolated TBI with a median Injury Severity Score of 17. 87.5% of patients had an adjacent skull fracture or intracranial haematoma. 82.5% were commenced on prophylactic low molecular-weight heparin (pLMWH) ≦72 hours post-TBI (50% within 24 hours), whilst 5% of patients were commenced ≧5 days post-TBI and 7.5% were not given LMWH at all. 13 patients were uptitrated to treatment-dose LMWH (tLMWH) a median of 8 days following TBI, whilst two other patients were initiated on tLMWH 5 and 13 days post-TBI respectively. 25 patients required neurosurgical intervention, of whom 15 required a decompressive craniectomy. Although there were no haemorrhagic complications following LMWH therapy, two patients had venous infarcts resulting from tCVST. During follow-up imaging, 52.5% had full or partial recanalization of the affected venous sinus(es), which was independent of anticoagulation strategy.

Conclusion: In our cohort, tCVST was associated with severe TBI and adjacent skull fractures and intracranial haematomas. Although an early anticoagulation strategy following TBI was generally favoured, there were no haemorrhagic complications. Larger-scale prospective studies may identify optimal treatment strategies in tCVST.

D079 Acute measurement of communication within brain networks to predict cognitive recovery in mTBI patients

Phoebe Imms,¹ Nahian Chowdhury,¹ Nikhil Chaudhari,¹ Anar Amgalan,¹ Govinda Poudel,² Karen Caeyenberghs,³ Benjamin Hacker,¹ Andrei Irimia ¹

¹University Of Southern California, Los Angeles, United States, ²Australian Catholic University, Melbourne, Australia, ³Deakin University, Burwood, Australia

Introduction: An ill-defined proportion of mild traumatic brain injury (mTBI) victims experience enduring cognitive sequelae. Early identification of such victims would facilitate timely neurorehabilitation. This study explores the potential of communication measures within brain networks to forecast patients’ risk for developing unfavorable cognitive outcomes six months post-mTBI.

Methodology: Diffusion and T₁-weighted MRI was conducted alongside cognitive assessments one week and six months post-injury for a cohort of 113 adult mTBI patients (71 males), enabling connectome mapping. The association between communication efficiency within brain networks and cognitive performance was analyzed. Measures of network communication (characteristic path length, global efficiency, and navigation efficiency) were correlated with cognitive changes, adjusting for age and sex covariance.

Results: Findings indicate moderate correlation between communication metrics within task-related brain networks and cognitive outcomes. Unfavorable network communication correlated with worse outcomes within cognitive domains frequently impacted by mTBI (episodic and working memory, verbal fluency, inductive reasoning, and processing speed). Impairments in networks associated with verbal fluency were particularly predictive of poor cognitive recovery. Significantly longer and less efficient network pathways (all t > 2.786, p < 0.006) correlated to poorer verbal fluency, highlighting language’s vulnerability to mTBI. Relative inefficiency in task-related networks one week post-injury was associated with up to an eightfold increase in the likelihood of adverse cognitive outcomes pertaining to the network’s function.

Conclusions: This study highlights how acute network communication measures can predict patients at decreased risk for long-term cognitive deficits following mTBI. By classifying patients based on their risk for developing persistent cognitive sequelae, our findings can help stratify persons with mTBI according to their expected need for follow-up and/or neurorehabilitation. The emphasis on task-related network efficiency in domains such as verbal fluency can help tailor clinical interventions in vulnerable cognitive domains essential to core cognitive functions, ultimately facilitating better patient outcomes.

D080 A Digital Health Intervention for Concussion: Development and Clinical Feasibility Study

Mr Aimun Jamjoom ,^1,2 Dr Xin Yi Ng,³ Ms Christine d'Offay,² Dr Laura Alexander,⁴ Prof Alan Carson,⁴ Dr David Gillespie,^2,4 Prof Matthew Reed⁵

¹Queens Hospital, Romford, United Kingdom, ²HeadOn Health Ltd, Edinburgh, United Kingdom, ³The University of Edinburgh Medical School, Edinburgh, United Kingdom, ⁴Centre for Clinical Brain Sciences, The University of Edinburgh, Edinburgh, United Kingdom, ⁵The Emergency Medicine Research Group Edinburgh (EMERGE), Royal Infirmary of Edinburgh, Edinburgh, United Kingdom

Introduction: Concussion is a common condition that can lead to a constellation of symptoms that impact quality of life. There are several evidence-based interventions that have been found to improve postconcussion symptom burden. However, these are not routinely delivered, and individuals receive limited support during their concussion recovery.

Methods: A mixed methodology study involving a scoping review (n=21), behavioural analysis, and logic model to inform the intervention design. The intervention was optimised with feedback from individuals who had experienced concussions (n=12). The intervention was then offered to patients presenting to the emergency department with a concussion (n=50). Participants used the intervention freely and input symptom data as part of the program. A number of outcome measures were obtained, including participant engagement with the intervention, postconcussion symptom burden, and attitudes toward the intervention. A selection of participants (n=15) took part in in-depth qualitative interviews to understand their attitudes toward the intervention and how to improve it.

Results: Metrics indicated high levels of early engagement that trailed off throughout the course of the intervention, with an average daily completion rate of the symptom diary of 28.23% (494/1750). A quarter of the study participants (13/50, 26%) were classified as high engagers who interacted with all the functionalities within the intervention. Quantitative and qualitative feedback indicated a high level of usability and positive perception of the intervention. Daily symptom diaries (n=494) demonstrated a wide variation in individual participant symptom burden but a decline in average burden over time. For participants with Rivermead scores on completion of HeadOn, there was a strong positive correlation (r=0.86; P<.001) between their average daily HeadOn symptom diary score and their end-of-program Rivermead score.

Conclusion: Using this systematic approach, we developed a digital health intervention for individuals who have experienced a concussion that is designed to facilitate positive behaviour change.

D081 Unmet Healthcare Needs in Concussion

Ms Elika Karvandi ,¹ Mr Adel Helmy, Prof Peter Hutchinson, UK Head Injury Network

¹University Of Cambridge, Cambridge, United Kingdom

Objectives: To assess the unmet health needs in the concussion population from the healthcare provider’s perspective and identify priorities for improving concussion care in England.

Methods: The participants were clinical specialists in head injury practising in emergency medicine, neurology, neuropsychology, neurosurgery, paediatric medicine, rehabilitation medicine and sports and exercise medicine in England. All participants are involved in the UK Concussion Network. Qualitative study using a written open-ended survey format. This was completed and collected over a few months in 2022. Data was analysed using inductive thematic analysis using the six-step process framework.

Results: 14 reports were collected from the healthcare providers. The thematic analysis resulted in four main themes; (1) clinical practice, (2) government support to meet patient needs, (3) education and awareness, and (4) research.

Conclusions: The reported gaps in clinical care and priorities for improvement largely overlapped between healthcare professionals. This research provides the foundation to improve concussion care in England.

D082 Virtual assessment after mild traumatic brain injury: A randomised controlled trial

Dr Alexander Olaussen ,^1,2 Miss Madison Essery,^1,2 Mrs Abha Somesh,^1,2 Miss Carly Talarico,^1,2 Dr Jennifer Makovec-Knight,^1,2 Professor Biswadev Mitra^1,2

¹Alfred Health Emergency, Melbourne, Australia, ²Monash University School of Public Health and Preventive Medicine, Melbourne, Australia

Introduction: After a concussive head impact many people have persistent ongoing symptoms. Diagnosing concussion can be challenging for several reasons, including limited access to specialist care. Telehealth is an emerging alternative strategy. The aim of this study was to determine whether specialised telehealth consultation resulted in increased rates of diagnosis of concussion.

Methodology: A prospective, single-centre, open-label randomised controlled trial is being performed. People with a sustained head injury or suspected concussion (P), were randomised to the telehealth arm (I) or standard care arm (C), to assess the proportion of concussion diagnoses at 7 days post-injury (O). Two episodes of telehealth consultations were provided by a neuropsychologist. Data were collected for all patients through clinical nurse follow-up. The outcomes were assessed in the intention-to-treat population. The analysis of the primary outcome used log-binomial regression to estimate the risk ratio and 95% confidence intervals.

Results: Of the initial 82 patients screened, 14 had incomplete entries or were ineligible. Two were lost to follow-up, leaving 66 patients for analysis. The cohort had a mean age of 32.8 years (SD: 9.6) and were predominately females (n=39/66, 59.1% vs 40.9% males). At baseline, there were no statistically significant differences between the arms with respect to age, sex, previous concussions, or symptom severity as assessed by Rivermead Post Concussion Symptoms Questionnaire. Concussion diagnosis at day 7 post-injury was significantly higher in the intervention arm compared to the control (n=26/33, 79%, vs 11/33, 33%, Chi2 = 13.8, p<0.001). The risk ratio of being diagnosed with a concussion in the intervention arm was 2.4 (95%CI: 1.4 – 4.0, p<0.001).

Conclusions: This preliminary report shows that telehealth significantly increases the proportion of concussion diagnosed. Telehealth provides early access to specialists and may be a viable addition to support clinicians to improve concussion care.

D083 Contribution of psychological resilience, psychological distress and coping style to longitudinal perception of symptom burden and recovery after mild traumatic brain injury

Miss Melissa Papini ,^1,2 Ms Jacinta Thorne,^1,2 Dr Aleksandra Gozt,^1,2 Mr Andre Avila,^1,2 Ms Caerwen Beaton,^1,2 Ms Geena Gill,^1,2 Dr Elizabeth Thomas,^3,4 Dr Francesca Buhagiar,⁵ Mr Alexander Ring,^6,7 Dr Glenn Arendts,^8,9 Dr John Charles Iliff,^14,15,16,17 Professor Antonio Celenza,^18,19 Dr Sjinene Van Schalkwyk,¹¹ Dr Philip Brooks,^12,13,14 Dr Daniel Xu,^3,14,21 Dr Stephen Honeybul,^8,16,18,22 Dr Gill Cowen,^1,14 Associate Professor Carmela Pestell,^1,5 Professor Daniel Fatovich,^9,24,25 Dr Ben Smedley,¹⁰ Dr Ashes Mukherjee,²⁰ Dr Michael Bynevelt,²³ Professor Melinda Fitzgerald,^1,2 Dr Sarah Hellewell^1,2

¹Curtin Health Innovation Research Institute, Faculty of Health Sciences, Curtin University, Perth, Australia, ²Perron Institute of Neurological and Translational Science, Perth, Australia, ³Centre for Clinical Research Excellence, School of Population Health, Curtin University, Perth, Australia, ⁴Division of Pathology and Laboratory Medicine, School of Medicine, The University of Western Australia, Perth, Australia, ⁵School of Psychological Science, The University of Western Australia, Perth, Australia, ⁶6 Institute for Immunology and Infectious Diseases, Murdoch University, Perth, Australia, ⁷School of Physiotherapy and Exercise Science, Faculty of Health Sciences, Curtin University, Perth, Australia, ⁸Emergency Department, Fiona Stanley Hospital, Perth, Australia, ⁹Centre for Clinical Research in Emergency Medicine, Harry Perkins Institute of Medical Research, Perth, Australia, ¹⁰Emergency Department, Rockingham General Hospital, Perth, Australia, ¹¹Emergency Department, Joondalup Health Campus, Perth, Australia, ¹²Emergency Department, Saint John of God Midland Public Hospital, Perth, Australia, ¹³School of Medicine, The University of Notre Dame, Perth, Australia, ¹⁴Curtin Medical School, Curtin University, Perth, Australia, ¹⁵Emergency Department, Saint John of God Hospital Murdoch, Perth, Australia, ¹⁶Emergency Department, Royal Perth Hospital, Perth, Australia, ¹⁷Royal Flying Doctor Service- Western Operations, Perth, Australia, ¹⁸Emergency Department, Sir Charles Gairdner Hospital, Perth, Australia, ¹⁹Division of Emergency Medicine, School of Medicine, The University of Western Australia, Perth, Australia, ²⁰Emergency Department, Armadale Health Service, Perth, Australia, ²¹The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China, ²²Department of Health, Government of Western Australia, Perth, Australia, ²³The Neurological Intervention & Imaging Service of Western Australia, Sir Charles Gairdner Hospital, Perth, Australia, Perth, Australia, ²⁴Emergency Medicine, Royal Perth Hospital, Perth, Australia, ²⁵University of Western Australia, Perth, Australia

Introduction: Psychological resilience may alter perception of symptom burden and influence recovery from mild traumatic brain injury (mTBI). We investigated whether psychological resilience was associated with symptom burden and recovery at 3-,6- and 12-months and if sex and psychological distress were modulators. We hypothesised that lower resilience, and/or depression, anxiety or stress would be associated with increased symptom burden and recovery time and modified by sex and other factors.

Methodology: mTBI participants recruited within 7d of injury completed the Brief Resilience Scale (resilience), Depression Anxiety and Stress Scale-21 (psychological state) and Utrecht Coping List (coping style). Post-Concussion Symptom Scale (PCSS) scores at 3-,6- and 12-months were used to assess symptom burden and recovery (symptom severity score: males: >6, females: >7).

Results: 33 participants (36.8±13.5 years; 14F) were assessed within 7d (5.9±2.1) and were followed-up at 3-(n=30), 6-(n=29) and 12-months (n=28). Males were more likely to recover at 3-months only (M:50%, F:17% p=0.018). Univariate regression determined that low resilience (LR) significantly contributed to increased symptom severity at 3-months (β=19.09, p=0.022) but not beyond this time. LR contributed to increased 3-month symptom severity in multivariate models including sex (LR:β=20.36, p=0.022) and coping style (LR:β=19.97, p=0.012) alongside high avoidant and emotional coping styles. Models of resilience and DASS-21 determined that only elevated stress was associated with increased 3-month symptom severity (β=23.62, p=0.002). Adjusted for sex, resilience was not associated with recovery; elevated psychological distress increased the likelihood of prolonged 3- (OR: 1.2, 95% CI: 1.01–1.33, p=0.037) and 12-month recovery (OR: 1.4, 95% CI: 1.04–1.97, p=0.03). High active coping style increased the likelihood of prolonged 6-month recovery (OR: 1.6, 95% CI: 1.01–2.47, p=0.046).

Conclusions: Perceived symptom severity increases with low resilience, avoidant and emotional coping styles at 3-months. Elevated levels of stress and coping style supersede resilience as contributors to symptom burden and recovery.

D084 Estimating traumatic intracranial lesion risk in mild traumatic brain injury: preliminary results from the Stockholm score of lesion detection on computed tomography following mild traumatic brain injury (SELECT-TBI) study

Dr. Li Jin Yang ,^1,2 Dr. Alexander Fletcher-Sandersjöö,^3,4 Dr. Philipp Lassarén,³ Dr. Jonathan Tjerkaski,³ Dr. Erica Bergman,³ Dr. Frida Björkman,³ Dr. Jonas Bronge,³ Dr. Julia Antonsson,³ Dr. Kasper Teromaa,¹ Dr. Mariya Solodovnykova,³ Dr. Simon Örtqvist,¹ Dr. William Kylander,³ Dr. William Lindqvist,³ Dr. Kristian Ängeby,^1,2 Dr. Rebecka Rubenson Wahlin,^1,6,7 Dr. Eric Thelin^3,5

¹Department of Clinical Sciences and Education, Karolinska Institutet, Stockholm, Sweden, ²Department of Emergency Medicine, South General Hospital, Stockholm, Sweden, ³Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden, ⁴Department of Neurosurgery, Karolinska University Hospital, Stockholm, Sweden, ⁵Department of Neurology, Karolinska University Hospital, Stockholm, Sweden, ⁶Department of Perioperative Medicine and Intensive Care, Karolinska University Hospital, Stockholm, Sweden, ⁷Ambulance Medical Service in Stockholm (AISAB), Stockholm, Sweden

Introduction: Mild traumatic brain injury (mTBI) is a common cause of emergency department (ED) visits. However, only approximately 10% of these patients will develop an intracranial lesion (ICL) and 1% will require neurosurgical intervention due to their injury. Many existing guidelines have high sensitivity but low specificity, leading to overuse of computed tomography (CT) in the management of these patients in the ED. The aim of the Stockholm score of lesion detection on computed tomography following mild traumatic brain injury (SELECT-TBI) study is to generate a scoring algorithm to provide an individualized estimation for the risk of developing an ICL and requiring neurosurgical intervention in patients with mTBI.

Methodology: In this retrospective observational study, patients ≥ 15 years old who have undergone a CT examination for their mTBI were included from seven emergency departments in Stockholm, Sweden between 2015–2020. A total of around 70 variables available from the emergency department were collected. The data analysis of the first 5000 patients in the cohort has been performed to test a preliminary model for the detection of ICL.

Results: The preliminary analysis of interim data produced a model using 10 variables and provided an AUC of 0.68, outperforming four other currently available decision algorithms in our cohort of mTBI patients.

Conclusion: The preliminary result from our study demonstrates the potential for a data-driven approach to generate a new personalized risk estimation tool to guide clinicians in the management of mTBI patients. The final retrospective cohort is expected to contain 30000 patients and will allow us to generate further optimized models for the outcomes ICL and neurosurgical intervention.

D085 Changes in brain structure and age in Veterans with TBIs following treatment with Magnesium-Ibogaine

Mr. Andrew Geoly,¹ Dr. John Coetzee ,^1,2 Dr. Wiebke Struckman,¹ Dr. Derrick Buchanan,¹ Azeezat Azeez^{1, 2}, Dr. Bora Kim,¹ Dr. Kirsten Cherian,¹ Dr. Nimrod Keynan,¹ Dr. Maheen Adamson,^1,2 Dr. Nolan Williams¹

¹Stanford University, Stanford, United States, ²VA Palo Alto Health Care System, Palo Alto, United States

Introduction: TBI is common among Veterans of recent conflicts, and may lead to a range of symptoms, as well as accelerated brain aging.

Ibogaine, a psychoactive alkaloid, has neuroplasticity-promoting properties. It may help remodel neural circuitry and improve functioning in Veterans with TBI.

Methods: We conducted an observational study with 30 Veterans with multiple blast TBI (mbTBI) and complex clinical problems who received ibogaine treatment, preceded and followed by preparation and integration. At baseline, immediate post, and 1-month, we performed clinical assessments and structural MRI scans. We derived cortical thickness (CT) measures with the ANTs longitudinal CT pipeline and evaluated CT and volume in cortical and subcortical gray matter, and in cerebellar ROIs. To evaluate longitudinal changes in CT across ROIs, we used linear mixed effects (LME) models. We used the algorithm brainageR to measure brain age.

Results: A Wald Χ2 test of regional LME models revealed a significant (pFDR<0.05) effect of study visit on CT in 13 ROIs. Pairwise t-tests demonstrated significant (pholm<0.05) increases in CT following ibogaine relative to baseline visit in 11 regions. For subcortical volume, Wald Χ2 test of the subcortical LME models revealed a significant (pFDR<0.05) main effect on the log-jacobian determinant in the Right Ventral Diencephalon. Wald Χ2 test of the LMEs revealed a significant change in brain age across time points [Χ2(2)=10.64, p=0.0049]. Post-hoc t-tests gave a significant (pholm<0.05) reduction of 1.60 years in predicted brain age relative to baseline one month after treatment (t=3.18, p=0.0082, d=1.035).

Conclusions: This provides the first evidence of measurable brain morphometric changes in humans following ibogaine therapy. More research is needed to understand the mechanisms by which ibogaine works and to determine long-term impact on cortical structure.

D086 Childhood Maltreatment affects White Matter Microstructure independent of mild Traumatic Brain Injury and Post-traumatic Stress Disorder

Luisa Berger ,^1,2 PhD Philine Rojczyk,^1,2 PhD Johanna Seitz-Holland,^1,3 B.Sc. Twishi Puri,¹ Prof. Fan Zhang,^1,4,5 PhD LJ O’Donnell,⁴ Prof. Sylvain Bouix,^1,6 PhD Yogesh Rathi,¹ PhD Ofer Pasternak,^1,3,4 Prof. MB Stein,^7,8,9 Prof. ME Shenton,^1,4 Prof. IK Koerte^1,2,3,10

¹Psychiatry Neuroimaging Laboratory, Department of Psychiatry, Brigham and Women’s Hospital, Harvard Medical School, Somerville, USA, ²cBRAIN, Department of Child and Adolescent Psychiatry, Psychosomatic and Psychotherapy, Ludwig-Maximilians-Universität, Munich, Germany, ³Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School, Boston, USA, ⁴Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, USA, ⁵University of Electronic Science and Technology of China, Chengdu, China, ⁶Department of Software Engineering and IT, École de Technologie Supérieure, Montreal, Canada, ⁷Department of Psychiatry, University of California San Diego, La Jolla, USA, ⁸School of Public Health, University of California San Diego, La Jolla, USA, ⁹Psychiatry Service, VA San Diego Healthcare System, San Diego, USA, ¹⁰Graduate School of Systemic Neuroscience, Ludwig-Maximilians-Universität, Planegg, Germany

Introduction: Mild traumatic brain injury (mTBI) and post-traumatic stress disorder (PTSD) are the signature wounds of military service and have been linked to alterations in brain white matter microstructure. Childhood maltreatment is common and may predispose to psychopathology. Childhood maltreatment has been associated with alterations in white matter microstructure in adulthood. This study investigated the effect of mTBI and/or PTSD on the relationship between childhood maltreatment and adult white matter alterations.

Methods: Clinical and diffusion-weighted magnetic resonance imaging data were collected from 299 participants of the Injury and Traumatic Stress (INTRuST) study, including individuals with and without mTBI and/or PTSD. Regression models were used to test associations between the severity of childhood maltreatment and white matter microstructure characteristics (Mean diffusivity, MD; Axial diffusivity, AD; Radial diffusivity RD; Fractional anisotropy, FA) of major white matter fiber tracts (Arcuate Fasciculus, Cingulum Bundle, Inferior Longitudinal Fasciculus, Inferior Occipito-frontal Fasciculus, Superior Longitudinal Fasciculus, Uncinate Fasciculus, and Corpus Callosum). Moderation models were employed to assess the impact of mTBI and/or PTSD on these associations. All statistical analyses were conducted using R and controlled for age, gender, and race.

Results: Higher childhood maltreatment severity was associated with lower MD (pFDR < .05) and lower AD (pFDR < .05) of major white matter fiber tracts. Notably, these associations were independent of a diagnosis of mTBI (all pFDR > .05), PTSD (all pFDR > .05), or comorbid mTBI+PTSD (all pFDR > .05).

Conclusions: Results from this study suggest that the association between childhood maltreatment and white matter microstructure is more attributable to childhood maltreatment than to mTBI and/or PTSD diagnoses. These findings highlight the importance of considering childhood maltreatment as a contributing factor when examining mTBI and PTSD in military personnel.

D087 The Role of ICP Monitoring and Decompressive Craniectomy in War-related TBI

Dr Yevgeny Karepov ,¹ Dr Jane Skidan,¹ Dr Mohamad Hmaeed Asdi,¹ Dr Orit Lesman-Segev,^2,3,7 Dr Avital Perry,^2,4 Dr Yael Rosen Lang,² Dr Anton Peled,⁴ Prof Israel Melamed,⁶ Dr Elad Avraham,⁶ Dr Amit Azriel,⁶ Dr Tomer Talmy,^8,9 Dr Saadit Sarah Houri,⁵ Dr Raquel C. Gardner,² Prof Guy Rosenthal⁵

¹Department of Neurosurgery, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel, ²Sagol Neuroscience Center, Sheba Medical Center, Ramat Gan, Israel, ³Faculty of Medical and Health Science, Tel Aviv University, Tel Aviv, Israel, ⁴Department of Neurosurgery, Sheba Medical Center, Ramat Gan, Israel, ⁵Department of Neurosurgery, Hadassah Medical Center, Hebrew University of Jerusalem, Jerusalem, Israel, ⁶Department of Neurosurgery, Soroka Medical Center, Beersheba, Israel, ⁷Department of Diagnostic Imaging, Sheba Medical Center, Ramat Gan, Israel, ⁸Division of Anesthesia, Intensive Care & Pain Management, Tel-Aviv Sourasky Medical Center, Tel Aviv, Israel, ⁹Israel Defense Forces Medical Corps, Tel Hashomer, Ramat Gan, Israel

Introduction: Traumatic brain injury (TBI) sustained in combat requires a different approach than civilian TBI. This study reports acute features, neurosurgical management, including use of intracranial pressure (ICP) monitoring and decompressive craniectomy (DC), among patients with war-related TBI (WrTBI).

Methodology: We established a 4-Hospital WrTBI Registry that is collecting common data elements (CDEs) on WrTBI in Israel. The current report includes consecutive WrTBI patients, with evidence of intracranial trauma on CT, admitted to one of the participating centers (Tel Aviv Sourasky Medical Center) between 7/10/2023–1/4/2024. Demographics, injury characteristics, neurosurgical interventions, and postoperative complications were recorded. Glasgow Outcome Scale-Extended (GOS-E) score at hospital discharge was estimated via chart review.

Results: Among 23 patients meeting inclusion criteria, 16 (70%) were male, median age was 32y (range 8–87y), 11 (48%) had GCS≤8 on arrival (of whom 9 (39%) had GCS 3). 12 (52%) sustained penetrating TBI (pTBI) with foreign body on CT, 11 (48%) sustained frontal lobe injury, and 8 (35%) sustained temporal lobe injury. Overall, 13 patients (57%) underwent neurosurgical intervention: 6 (26%) wound debridement, 3 (13%) primary DC, 4 (17%) insertion of ICP monitor, 2 (9%) secondary DC due to intractable ICP. 4 (17%) had CNS infection during hospitalization; 3 (13%) had systemic infection. In-hospital mortality was 13% overall (27% among pTBI GCS≤8). Estimated discharge GOS-E was >4 in 74% (45% among pTBI GCS≤8). All patients that died during acute hospitalization had GCS 3 on arrival with at least 1 dilated non-reactive pupil.

Conclusions: The experience of our single center highlights the complexity of WrTBI, and the important role of ICP monitoring and DC in management of WrTBI. Data collection, entry, and pooling is ongoing from all four hospitals to allow more comprehensive results and comparative effectiveness studies.

D088 “Melting” brain vs. infection: a longitudinal imaging study of complex high-velocity penetrating traumatic brain injury

Dr. Orit Lesman-segev ,^1,2,3 Avital Perry,^2,4 Yael Rosen Lang,² Anton Peled,⁴ Saadit Sarah Houri,⁵ Guy Rosenthal,⁵ Esther Yuh,⁶ Israel Melamed,⁷ Elad Avraham,⁷ Amit Azriel,⁷ Jane Skidan,⁸ Yevgeny Karepov,⁸ Mohamad Hmaeed Asdi,⁸ Tomer Talmy,^10,11 Adi Gidali,^2,12 Dafna Yahav,^9,3 Chen Hoffman,^1,3 Raquel C. Gardner,² Gahl Greenberg^1,3

¹Department of Diagnostic Imaging, Sheba Medical Center, Ramat Gan, Israel, ²Sagol Neuroscience Center, Sheba Medical Center, Ramat Gan, Israel, ³Faculty of Medical and Health Science, Tel Aviv University, Tel Aviv, Israel, ⁴Department of Neurosurgery, Sheba Medical Center, Ramat Gan, Israel, ⁵Department of Neurosurgery, Hadassah Medical Center, Hebrew University of Jerusalem, Jerusalem, Israel, ⁶Department of Radiology, University of California San Francisco, San Francisco, USA, ⁷Department of Neurosurgery, Soroka Medical Center, Beersheba, Israel, ⁸Department of Neurosurgery, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel, ⁹Unit of Infectious Diseases, Sheba Medical Center, Ramat Gan, Israel, ¹⁰Division of Anesthesia, Intensive Care & Pain Management, Tel-Aviv Sourasky Medical Center, Tel Aviv, Israel, ¹¹Israel Defense Forces Medical Corps, Tel Hashomer, Ramat Gan, Israel, ¹²Department of Physical Therapy, Sheba Medical Center, Ramat Gan, Israel

Introduction: Penetrating traumatic brain injury (PTBI) is associated with a high risk of infection, particularly in the presence of dura violation, air sinus penetration, or retention of contaminated objects. However, the natural history of PTBI may independently result in an intraparenchymal collection/edematous changes even in the absence of infection. Longitudinal PTBI imaging studies are needed to elucidate neuroimaging features of sterile vs. infectious sequelae of PTBI to guide management. We aim to characterize longitudinal CT features and infectious work-up/outcomes in a series of patients with PTBI who developed parenchymal ring-enhancing collection/edematous lesions, consistent with potentially infectious lesions.

Methodology: We included consecutive patients that presented to Sheba Medical Center (Oct.-Dec. 2024), following gunshot, shrapnel, and/or blast injury to the head resulting in PTBI, who developed potentially infectious intra-axial collection/edematous lesions, as observed on subsequent CTs. Two neuroradiologists, blinded to outcome, visually rated baseline CT TBI Common Data Elements (CDEs; e.g., contusion, hemorrhage, fractures, shift, etc.), additional features specifically pertinent to PTBI (metal/non-metal foreign material, entry/exit wounds, trajectory, vascular injury, etc.), and features of the evolving intra-axial lesions longitudinally.

Results: Of 24 patients admitted with PTBI, six patients met inclusion criteria and were followed up radiographically for at least 4 months (mean age 32y [range 21–46y], all males, 1 civilian, 5 soldiers). Peripheral enhancement was first observed on average on day 21 (range 9–38d) post-injury. All were in close proximity to bony defects/foreign material entry site, mostly near skull base/sinus/orbit walls, and developed in the vicinity of the foreign material.

Conclusions: We report on intra-axial lesion evolution following PTBI, presenting a clinical dilemma between infectious collection and sterile liquidation. In our study, 25% of patients with PTBI developed lesions concerning for brain abscess. Analyses are ongoing. More detailed imaging characteristics, infectious work-up results, and clinical outcomes will be available for the conference.

D089 Forced polarisation of microglia by IL-13 is modified by microenvironmental context

Ms Emmanuelle Aiyegbusi ,^1,2 Dr James P. Reynolds,^1,2 Dr Dearbhaile Dooley^1,2

¹School of Medicine, University College Dublin, Belfield, Ireland, ²Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Ireland

Introduction: Traumatic spinal cord injury (SCI) is a severe clinical challenge involving a primary physical damage phase, and an inflammatory secondary phase driven by microglia and other infiltrating immune cells. Immunomodulatory therapies may help promote healing and restrict secondary damage.

Methodology: We have previously demonstrated that interleukin (IL)-13 delivery improves functional and histopathological recovery after SCI in murine models primarily by polarising macrophages towards an alternatively activated pro-reparative M2-like phenotype and reducing axonal contacts. Although microglia respond robustly to IL-13 in vitro, polarisation of microglia in vivo is more difficult.

To better understand what conditions may restrict microglial responses to IL-13 in vivo, we sought to examine the effect of cellular context or microenvironment on IL-13 efficacy in forcing microglia polarisation in vitro. : Transcriptional changes (RT-qPCR) and cytokine release (ELISA) were examined in BV2 microglia following IL-13 treatment. High content screening was used to quantify expression of anti-inflammatory markers in response to various immunomodulatory stimuli.

Results: IL-13 leads to increased expression of the anti-inflammatory marker arginase-1 while lowering expression and secretion of the pro-inflammatory markers IL-1β, iNOS, and TNFα, signifying effective polarisation of microglia. Concomitant administration of lipopolysaccharide (LPS) with IL-13 reduces IL-13 polarisation efficacy in BV2 cells, suggesting that IL-13 efficacy is reduced in inflammatory contexts. Other microenvironmental conditions, including excess glutamate, extracellular matrix proteins, acidosis, and hyperkalaemia were also examined, revealing key constraints that act to limit IL-13 efficacy in forcing microglial polarisation. A combination of TGFβ1, TGFβ2, cholesterol, and M-CSF was found to boost IL-13′s polarising efficacy on microglia.

Conclusion: IL-13 polarising efficacy is limited by factors within the cellular microenvironment. However preliminary results indicate that the addition of certain cytokines and growth factors boosts IL-13 activity on microglia under inflammatory conditions.

D090 Modelling stem cell viability in neurotrauma: a novel ex vivo strategy to understand neural crest-derived stem cell efficacy in simulated multicellular microenvironments

Sandra Jenkner ,^1,2 Professor Jillian Clark,^2,3 Professor Stan Gronthos,^1,4 Dr Ryan O’Hare Doig²

¹School of Biomedicine, The University Of Adelaide, Adelaide, Australia, ²Neil Sachse Centre for Spinal Cord Research, South Australian Health and Medical Research Institute, Adelaide, Australia, ³Centre for Orthopaedic and Trauma Research, Adelaide Medical School, The University of Adelaide, Adelaide, Australia, ⁴Mesenchymal Stem Cell Laboratory, South Australian Health and Medical Research Institute, Adelaide, Australia

Introduction: Stem cell survival rates of <3% post-transplantation into the central nervous system (CNS) hinder the clinical translation of stem cell therapeutics. This study attempts to unravel the mechanisms of this low viability. Ex vivo pre-conditioning of stem cells with inflammatory cytokines can significantly improve stem cell efficacy. We also have demonstrated that pre-conditioning of neural crest-derived dental pulp stem cells (DPSC) with peripheral blood mononuclear cells (PBMC) improves stem cell viability. Whether this is maintained within prohibitive multicellular neurotrauma-like stress microenvironments is unresolved.

Methodology: Human donor DPSC were co-cultured ex vivo with naïve PBMC (n=6) and/or recombinant cytokines (TNFα, IFNγ and/or IL17A), and exposed to cellular (activated PBMC) or acellular (no PBMC) stressors [endotoxic (LPS), excitotoxic (glutamate), hypoxic (cobalt chloride), oxidative stress (H2O2) or combined]. Lactate dehydrogenase assays measured DPSC cytotoxicity and viability. Flow cytometry and qPCR analysed viability and functional phenotypes of DPSC and PBMC.

Results: DPSC showed high resilience to acellular stressors. In contrast, inflammatory cytokines and/or activated PBMC significantly reduced DPSC viability (10–70% vs 97% unstressed control; p<0.0001) and increased cytotoxicity (46–60% vs 15% unstressed control; p<0.0001), particularly under hypoxic and combined stressors. Microenvironments resulting in poor DPSC viability were associated with higher proportions of cytotoxic T cells, NK cells, dendritic cells and classical monocytes, while non-classical monocytes were depleted. Poor DPSC viability correlated with a downregulation of stem cell immunomodulatory, cell-to-cell interaction and stemness markers, including IDO1, CD73 and NFkB (R2 = 0.6).

Conclusions: Routine ex vivo pre-conditioning protocols may compromise stem cell survival and therapeutic efficacy in the context of multicellular neurotrauma-like microenvironments. Current investigations in our lab are expanding on these findings using activated PBMC derived from spinal cord injured donors, known to have distinct inflammatory phenotypes. Immunomodulatory approaches may enhance stem cell viability, efficacy and clinical translation.

D091 The Clinically Used Kappa Opioid Receptor Agonist Nalfurafine Improves Inflammation In A Rodent Model Of Spinal Cord Injury

Mr Kevin Roy,¹ Miss Barbara Fackelmeier,¹ Dr Bronwyn Kivell,² Dr Simon O’Carroll ¹

¹Department of Anatomy and Medical Imaging, Faculty of Medical and Heath Sciences, University of Auckland, Auckland, New Zealand, ²School of Biological Sciences, Victoria University of Wellington, Wellington, New Zealand

Introduction: Previous studies have shown that activation of the kappa opioid receptor (KOR) may be detrimental following spinal cord injury. The KOR agonist dynorphin increases damage after injury however dynorphin can activate both GPCR pathways, which are involved in normal cell function and neuroprotection, and β-arrestin pathway and p38 pathways that lead to cell death, which is a major contributor to SCI injury and inflammation. KOR such as nalfurafine reduce inflammation, modify the immune response and promote remyelination in multiple sclerosis and this is believed to be through its bias for activation of the activation of GPCR pathways as opposed to β-arrestin pathway and p38 activation. Therefore, we are testing the G-protein biased KOR nalfurafine for its ability to improve inflammation, improve remyelination and modify immune cell infiltration after SCI.

Methods: Sprague-Dawley rats were given a moderate (175 kDyne) contusion spinal cord injury at T10 using the Infinite Horizons impactor. Animals we randomly divided into 3 groups (n = 8/group) and received daily i.p injections of vehicle, 0.03 or 0.1 mg/kg nalfurafine daily for 4 weeks. Animals underwent open field testing to measure hind limb function and co-ordination using the Basso-Beattie-Bresnahan (BBB) rating scale and tissue was collected at the end of the study for immunohistochemical analysis.

Results: Treatment with 0.03 mg/kg showed a significant decrease in total and M1 microglia compared to vehicle (one-way ANOVA, Bonferroni’s post-hoc test, p < 0.05). No significant difference in M2 microglia was seen between the groups. No difference in astrocytes was seen between the groups. Although not powered to determine a difference in the BBB score, 0.03 mg/kg treated animals showed improved hind-limb function at 4 weeks (control, 10.4 ± 0.64 vs nalfurafine, 12.2 ± 0.28). This preliminary study suggests that treatment with nalfurafine may be a potential treatment for spinal cord injury.

D092 Can the activation of angiotensin II receptor type 2 potentiate the diminished intrinsic axonal regenerative capacity?

Dr. Jaroslav Pavel ,¹ Dr. Jana Snopkova,¹ Dr. Veronika Liptakova,¹ MSc. Erika Hatalova¹

¹Institute of Neurobiology, Biomedical Research Center Of The Slovak Academy Of Sciences, Kosice, Slovakia

Introduction: Spontaneous regeneration in the mature CNS is very limited. It is evident that a hostile environment is not only a critical determinant of axonal regeneration failure but also that other signalling pathways are involved. The AT2 receptor, an integrative part of the “protective arm” of the renin–angiotensin system, is abundant during the prenatal period. However, its expression gradually decreases with postnatal age, and it is expressed only in discrete areas in adults. Interestingly, it is upregulated after pathological conditions, including CNS injuries. Time-dependent receptor expression was studied after severe spinal cord injury to determine the appropriate timing of pharmacological intervention.

Methodology: Severe spinal cord compression (40 g for 15 min) at the Th9 spinal level was induced in female adult Wistar rats. The injured spinal cord was studied within a four-week survival period. The AT2 receptors were continually stimulated using the selective ligand CGP42112 or blocked with PD123319 via osmotic minipumps. Locomotor activity, bladder function, body weight, transcranial motor evoked potentials, spinal cord histopathology and the gene and protein expression of the AT2 receptor, GAP-43, neurofilaments, MBP, CNPase, PLP1, and Olig2 were analysed.

Results: A delayed and transient increase in receptor expression was observed from the 14th to 21st day postinjury. Surprisingly, it normalized almost to that of the control one week later. The period of receptor upregulation was selected as the most accurate interval for AT2 receptor stimulation. Improvements in monitored functional parameters, increased spinal cord tissue preservation, a decreased number of cysts and cavities, and increased axonal outgrowth and major structural components were detected. These beneficial effects were prevented by the administration of an AT2 receptor antagonist.

Conclusions: We can conclude that the stimulation of late-expressed AT2 receptors improved functional parameters and potentiated diminished intrinsic axonal regeneration after spinal cord injury.

Supported by APVV-22-0248, VEGA No. 2/0123/23.

D093 Hyperspectral imaging successfully detects tissue damage in a spinal cord injury rat model

Obada Alhalabi,¹ Dr. Raban Heller,^2,3,4 Dr. Alexander Studier-Fischer,⁵ Maria Maria Maares,³ Dr. Hajo Haase,³ Lutz Schomburg,² Prof. Arash Moghaddam,⁶ Dr. Bahram Biglari,⁷ Prof. Sandro M. Krieg,¹ Pd Dr. Med. Alexander Younsi ¹

¹Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Germany, ²Institute for Experimental Endocrinology, Charité—Universiẗ atsmedizin Berlin, corporate member of Freie Universitaet Berlin, Humboldt-Universitaet zu Berlin, Berlin Institute of Health, Berlin, Germany, ³Department of Food Chemistry and Toxicology, Technische Universitaet Berlin, Berlin, Germany, ⁴Bundeswehr Hospital Berlin, Department of Traumatology and Orthopaedics, Septic and Reconstructive Surgery, Berlin, Germany, ⁵Department of General, Visceral and Transplantation Surgery, Heidelberg University Hospital, Heidelberg, Germany, ⁶Orthopedic and Trauma Surgery, Achaffenburg, Germany, ⁷BG Trauma Centre Ludwigshafen, Department of Paraplegiology, Heidelberg, Germany

Introduction: In situ evaluation of the extent of spinal cord injury (SCI) could provide a further modality for intra- and perioperative decision-making. Hyperspectral Imaging (HSI) is a novel, non-invasive tool that measures variations in tissue oxygenation and perfusion as a possible characterization surrogate for tissue injury, indicative of underlying neural damage. This proof-of-concept study aimed to determine the efficacy of HSI in informing on the extent of tissue injury after SCI in a rat model.

Methodology: After T9/10-laminectomy, the TIVITA® Tissue system (HSI camera) was used to acquire spectral data before and after experimentally induced clip-compression/contusion SCI in n=6 Wistar rats. The imaging process involved acquiring data across a broad wavelength range mostly around the visible to near-infrared spectrum. Animals were sacrificed 14 days after trauma and HSI findings were correlated with conventional immunohistochemistry analyses of explanted and cryosectioned spinal cords, serum cytokine data and functional outcomes (using an open-field locomotor scale and the CatWalk XT® computer-based gait analysis).

Results: Global principal component analyses (PCA) of multidimensional HSI data revealed distinct clustering between injured and non-injured spinal cord tissues based on mere spectral signatures. Moreover, quantitative analysis of HSI data established a robust correlation with injury extent 14 days after SCI, as informed by cyst size in isolated spinal cord tissue, pro-inflammatory serum cytokine levels at day 14 after SCI, and endpoint neurobehavioral parameters pertaining to locomotion and coordination of hind limbs.

Conclusions: HSI-derived data successfully identified tissue injury in situ after SCI in a rat model. Its ability to provide detailed, post-injury insights into the perfusion state of spinal cord tissue holds promising potential for both animal studies and future clinical applications. Further data to better characterise the prognostic value of HSI in the context of neurotrauma are still needed.

D094 Traumatic brain injury in UK road traffic collisions: the effect of collision dynamics and age on injury

Dr Claire Baker,¹ Dr Phil Martin,² Prof Mark Wilson,¹ Dr Lucia Li,¹ Dr Mazdak Ghajari ,¹ Prof David Sharp¹

¹Imperial College London, London, United Kingdom, ²Transport Research Laboratory (TRL), Wokingham, UK

Introduction: Road traffic collisions (RTC) are a leading cause of global morbidity and often cause traumatic brain injury (TBI) [1]. Age and collision dynamics separately influence TBI outcomes, but their interaction is unstudied. UK Road Accident In-depth Studies (RAIDS) data uniquely captures collision dynamics and clinical data. We previously showed that change-in-velocity during impact (“delta-V”) is a key injury determinant. Here we examine age-dynamics influence on TBI prevalence/type.

Methods: 5743 subjects were investigated (2013–2024 RAIDS data), with 971 vehicle occupants and 147 pedestrians with known age/speed information. Mayo TBI was classified as previously described [2]. Injury rates were studied in three age groups (18–30/31–69/70+yrs). TBI/age/delta-V interaction was analysed with Chi-squared and Mann-Whitney statistics.

Results: 3345 injured subjects (58.1%) included 291 fatalities (8.7%). 669 TBI subjects (68.7% of fatalities, 21.9% of injured subjects) included 365 (54.6%) moderate-severe, 152 (22.7%) mild and 152 (22.7%) symptomatic and 197, (29.4%) skull fracture, 195 (29.1%) subarachnoid haemorrhage (SAH), 114 (17.0%) subdural haematoma (SDH) and 29 (4.3%) axonal injury.

TBI rate was higher in younger (18–30yrs) and older (70+yrs) adults compared to 31–69-year-olds. Intracerebral haemorrhages (SAH/SDH) but not skull fracture were more prevalent for 70+yrs vs. other groups (no difference in younger groups).

TBI differences were not driven by higher RTC dynamics in cars or pedestrians. Average dynamics did not differ significantly, although they were generally numerically higher in younger vs. older adults. In car subjects, TBI: 40.6km/h (18–30yrs), 37.6km/h (31–69yrs), 33.6km/h (70+yrs); moderate-severe TBI: 53.4km/h (18–30yrs), 49.6km/h (31–69yrs), 38.2km/h (70+yrs). Average delta-V in fifteen 70+yrs SDH subjects was significantly lower (than 31–69yrs) despite their higher SDH prevalence. In pedestrians with TBI, post-impact speeds were 30.5km/h (18–69yrs) vs. 26.5km/h (70+yrs).

Conclusions/Discussion: Older adults are more vulnerable to TBI (especially SAH/SDH), despite being exposed to similar RTC dynamics, suggesting increased vulnerability to intracerebral haemorrhage given particular dynamics.

D095 A Retrospective Cohort Study on Early Propranolol Administration and its Effects on Mortality from Severe TBI

Dr. David Hallan, Dr. Francis Jareczek, Mr. Mason Stoltzfus, Mr. Zachary Freedman, Dr. David Bailey, Mr. Jinpyo Hong ,¹ Dr. Elias Rizk, Dr. Haejoe Park

¹Penn State Health Milton S Hershey Medical Center, Hershey, USA

Background: The role of beta blockers in severe, traumatic brain injury (TBI) management is debated. Severe TBI may elicit a surge of catecholamines, which has been associated with increased morbidity and mortality. We hypothesize that catecholamine levels within 48 hours of TBI can be used to predict patient mortality within 30 days of injury, and leveraged the TriNetX database to determine whether administering a beta- adrenergic blocker within 48 hours of severe TBI improves 30-day outcomes.

Methods: The TriNetX Research Network was used to form two cohorts using retrospective data from 106,294,356 patient profiles from 9/10/2022. Cohort 1 included all patients who received the first-instance diagnosis of severe TBI (GCS 3–8) and propranolol within 48 hours of injury. Cohort 2 included all patients with the same diagnosis of severe TBI but did not receive betablockers. The primary outcome of interest was mortality at 30 days. Secondary outcomes included gastrostomy tube placement, neurosurgical intervention in the form of craniotomy, craniectomy, burr hole drainage, seizure, cardiac arrest, bradycardia, and intensive care unit (ICU) length of stay (LOS).

Results: At 30 days post-severe TBI, 22.7% (84) of patients from the cohort that received propranolol, and 30.77% (116) from the cohort that did not, were deceased (OR 0.66), 95% CI [0.48, 0.92]), (p 0.01). TBI patients who received propranolol were also less likely to require neurosurgical intervention, experience seizures, cardiac arrest, and bradycardia, but were more likely to require longer ICU length of stay.

Conclusion: The results of this study demonstrate significantly reduced mortality at 30 days and fewer neurosurgical interventions, seizures, episodes of cardiac arrest, and bradycardia in severe TBI patients who received propranolol within 48 hours of injury.

D096 Head Injury and Single and Dual Task Gait in Older Adults in the Atherosclerosis Risk in Communities (ARIC) Cohort

Dr. Katie Hunzinger ,¹ Dr. Laura Skow,² Mr. Connor Law,³ Dr. Alexa Walter,³ Dr. Priya Palta,⁴ Dr. Pamela Lutsey,⁵ Dr. Gwen Windham,⁶ Dr. Andrea Schneider³

¹Thomas Jefferson University, Philadelphia, United States, ²National Institutes of Health, Bethesda, USA, ³University of Pennsylvania-Perelman School of Medicine, Philadelphia, USA, ⁴University of North Carolina at Chapel Hill, Chapel Hill, USA, ⁵University of Minnesota, Minneapolis, USA, ⁶University of Mississippi Medical Center-the MIND Center, Jackson, USA

Introduction: Older adults have the highest rates of head injury, yet the long-term effects of prior head injury on late life gait are unknown. We examined associations of prior head injury with mobility using single task (ST) and dual task (DT) gait in a cohort of community-dwelling older adults.

Methodology: 336 participants in the Atherosclerosis Risk in Communities (ARIC) Study completed 2 trials of ST and 2 trials of DT gait using the ProtoKinetics Zeno Walkway in 2020. Prior head injury was defined by self-report and ICD-9/10 codes. Gait data (mean velocity, mean stride length, mean stride width) from both ST and DT conditions were used, as well as the DT cost (DTC), which represents the percent difference between undivided (ST) and divided (DT) attention conditions for each metric. Linear regression models adjusted for age, sex, race, education, and body mass index were used to estimate associations between head injury and each gait metric.

Results: Overall, participants were a mean age of 81.6 years, 33.1% had a history of head injury, 51.2% were female, 16.3% were of self-reported Black race. Prior head injury was associated with slower gait speeds (ST: -0.077m/s [95% CI: -0.120, -0.034], DT: -0.078m/s [95% CI: -0.129, -0.027]), shorter stride lengths (ST: -0.071m/s [95% CI: -0.109, -0.032], DT: -0.079m [95% CI: -0.122, -0.037]), but similar stride widths across conditions (ST: 0.002m [95% CI: -0.006, 0.011]); DT: 0.002m [95% CI: -0.008, 0.012]). The percent difference between ST and DT conditions by head injury status was similar (p>0.05 for all DTC metrics).

Conclusions: Older adults with a history of head injury exhibit reduced gait performance in ST and DT conditions compared to those without a history of head injury, suggesting that prior head injury may have lasting effects on gait into late life.

D097 Exploiting ectopic claudin-1 expression for targeted delivery to injured and aging brain

Dr. Saiprasad Gowrikumar,² Dr. Aria Tarudji,¹ Dr. Punita Dhawan,² Dr. Forrest Kievit,¹ Brandon McDonald ²

¹University Of Nebraska, Lincoln, United States, ²University of Nebraska Medical Center, Omaha, United States

Introduction: Because of a current lack of treatment options, there is a critical need to develop active targeting strategies to gain access specifically into diseased brain regions. The structural integrity of the BBB governing its tightness is linked to the correct regulation of tight junction (TJ) proteins. Dysregulation of claudins and other TJ proteins that lead to disruptions in the BBB are likely culprits in progressive, secondary consequences of TBI. CLDN5 and occludin constitute the major TJs at the BBB whereas CLDN1 is minimally present. Here we sought to identify a role of CLDN1 in TBI and aging brain in order to establish new active targeting strategies for improving delivery into specific brain regions that are experiencing dysfunction.

Methodology: We utilized mouse models of severe (CCI) to mild TBI (mCCI) and aging (12 months old) as well as in vitro exposure of brain endothelial cells to gain mechanistic insight into alterations observed at the BBB. In addition, CLDN1 targeting nanoparticles were synthesized and tested for CLDN1 binding in vitro and in our in vivo models using biophysical and medical imaging approaches.

Results: We found that CLDN1 was significantly increased in brain endothelium in our mouse models and in vitro. The observed increase of CLDN1 expression correlated with down-regulation of CLDN5 and occludin, thereby altering BBB integrity by decreasing TEER and increasing permeability. Knockdown of CLDN1 showed stability of the endothelial junctional proteins indicating a direct role of CLDN1 in BBB dysfunction. Nanoparticle-mediated targeting of CLDN1 in TBI and aging brain showed enhanced accumulation specifically in regions with increased CLDN1 expression.

Conclusions: Our findings provide new insights into BBB deregulation and new delivery opportunities for TBI and aging. In addition, inhibition of CLDN1 and its mediators may be a promising approach to help curb progression of disease mediated by BBB dysfunction.

D098 The preventive measures of traumatic brain injury in the elderly

Associate Professor Takeshi Maeda ,^1,3,4 Kazuko Kamiya,² Instructor Ryo Otaki,¹ Adjunct Assistant Profesor Takahiro Kumagawa,¹ Profesor Yoichi Katayama,^1,3 Profesor Atsuo Yoshino¹

¹Department of Neurological Surgery, Nihon University School of Medicine, Tokyo, Japan, ²Japan Medical Association Research Institute, Tokyo, Japan, ³Center for Brain and Health Sciences, Aomori University, Aomori, Japan, ⁴Anesthesiology, Nihon University Hospital, Tokyo, Japan

In rapidly ageing society, head injuries among the elderly are on the increase and preventive measures are urgently needed. Prognosis in the most recent data showed that there was a significant increase in falls and falls injuries with older age (P<0.05). In this study, the current situation of CSDH was analyzed and the involvement of drugs with side effects such as “muscle weakness and dizziness” in CSDH was investigated, as prevention of falls is an effective way to prevent head injury in the elderly. The target population was 36375 surgical cases of CSDH extracted from the National data base in Japan. The number of people with CSDH per 100,000 population increased with age, peaking at 80–84 years and 67% male. The proportion of drug users to the estimated population was 13.5% for hypnotic sedatives and anxiolytics, 7.1% for antithrombotic drugs, 1.8% for antiepileptic drugs and 5.1% for alpha-1 receptor blockers used to treat enlarged prostate. The proportion of users with CSDH was 39.2%, 30.8%, 7.5% and 27.1% respectively, with a significant increase in CSDH for all drugs (p<0.0001). Especially, 50% of CSDH aged 75 years and over were found to be taking oral hypnotic sedatives/anxiolytics. In another series, a lower Body Mass Index was associated with a significantly lower Modified Rankin Scale (p<0.05). As CSDH is the leading cause of head injury in the elderly, it is incidence reflects the frequency of head injury among the elderly. The higher the frequency of head contusions, the higher the frequency of serious injuries among them. We must be aware that drugs commonly taken by the elderly include drugs that increase head bruising. From the perspective of preventing head injuries, medication for the elderly should be used appropriately, considering the overall quality of life.

D099 Development of the first cross-disciplinary guideline for the care of chronic subdural haematoma (cSDH)

Dr Daniel Stubbs ,¹ Dr Benjamin Davies,¹ Dr Ellie Edlmann,² Prof Mary Dixon-Woods,¹ Prof David Menon,¹ Prof Peter Hutchinson¹

¹University Of Cambridge, Cambridge, United Kingdom, ²University of Plymouth, Plymouth, United Kingdom

Introduction: cSDH is a common neurosurgical condition in older patients, who are often frail, medically complex, and require significant perioperative input. No guideline exists to define best practice. We report on the co-design of multidisciplinary guidelines for cSDH and ongoing work towards implementation.

Methodology: Following system mapping and stakeholder identification, we formed a multidisciplinary steering group with patient representation. Based on facilitated discussion and literature review, 79 draft statements were formed and consensus-built through a 2-round online modified Delphi of the professional community (June 2023), with a threshold of 66% agreement for inclusion of statements and use of content analysis of free-text comments to refine wording and content. Final inclusion decisions were made at an in-person consensus meeting including patients. The resulting guideline was submitted to professional societies for endorsement. Emerging work is examining influences on implementation health economics considerations.

Results: 139 participants, from 12 different disciplines participated in the Delphi. Of 79 statements reviewed in round 1, 76 met the pre-specified inclusion threshold. 3 statements on middle meningeal artery embolization, referral criteria, and surgical indications were reviewed in the second round, which involved 88 participants (55% retention rate). Following content analysis of >1000 free-text comments, 18 statements were refined and discussed at our consensus meeting. A final guideline with 67 statements was then formed. It has been reviewed and endorsed by UK societies for neurosurgery, neuroanaesthesia and critical care, geriatric medicine, and perioperative care. In May 2024 we will launch an online survey examining influences on implementation and will conclude baseline health economic analyses.

Conclusion: Multidisciplinary guidelines for cSDH have been developed through co-design methodology. A roadmap to implementation and the establishment of national audit infrastructure is now planned. Our methods are transferable. The findings are of relevance to the care of this and similar complex neurotrauma conditions.

D100 Impacts of DAPT (dual antiplatelet therapy) on geriatric traumatic brain injury patients

Professor Eiichi Suehiro ,¹ Dr Ryohei Sashida,¹ Associate Professor Tatsuya Tanaka,¹ Professor Akira Matsuno¹

¹International University of Health and Welfare, School of Medicine, Narita, Japan

Objective: In Japan, the aging of traumatic brain injury (TBI) patients is becoming noticeable. In such a situation, DAPT (dual antiplatelet therapy) has become popular among antiplatelet therapies. On the other hand, there are concerns about an increase in bleeding complications. This study investigated the impacts of DAPT compared to SAPT (single antiplatelet therapy) in geriatric TBI patients.

Methods: A prospective, multicenter, observational study was conducted to monitor the use of antithrombotic and reversal drugs in TBI patients (n=721). As a secondary analysis, we examined the impacts of DAPT on TBI. 132 patients taking only antiplatelet drugs were selected. We investigated the factors that influence the severity of TBI patients. We divided patients with GCS 13–15 on admission into a mild group (n=95) and patients with GCS 3–12 into a moderate-severe group (n=37). Subsequently, the group was divided into a group receiving SAPT (n=106) and a group receiving DAPT (n=26). GCS on admission, course during hospitalization, and outcome were compared between the two groups, respectively. Multivariate logistic regression analysis was used to identify independent predictors.

Results: When comparing the mild and moderate-severe groups, the rate of DAPT was significantly higher in the moderate-severe group (14.7% vs. 32.4%). DAPT was the only factor that had a significant effect on severity on admission.The DAPT group had a significantly higher severity on admission, and had a higher frequency of brain herniation findings on head CT (21.7% vs. 46.2%), resulting in a significantly higher mortality rate (12.3% vs 30.8%). The only significant factor influencing mortality was severity on admission.

Conclusion: Compared to SAPT, patients with TBI undergoing DAPT have worse severity on admission. The severity of the injury on admission influences the outcome six months after the injury.

D101 The effect of age in patients years after acute acquired brain injury: the BRAIN-ReADAPT effect

Msc Nikki Thüss ,¹ dr Heleen den Hertog,² prof dr Coen van Bennekom,³ prof dr Jacoba Spikman,¹ prof dr Joukje van der Naalt¹

¹University Medical Center Groningen, Groningen, Netherlands, ²Isala Hospital, Zwolle, Netherlands, ³Heliomare Rehabilitation Center, Wijk aan Zee, Netherlands

Objectives: Acute Acquired Brain Injury (ABI), including traumatic brain injury and stroke, is highly prevalent with increasing numbers due to higher life expectancy. Both acute ABI and aging can cause cognitive decline resulting in decreased participation levels, which we refer to as: the BRAIN-ReADAPT effect. This study aims to describe characteristics of a large cohort of acute ABI patients more than 5 years after injury focusing on changes in participation level, considering age and injury chronicity.

Methodology: 541 acute ABI patients aged 50 to 68 years (M=60.4, SD=4.87) participated in a retrospective cohort study (the BRAIN-ReADAPT study). Questionnaires were administered 5 to 30 years after injury (M=10.7, SD=4,4). Participation was assessed by the Utrecht Scale for Evaluation of Rehabilitation-Participation (USER-P) comprising 3 subscales: frequency, restrictions and satisfaction. Fatigue, mental distress, and social support were determined with validated questionnaires. Self-rated decline in participation level (BRAIN-ReADAPT effect) was scored as a binary outcome measure: presence (BR) or absence (no-BR) of decline in participation.

Results: 43% of patients reported decline in participation level compared to their maximum participation level after injury. BR-patients, showed no difference in frequency of participation compared to no-BR-patients but experienced increased restrictions (p = 0.047) and less satisfaction with their participation level (p=0.002). BR-patients showed significantly higher scores on measures of cognitive complaints (slowness, forgetfulness and concentration problems), fatigue, emotional distress and (p<0.001). No differences were found between BR-patients and no-BR-patients for age at time of injury, gender, educational level, social support, ABI type and injury chronicity.

Conclusions: Almost half of ABI patients experience a decline in participation level more than 5 years after injury: the BRAIN-ReADAPT effect. No specific characteristics underlying this effect have been identified yet. More research is necessary, focusing on cognitive reserve, needs in this chronic phase after injury, along with potential contributing psychological factors.

D102 Traumatic Brain Injury and Changes in Plasma Biomarkers of Neurodegeneration Over 26 Years

Dr Alexa Walter ,¹ James Pike,² Dr Josef Coresh,² Dr Ramon Diaz-Arrastia,¹ Dr David Menon,³ Dr Rebecca Gottesman,⁴ Dr Priya Palta,⁵ Dr Andrea Schneider¹

¹University of Pennsylvania, Philadelphia, United States, ²New York University, New York, United States, ³University of Cambridge, Cambridge, England, ⁴National Institute of Neurological Disorders and Stroke, Bethesda, United States, ⁵University of North Carolina, Chapel Hill, Chapel Hill, United States

Introduction: Limited data exist on changes in plasma biomarkers of neurodegeneration before and after traumatic brain injury (TBI). Our objective was to characterize trajectories of plasma glial fibrillary acid protein (GFAP), neurofilament light chain (NfL), phosphorylated tau 181 (p-tau181), and amyloidβ 42/40 ratio (Aβ42/40) over 26 years of follow-up among community dwelling older adults with and without TBI.

Methodology: 1,150 participants in the Atherosclerosis Risk in Communities (ARIC) study with at least two biomarker measurements and without a history of TBI prior to the first biomarker assessment were included. Incident TBI was defined by self-report and ICD-9/10 codes. Plasma biomarkers (Aβ40, Aβ42, NfL, GFAP, and p-tau181) were measured from stored frozen samples from ARIC Visit 3 (1993–1995), Visit 5 (2011–2013), and Visit 6/7 (2016–2019) using the Quanterix HD-X. Biomarkers were log₂ transformed for analysis. Linear mixed effects models incorporating sampling weights and time splines were adjusted for age, sex, race-center, and education as time-invariant covariates, and body mass index and estimated glomerular filtration rate as time-varying covariates.

Results: Overall, participants were a mean age of 59.1 years at the time of the first biomarker measurement, 65.8% female, 28.8% Black, and 14.5% had an incident TBI over study follow-up. Compared to individuals who never had a TBI event, individuals with an incident TBI had a higher rate of log₂GFAP increase per decade both prior to (ß=0.042, 95%CI=-0.100,0.184) and after their TBI event (ß=0.143, 95%CI=-0.113,0.399), an increase of 0.101 (95%CI=-0.211,0.413) post-injury. GFAP remained elevated for over 10 years among individuals with TBI. Similar patterns were observed for NfL, whereas Aβ42/40 and p-tau181 remained elevated compared to individuals without TBI for approximately 6 years.

Conclusions: Plasma biomarkers of neurodegeneration remain elevated after TBI for at least 6–10 years. Future work will investigate patterns of biomarker changes in relation to TBI-related dementia risk.

D104 Community re-integration among traumatic brain injury survivors: the Indian scenario

Dr Priya Baby ,¹ Jayanthi Palaninathan,¹ Dwarkanath Srinivas,¹ Sivakami Subramanian¹

¹National Institute of Mental health and Neurosciences, Bangalore, India

Introduction: Traumatic brain injury can cause impairments of physical, cognitive, emotional, and psychosocial functioning of a person affecting their community re-integration. Very little is understood about the stride of recovery and community re-integration of these patients. The purpose of this study was to understand the level of community re-integration in adult chronic TBI patients in India and explore the caregiver reported psychological difficulties faced by these patients.

Methods: A cross- sectional survey was conducted among the TBI patients and their caregivers attending an Outpatient department of a tertiary hospital. Community –integration questionnaire, and Visual analog scale ranging from 0 to 10 to measure the extent of psychological issues among TBI survivors was used to for the study.

Results: Seventy TBI survivors aged 18 years to 71 years (39.3, S.D 13.5) and their caregivers were included in the study. Out of the 70 caregivers, 57.8% reported changes in the patient’s behavior and personality and 58.8% of them expressed that they witnessed mood swings in them after the injury. Anger issues in the patients were reported by 53.7% of the caregivers. The mean community re-integration score was very poor [mean 9.83 (S.D 6.1)] when the maximum possible score is 29. Home –integration, social integration and productivity scores were also significantly affected.

Conclusion: Chronic TBI survivors are poorly integrated back into the community and most of them face several psychological issues. The study points towards the need for concerted efforts that are needed towards community re-integration of TBI patients in India.

D105 TBI Prevention - A Collaborative Effort with Non-Governmental Organizations

Mr. Madhu Narayana Rao Kottakki ,¹ Dr Vijaya Sekhar Manda,² Dr. Jogi Pattisapu³

¹King George Hospital, Visakhapatnam, AP India, ²Rangaraya Medical College, Kakinada AP India, ³University of Central Florida College of Medicine

Road accidents are a primary cause of Traumatic Brain Injury (TBI), with a serious prevention challenge due to the complex social systems. Government initiatives are playing a vital role in dealing with these issues, but it is need of the hour to have a multifaceted system to facilitate effective strategies in mitigating TBI cases.

Along with local Rotary clubs and Lions Clubs, we conducted awareness sessions for over 5000 college students, working communities and public over the last 12 months. The core team of 30 volunteers observed emotional transformation and increasing support making the ongoing efforts more productive. Head injury awareness and helmet usage increased (especially in the youth), adding to the success of the campaigns. Efforts targeting school children for TBI prevention education are being developed. These positive works, current models and future strategies will be presented.

NGOs are best platforms to combat with TBI by providing education, raising awareness, advocating best environmental policies and supporting research. They also provide first aid kits, sign boards, helmets and medical aids in improving awareness for the prevention of TBI cases.

Non-governmental Organizations (NGOs) are large systems with active involvement by various groups of the society. They can be powerful allies for education, carrying prevention campaigns to several layers of society and connect with local philanthropists. Collaboration with NGOs provides a great advantage with significant outreach and effective communication/action with local authorities, municipalities for TBI prevention.

As we observed, prevention efforts supported by NGOs can positively alter TBI statistics.

D106 Public Awareness and Law Enforcement Strategies for Preventing Traumatic Brain Injuries

Alakananda Godavari1,¹ Mr. Madhu Narayana Rao Kottakki,² Dr Vijaya Sekhar Manda ,¹ Dr Padmavathi,¹ Dr. Jogi Pattisapu³

¹Rangaraya Medical College, Kakinada AP India, ²Andhra Medical College, King George Hospital, Visakhapatnam, AP India, ³University of Central Florida College of Medicine, Orlando, FL USA

Traumatic Brain Injuries (TBIs) pose a significant public health concern worldwide, with substantial socio-economic costs, especially in the developing countries and LMICs. Preventive strategies rely heavily on public awareness and law enforcement efforts and we review conditions and recent progress in the Indian sub-continent.

Effective TBI awareness campaigns and law enforcement prevention strategies in past 18 months will be presented. Multi-media outreach efforts aimed at college students, truck drivers, auto-rickshaw (3-wheeler) and helmet safety for 2-wheeler riders (including pillion/rear passenger) audiences have been completed. Individual driving habits and risk factors were analyzed, explaining the safe driving practices and high-risk behaviors.

A questionnaire given to lorry drivers and their driving habits including cell phone driving, long hours, sleeplessness, alcohol intake etc. yielded good data. Student volunteers performed several helmet studies to gather data. About 33 % of 2-wheeler drivers were wearing helmets that improved after a 2-month campaign of education and strict law enforcement. The percentage of drivers wearing helmets was 70 - 80 % when reviewed after 3 months.

Interviews also taken from law enforcement police officials provided qualitative data with ideas for improved collaboration with public. The Indian government is supporting preventive strategies, including a Road Safety week conducted in February that entails multiple activities. Special helmet drives and good Samaritan law are emphasized, along with educational videos, first aid and special programs (details will be discussed).

Current challenges and opportunities highlighting the need for multi-faceted interventions that integrate public education with targeted law enforcement efforts will be reviewed.

D107 Targeting neuroinflammation in traumatic brain injury with immunomodulatory cell therapies

Mr James Aspden ,^1,2,3 Dr Ruxandra Sîrbulescu^1,3

¹Harvard Medical School, Boston, United States, ²Edinburgh Medical School, Edinburgh, United Kingdom, ³Massachusetts General Hospital, Boston, United States

Introduction: Traumatic brain injury (TBI) remains a significant cause of morbidity and mortality. Secondary injury cascades following the initial insult play a significant role in mediating neurological dysfunction in TBI and are driven by inflammatory responses to injury. Despite significant medical need, there are currently no immunomodulatory therapeutic agents available for patients with TBI. We have previously shown the utility and efficacy of a novel B cell-based immunomodulatory therapy to promote structural and functional neuroprotection after central nervous system (CNS) injury. In this study we investigated the cellular interactions that underlie this neuroprotective effect.

Methodology: B cells and monocytes were isolated from murine splenocytes using immunomagnetic separation and cocultured overnight at specified concentrations and ratios. Subsets of cells were stimulated with specific Toll-like receptor (TLR) agonists to simulate an inflammatory response. Cellular responses were interrogated using immunofluorescence and flow cytometry.

Results: We found that co-culturing purified splenic B cells and monocytes at high local concentrations induced significant expression of regulatory cytokines (IL-10, TGF-beta) both in the B cells and in Ly6C+ classical monocytes. Inflammatory stimulation with LPS did not affect this regulatory shift in B cells. Inflammatory cytokines (IL-6, TNF-alpha, IFN-gamma) followed the same pattern, however the response amplitude was much lower. Monocytes increased the production of regulatory cytokines in the presence of B cells but showed a stronger inflammatory response to LPS. The regulatory function of B cells required IL-10 since the coculture of monocytes with IL-10-/- B cells did not modulate the monocyte inflammatory response to LPS.

Conclusions: These findings establish a foundation for the clinical development of a novel, cost-effective therapy that uses safe, readily-available cells, while addressing a major unmet medical need. Understanding the local microenvironment in TBI will be an integral step in determining whether other cells may be involved in mediating regulatory effects.

D108 Inhibition of Type I Interferon signaling alleviates Microglial inflammation following Traumatic Brain Injury in Aged Mice

Dr. James Barrett ,¹ Dr. Alan Faden,¹ Dr. Bogdan Stoica^1,2

¹Department of Anesthesiology and Shock, Trauma and Anesthesiology Research (STAR) Center, University of Maryland, Baltimore, United States, ²Veterans Affairs Maryland Health Care System, Baltimore VA Medical Center, Baltimore, United States

Introduction: Experimental evidence has demonstrated that following TBI, age is associated with worsened neurological dysfunction and exacerbated neuroinflammation[1, 2]. We have demonstrated that Type I Interferons (IFN‐I) are involved in the development of neurological dysfunction, neuroinflammation, neurodegeneration after traumatic brain injury (TBI)[3]. Our recent work showed that TBI in aged animals induces a robust neuroinflammatory response that is associated with increased expression of DNA/RNA recognition pathways and IFN‐related genes[4]. In the present study, we set out to examine the impact of IFN‐I inhibition on these age‐related responses following injury.

Methods: Young (12‐week‐old) and Aged (20‐month‐old) C57BL/6 mice underwent controlled cortical impact (CCI) or Sham surgery. Mice received intraperitoneal injections of anti‐IFNAR (IFN Receptor) monoclonal antibody (MAR1‐5A3) or appropriate Igg control at 1 hour, 24 hour and 48 hours post-injury. Animals were sacrificed at 3-, 7- and 28-days after injury; using magnetic bead isolation, CD11b+ cells were isolated from the hippocampus and cortex and prepared for RNAseq analysis. Using flow cytometry, we also assessed the impact of anti‐IFNAR on microglial function.

Results: RNAseq analysis revealed increased expression of pro‐inflammatory mediators in microglia from TBI mice and that these responses were amplified in aged mice. Pathway analysis revealed that TBI in aged mice was associated with elevated expression of viral response genes, including many IFN‐related genes. Anti‐IFNAR treatment significantly reduced the expression of these viral response genes in both young and aged animals. Significantly the age‐related increase in several key pro‐inflammatory pathways was significantly reduced in aged microglia.

Conclusions: IFN‐I signaling may prove to be a key pathophysiological mechanistic link inducing microglial dysfunction and neurodegeneration in the aged TBI brain. Inhibition of this pathway may prove to be a promising therapeutic target following TBI, especially in aged animals.

D109 Molecular and genetic overlap in TBI and dementia risk-factors

Dr Chiara de Lucia ,¹ Dr Aminul Ahmed, Prof Sandrine Thuret

¹King's College London, London, United Kingdom

Traumatic Brain Injury (TBI) presents a significant global health burden, affecting millions annually and costing billions of dollars, with potential long-term consequences including increased susceptibility to neurodegenerative diseases. Despite its prevalence and impact, effective therapies and comprehensive understanding of the molecular mechanisms underlying TBI and its subsequent complications remain elusive. Here we present results from a literature review carried out to investigate which genetic and molecular risk factors are shared between TBI and neurodegeneration. These findings form the basis of our research project aimed at elucidating the molecular and genetic underpinnings of TBI susceptibility and its association with neurodegenerative conditions using advanced genetic epidemiology approaches.

D110 Investigating succinate as a novel protective agent against metabolic dysfunction following traumatic brain injury: in vitro modelling and therapy development

Mr Cameron Hall ,¹ Dr Koby Baranes,^1,2 Mr Mark Kotter,^1,2 Dr Keri Carpenter,¹ Prof Peter Hutchinson¹

¹University of Cambridge, Cambridge, United Kingdom, ²Wellcome Trust-MRC Stem Cell Institute, University of Cambridge, Cambridge, United Kingdon

Introduction: Following traumatic brain injury (TBI), several downstream cascades are initiated, including deranged cerebral metabolism and inflammation, contributing towards secondary injury. This often manifests in patients as a high extracellular lactate/pyruvate ratio (LPR), which correlates significantly with unfavourable clinical outcome. Initial in-vitro research has shown that succinate, an intermediate of the tricarboxylic acid cycle, can protect against rotenone-induced metabolic dysfunction. Rotenone is an inhibitor of complex I of the mitochondrial electron transport chain. Succinate is a substrate for complex II, bypassing complex I injury. We further this research in human induced astrocytes (iAs) and neurons (iNs).

Methodology: Utilising the novel cellular reprogramming technique Optimised Inducible Overexpression, we produced iNs and iAs through transcription factor overexpression. These cultures were treated with varying concentrations of rotenone and disodium succinate. At select time points, the cellular metabolism and LPR were assessed using an ISCUSflex analyser and cell viability was determined using selected cellular stains. We also measured extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) in response to treatment.

Results: iAs behaved more consistently than iNs in the testing regime and produced results suggesting metabolic rescue, such that LPR was decreased, and cell viability increased following treatment, compared with non-succinate control cells. We also showed that ECAR and OCR were positively affected by succinate, in a dose-dependent manner. iNs responded to treatment with both rotenone and succinate, though not consistently. Co-culturing experiments with iAs and iNs were performed to investigate a mechanism whereby astrocytes may provide a supportive or protective role to neurons.

Conclusions: We showed that both iNs and iAs respond to rotenone treatment, confirming our metabolic injury model. Furthermore, treatment with succinate has provided evidence of metabolic rescue in iAs cultures and these studies have made progress in elucidating succinate’s potential to become a novel therapeutic agent for TBI.

D111 New Inhibitors of Quinone Reductase 2 Enhance and Accelerate Cognitive and Pathology Recovery After Traumatic Brain Injury in Mice

Dr. Sigal Liraz Zaltsman ,^1,2 Mrs. Chen Shemesh,¹ Mrs. Liora Omesi,¹ Mrs. Maria Foqara,³ Dr. Nethaniel Gould,⁷ Dr. Haneen Kayal,³ Dr. Efrat Edry,^3,4 Dr. David Last,⁵ Prof. Yael Mardor,^5,6 Prof. Kobi Rosenblum^3,4

¹The Joseph Sagol Neuroscience Center, Sheba Medical Center, Ramat Gan, Israel, ²Institutes for Health and Medical Professions, Department of Sports Therapy, Ono Academic College, Kiryat Ono, Israel, ³Sagol Department of Neurobiology, University of Haifa, Haifa, Israel, ⁴The Centre for Genetic Manipulation in the Brain, University of Haifa, Haifa, Israel, ⁵The Advanced Technology Center, Sheba Medical Center, Ramat Gan, Israel, ⁶Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel, ⁷Kavli Institute for Nanoscience Discovery, University of Oxford, Oxford, UK

Introduction: Quinone reductase 2 (QR2) is a removable memory constraint and is overexpressed during aging or Alzheimer's disease in the cortex. We described recently the QR2 pathway and developed a water-soluble, crossing the BBB, highly specific QR2 inhibitor (QR2i) that can be given directly to rodents in their daily water intake. Treating mice models of AD with the QR2i improved dramatically cognitive function and associated pathologies. Here, we aimed to evaluate the therapeutic potential of QR2i in traumatic brain injury (TBI) in mice.

Methodology: Moderate closed head injury or Sham was induced in 137 C57bl male mice using a well-calibrated weight drop model. Animals were treated with varying doses of QR2i (5, 25, 50 mg/kg) for 2 weeks and with the optimal dose for different durations post-injury (1,2, 4 weeks). Behavioral assessments were conducted over four weeks and pathological brain markers were measured. In addition, QR2 mRNA levels were measured in the injured and uninjured cortex or sham animals at different time points post-injury (0.5, 1, 3, & 7 days, 5 mice/group).

Results: QR2 mRNA expression significantly increased post-TBI, with distinct differences between the ipsilateral (2.5 folds) and contralateral cortex (1.5 folds). Treatment with QR2i, (25 or 50 but not 5 mg/kg), significantly improved behavioral outcomes, sustaining benefits up to six weeks post-injury. Two or four weeks of QR2i treatment (25 mg/kg), enhanced memory recognition, learning, short and long-term memory abilities. In addition, it reduces different measurements of pathologies in the brain.

Conclusions: Administration of 25 mg/kg QR2i in drinking water for two to four weeks post-TBI improved functional outcomes with sustained effects beyond treatment cessation. Next, we aim at developing the new QR2i as a novel therapeutic strategy for TBI.

D112 Astro but not micro gliosis is a pathological hallmark in post-traumatic epilepsy in mice

Miss Ilaria Lisi ,¹ Dr Federico Moro,¹ Dr Rossella Di Sapia,¹ Dr Annamaria Vezzani,¹ Dr Teresa Ravizza,¹ Dr Elisa R. Zanier¹

¹Mario Negri Intitute for Pharmachological research - Department of Acute Brain and Cardiovascular injury, Milano, Italy

Introduction: Post-traumatic epilepsy (PTE) is a severe neurological consequence of traumatic brain injury (TBI), with a 30-year cumulative incidence of 16.7%. Reactive astrogliosis is a common feature of TBI and epilepsy, and a key contributor to disease mechanisms; however, astrocytic alterations in PTE are not fully characterized. Here, we investigated functional and morphological changes in forebrain astrocytes by in vivo MRI and immunohistochemical analyses, in a well-characterized mouse model of TBI with ∼50% PTE incidence.

Methodology: Adult CD1 male mice underwent sham surgery (n=15) or TBI (n=30) by controlled cortical impact. Proton magnetic resonance spectroscopy (1H-MRS) was performed in the ipsilateral (il-) thalamus, before (3 weeks, w) and after (3 months, m) PTE onset. At 5m, mice were ECoG-monitored to assess PTE development, then sacrificed for histology.

Results: TBI induced alterations in astrocytic metabolism evidenced by a significant increase of myo-inositol and glutamate+glutamine in the il-thalamus at 3w (p<0.05 vs sham), as assessed by 1H-MRS. These alterations persisted up to 3m in PTE mice only (p<0.05 vs non-PTE), pointing to long-lasting astrocytic involvement in PTE. In the il-cortex and il-thalamus of PTE mice we observed an average 45% increase in both the number of GFAP+ cells and GFAP stained area vs non-PTE mice (p<0.05), denoting astrocytic activation in these areas. Notably, cortical and thalamic astrocytes in PTE mice displayed more elongated and polarized processes (cortex: p<0.05; thalamus: p<0.01 vs non-PTE), highlighting astrocytic morphological modifications in PTE. No differences in microglial activation in the il-cortex and il-thalamus were noted between PTE vs Non-PTE by IBA-1 immunohistochemistry. Efforts are ongoing to characterize at the gene and protein level the astrocytic alterations associated to PTE.

Conclusions: Our data show that astrocytic reactivity and metabolic and morphological changes occur in mice developing PTE, suggesting that astrocytic activation/dysfunction, but not microgliosis, mediate epileptogenesis after TBI.

D113 Development of a closed head impact acceleration model of traumatic brain injury in the sheep

Mr Charlie Magarey ,¹ Associate Professor John Finnie,² Professor Peter Blumbergs,² Mr Jim Manavis,² Professor Peter Cripton,³ Associate Professor Claire Jones¹

¹School of Electrical and Mechanical Engineering, University of Adelaide, Adelaide, Australia, ²Discipline of Anatomy and Pathology, Adelaide Medical School, University of Adelaide, Adelaide, Australia, ³School of Biomedical Engineering and Departments of Mechanical Engineering and Orthopaedics, University of British Columbia, Vancouver, Canada

Introduction: Most traumatic brain injuries (TBI) result from head impact without skull penetration (closed head impact), but the relationship between the biomechanics and the resultant brain pathology of such events is not fully understood. Large animal models of TBI provide benefits of physiological similarity, gyrencephalic brain structure, and scale. The aims of this study were to develop an ovine model of closed head impact induced TBI, and to characterize the resultant axonal pathology and head kinematics.

Methods: The heads of anaesthetised sheep were secured via a bite plate to a frame with fixed centre of rotation. The heads were impacted at the frontal bone, via an interface plate, with a custom impactor, inducing rapid sagittal plane head rotation. Impactor velocity was “low” (12.3 ± 0.4 m/s) (n=6), or “high” (15.2 – 15.4 m/s) (n=2). A sensor array attached caudal to the bregma measured head kinematics. Cranial fracture was assessed with computed tomography. Animals were maintained under anaesthesia until brains were perfusion fixed and harvested 6h post injury, and later analysed immunohistochemically for amyloid precursor protein (APP), the most sensitive early marker of axonal injury.

Results: No fractures of the cranial vault were observed. Peak angular accelerations of the head were 347 ± 82 krad/s² and 494–575 krad/s², at the “low” and “high” impact velocities, respectively. APP-immunopositive axons were consistently observed in the thalamus and, to a lesser degree, brainstem. The injured axons were greater in number in the “high” impact velocity group.

Conclusion: Preliminary results showed this ovine closed head impact model produced mild axonal pathology following sagittal-plane head impact, particularly in the thalamus. Impactor velocity was repeatable for the two reported impact velocities. Further work is needed to quantify the pathology produced at these and higher impact velocities, and to investigate the effects of head kinematics on pathology outcomes.

D114 Intraparenchymal B cell treatment promotes a neuroprotective microenvironment after traumatic brain injury through reciprocal immunomodulation with infiltrating peripheral myeloid cells

Liam Dwyer,² Dr Saumya Maheshwari ,² Emily Levy,³ Dr Mark Poznansky,^2,3 Dr Michael Whalen,^1,3 Dr. Ruxandra Sirbulescu^1,2,3

¹Harvard Medical School, Charlestown, United States, ²Vaccine and Immunotherapy Center, Boston, United States, ³Massachusetts General Hospital, Boston, United States

Introduction: Traumatic brain injury (TBI) remains a major cause of death and severe disability worldwide. Numerous neurodegenerative effects associated with TBI can be traced to aberrant inflammatory responses. Early intervention to modulate neuroinflammation may provide a successful therapeutic approach to TBI. Our previous work has demonstrated that exogenous mature naïve B220+/CD19+/IgM+/IgD+ B cells are potent modulators of inflammatory responses, and that B cell treatment was associated with structural and functional neuroprotection after TBI. Here, we further characterize the underlying cellular mechanisms of immunomodulation seen with intraparenchymal delivery of B lymphocytes in TBI.

Methodology: We used a mouse model of unilateral controlled cortical contusion TBI to assess post-injury cellular immune mechanisms after acute intraparenchymal delivery of B lymphocytes. At intervals between 18 hours to 2 months, immune infiltration and cytokine production in the ipsilateral and contralateral brain hemispheres of B cell-treated and control animals was assessed using multiplexed flow cytometry and histological analysis.

Results: B cell therapy increased the proportion of M2-like monocytes/macrophages as early as 2 days and up to 2 months after injury. Moreover, microglial activation was significantly reduced after 4 days and up to 2 months post-injury. B cells placed in the injured brain showed a complex time-dependent response, upregulating multiple immunomodulatory cytokines, including IL-10, IL-35, and TGF-beta. In the presence of B cells, infiltrating immune cells at the lesion site significantly increased their production of anti-inflammatory cytokines IL-10, TGF-beta, and IL-35 and reduced production of TNF-alpha and IL-2 as compared to saline-treated controls. Interestingly, ablation of peripheral monocyte/macrophages with clodronate liposomes demonstrated that this population was crucial for inducing a regulatory phenotype in exogenous B cells.

Conclusions: This study supports the hypothesis that introducing B cells into the injured CNS can initiate an immunoregulatory cascade with lasting neuroprotective effects, presenting a potentially promising cell-based therapy for TBI.

D115 Persistent changes in astrocyte-related MR measures are associated with post-traumatic epilepsy development in mice

Dr Federico Moro ,¹ Dr Ilaria Lisi,¹ Dr Rossella Di Sapia,¹ Dr Edoardo Micotti,¹ Dr Annamaria Vezzani,¹ Dr Kevin K. Wnag,² Dr Teresa Ravizza,¹ Dr Elisa R. Zanier¹

¹Mario Negri Institute for Pharmacological Research, Milan, Italy, ²Morehouse School of Medicine, Atlanta, USA

Introduction: The delay between experiencing a traumatic brain injury (TBI) and the emergence of epilepsy (PTE) provides a chance to intervene against epileptogenesis. Research of effective treatments is hampered by the lack of biomarkers apt to identify subjects at risk to enrich clinical trials population. Reactive astrogliosis is a common feature of both TBI and epilepsy. Our preliminary data show that astogliosis is a distinct pathological hallmark of PTE in mice 6 months post-TBI, and the modulation of specific astrocytic functions in animal models impacts epilepsy, thus suggesting that these cells are key contributors to disease mechanisms.

By magnetic resonance approaches, we investigated if astrocytic alterations may be assessed in vivo and have a role in PTE development in a TBI mouse model leading to 50% epilepsy incidence.

Methodology: Adult CD1 male mice underwent sham surgery (n=15) or TBI (n=30) by controlled cortical impact. Proton magnetic resonance spectroscopy (1H-MRS) was performed in the ipsilateral thalamus before (3w) and after (3mo) PTE onset; structural damage was assessed by diffusion tensor imaging (DTI) and 3D structural MRI at the chronic disease stage (5mo). After MRI analyses, mice were ECoG-monitored to assess PTE development and then sacrificed for brain histopathology.

Results: 1H-MRS showed that thalamic myo-inositol (an index of astrogliosis) and glutamate+glutamine (regulated by astrocytes) levels persistently increased only in PTE mice (p<0.05 vs non-PTE). Changes in DTI metrics compatible with astrocyte activation (increased FA and AD values) were observed in thalamic and cortical areas of PTE mice (p<0.05 vs sham) and associated with brain atrophy (p<0.05 vs non-PTE). The immunostaining of the astrocytic marker GFAP increased in the ipsilateral thalamus and perilesional cortex of PTE mice only (p<0.05 vs non-PTE).

Conclusions: Our data show that astrocytic activation and metabolism are persistently altered in PTE mice and can be monitored by in-vivo MR.

D116 Influence of APOE4 allele on TBI dependent Tau astrogliopathy

Dr Joseph Ojo ,¹ PhD Camila Ortiz,¹ PhD Andrew Pearson,¹ Miss MacKenzie Browning,¹ PhD Michael Mullan,¹ PhD Fiona Crawford¹

¹Roskamp Institute, Sarasota, United States

Introduction: Patients exposed to repetitive head injuries reveal tau-immunoreactive astrocytes many decades after their injuries. But it remains unclear whether this is a result of ageing. It is also unclear why some individuals exposed to r-mTBI develop tau astrogliopathy and others do not. APOE4 allele is a risk factor for late onset-AD, and influences TBI outcomes. In the brain, apoE4 is produced primarily in astrocytes. However, the role of APOE in driving TBI related Tau astrogliopathy remains unknown.

Methods: In this study, we expose our 20-hit r-mTBI/r-sham paradigm to APOE3-TR/GFAP-hTauWT and APOE4-TR/GFAP-hTauWT mice engineered to produce non-mutant hTau in astrocytes (including APOE3-TR and APOE4-TR as controls). We interrogate histopathological outcomes, tau biochemistry and ACSA2+ astrocyte RNAseq at 3- and 6-months post-injury. Microglia have been shown to drive the outcome and pace of tau proteinopathies in an APOE dependent manner, therefore we also explore whether microglial ablation with PLX3397 influences tau atrogliopathy and chronic outcomes in our mouse models.

Results: Our RNAseq data at the 3-month timepoint reveal 1735 dysregulated genes between APOE4-TR/GFAP-hTauWT vs APOE4-TR (sham) astrocytes. This was marked by an upregulation in PKA and synaptogenesis related pathways, and downregulation in CDK5, Rho-Family-GTpase and glutaminergic receptor signaling. But, no effects were observed in mice on the APOE3 background from this same comparison (i.e., APOE3-TR/GFAP-hTauWT vs APOE3-TR (sham) astrocytes). TBI augmented astrocyte response in mice on the APOE4-TR/GFAP-hTauWT strain (4530 dysregulated genes) compared to APOE3-TR/GFAP-hTauWT strain (no significant dysregulated genes reached statistical cut-off). These APOE4 mediated effects were marked by upregulation in fibrosis, phagosome formation, Endothelin-1 and CREB signaling, and downregulation of semaphorin and PTEN pathways. Additional interrogation for histopathology and tau biochemistry is ongoing.

Conclusions: Thus far, we have revealed that APOE4 genotype negatively influences astrocyte pathobiology in the context of tau astrogliopathy and injury. Further investigation is still ongoing.

D117 TBI is associated with fast splenic immune signaling and modulation

Mr. Florian Olde Heuvel ,¹ Ms. Sruthi Krishnamurthy,¹ Dr. Jin Zhang,¹ Mr. Fanjian Li,¹ Prof. Markus Huber-Lang,² Prof. Francesco Roselli^1,3

¹Ulm University, Ulm, Germany, ²University Clinic Ulm, Ulm, Germany, ³DZNE Ulm, Ulm, Germany

Introduction: Systemic inflammatory responses have been reported after traumatic brain injury (TBI), with almost all organs affected. The spleen, one of the most important immune regulatory organs, shows high interaction with the brain, known as the brain-spleen axis. Both brain-derived mediators as well as nerve fibres have been reported to directly affect immune cells in the spleen. We have previously investigated the effects of TBI on splenic immune cells, showing a fast maturation of splenic dendritic cells 3h after an experimental TBI. However, how the inflammatory cytokine response changes to TBI, and possibly DC maturation, remains largely unknown.

Methodology: We have performed large scale phospho-signaling, cytokine and angiogenesis arrays to determine early functional immune responses in spleen samples post TBI. We have used a larger sample size to investigate several cytokines in the spleen and plasma, using ELISA and qPCR.

Results: We found a specific signaling fingerprint with up almost 140 up or down phosphorylated proteins, pointing towards distinct cytokine related signaling pathways. The cytokine and angiogenesis arrays reveal the modulation of several cytokines, like an upregulation of CXCL1, CXCL13, CCL6 and a downregulation of IL-5. Several of these markers could be verified at protein level, however not at transcriptional level.

Conclusion: We have shown the early involvement and maturation of splenic dendritic cells upon TBI. Furthermore, this data points out that TBI is regulating splenic signaling, immune and vascular response. Showing the involvement of TBI on spleen functioning and its effect on systemic inflammatory responses.

D119 Administration of a synthetic cocktail based on mesenchymal stromal cell bioactive factors improves neurological outcome up to six months after traumatic brain injury in mice

Dr Francesca Pischiutta ,¹ Francesca Tribuzio,¹ Marta Magatti,² Laura Brunelli,¹ Giulia De Simone,¹ Fabrizio Ortolano,³ Luther Loose,¹ Costanza Bertani,¹ Federico Moro,¹ Antonietta Rosa Silini,² Roberta Pastorelli,¹ Ornella Parolini,^4,5 Elisa R Zanier¹

¹Istituto di Ricerche Farmacologiche Mario Negri IRCCS, Milano, Italy, ²Centro di Ricerca E. Menni, Fondazione Poliambulanza – Istituto Ospedaliero, Brescia, Italy, ³Neuroscience Intensive Care Unit, Department of Anesthesia and Critical Care, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Milan, Italy, ⁴Department of Life Science and Public Health, Università Cattolica del Sacro Cuore, Rome, Italy, ⁵Fondazione Policlinico Universitario “Agostino Gemelli” IRCCS, Rome, Italy

Introduction: Mesenchymal stromal cells (MSCs) form amniotic membrane induce neuroprotection in experimental TBI, via released bioactive factors (conditioned media, CM). We previously identified a low molecular weight (<700Da) CM fraction endowed with similar protective action than total CM after acute brain injury in vitro¹. Here we aimed at identifying the bioactive factors in the CM protective fraction, reconstituting a synthetic cocktail (SYNT) and testing its efficacy vs total CM in TBI models.

Methodology: Discovery: metabolomics profiling was performed by targeted and untargeted mass spectrometry. Factors enriched by 1.5-fold in the protective vs non-protective CM fraction were selected to reconstitute the synthetic cocktail (SYNT).

In vitro: TBI organotypic cortical brain slices treated with SYNT were evaluated for cell death, neuronal damage (NfL release) and gene expression analyses.

In vivo: adult C57BL/6J sham or severe TBI mice were randomized to receive daily intraperitoneal administration of CM, SYNT or control solution starting 3h post-TBI. Sensorimotor (SNAP and Neuroscore) and cognitive (Y maze) deficits and contusion volume (MRI) were longitudinally assessed up to 6-months.

Results: Metabolomic profiling identified more than 300 molecules, but only 4 were selectively enriched in the protective CM fraction: prostaglandins (PGA2,PGE2,PGJ2) and kynurenine. Their combination (SYNT) was protective in the in vitro TBI model, showing a reduced cell death in the lesion core (-18%), a decreased neuronal damage (NfL release:-45%) and a rescue effect on NeuN and BDNF mRNA downregulation induced by TBI.

TBI mice treated with CM or SYNT showed improvements in sensorimotor (AUC_0-6mo: CM+44%, SYNT+47%) and cognitive (4mo: CM:+50%, SYNT:+20%) outcomes, and a astrogliosis reduction in the corpus callosum (6mo: CM:-20%, SYNT:-23%). Instead, only CM treatment reduced the contusion volume (5mo:-25%).

Conclusions: MSC-released prostaglandins and kynurenine play a protective role, inducing functional improvement in TBI up to 6-months. Our data lay the basis for a synthetic cocktail reconstruction.

D120 Circular RNAs control synaptic stability and vulnerability in neurotrauma

Prof. Dr. Francesco Roselli ,¹ dr. Zhneghui Li,¹ dr. Albrecht Froehlich,¹ Florian olde Heuvel¹

¹Ulm University, Ulm, Germany

Circular RNA (CircRNA)are by-products of RNA splicing in which back-splicing of exons and introns results in a RNA molecule lacking open ends; circRNA are thought to be endowed with regulatory functions through interactions with microRNA, with RNA binding proteins and with long-noncoding RNAs. Here we explored the effect of TBI on a number of high-abundance neuronal circRNA derived from the splicing of genes coding for synaptic proteins. We identified a number of circRNA with distinct kinetics, being upregulated from the early stages (3h) to the subacute phase (7d). We confirmed that the elevation of specific circRNA is independent of the levels of the corresponding coding RNA, and it is selective of the hippocampus in the injured side (i.e., is not due to systemic stress). We confirmed the elevation of at least one circRNA in resection specimens obtained from human TBI patients. We engineered viral vectors to obtain either the knock-down or the upregulation in vivo of the circKhlh2 RNA, which displays a selective upregulation at 7d. Whereas the overexpression of circKhlh2 increased synaptic stability upon TBI, the KD resulted in an enhanced synaptic loss. Likewise, the upregulation of circKlhl2 accelerated the recovery of behavioural performance in the novel object recognition and Y-maze upon TBI. Thus, we demonstrate the functional relevance of circRNA upregulation in the recovery process that follows a traumatic injury.

D121 Limitation of Cerebral Blood Flow by Increased Venous Outflow Resistance in Elevated ICP

Yuliya Zadka,² Professor Guy Rosenthal ,¹ Dr. Omer Doron,³ Professor Ofer Barnea²

¹Hadassah-hebrew University Medical Center, Jerusalem, Israel, ²Tel Aviv University, Tel Aviv, Israel, ³Massachusetts General Hospital, Boston, USA

Background: Extensive investigation and modeling efforts have been dedicated to cerebral pressure autoregulation, primarily regulated by the cerebral arterioles ability to change their resistance and modulate cerebral blood flow (CBF). However, the impact of elevated intracranial pressure (ICP) on venous outflow resistance (VOR) and its consequential restriction of CBF have received less attention. This study introduces a novel venous component, the “Flow Control Zone” (FCZ), to simulate the effects of venous compression on CBF.

Methodology: We extended our prior intracranial fluid interactions model by incorporating the FCZ, which was designed using non-linear functions representing resistance as a function of cross-sectional area and the pressure-volume relations of the vessel wall. Utilizing a swine model of cerebral edema with graduated ICP elevation to calculate venous outflow resistance (VOR). We introduced the cerebral resistance index (CRI), a new metric indicating the ratio between VOR and cerebrovascular resistance (CVR).

Results: Model simulations correlated closely with experimental data, showing increased VOR with rising ICP (cross-correlation coefficient of 0.97, an MSE of 0.087, and an MAE of 0.15). CRI was strongly correlated to ICP in the swine model (r2 = 0.77, p = 0.00012, 95% CI [0.15, 0.45]). A CRI value of 0.5 was associated with clinically significant ICP elevations (>24 mm Hg), marking a point where arteriolar resistance regulations no longer significantly affect CBF. These findings highlight the critical role of venous compression at the FCZ in limiting CBF during high ICP states.

Conclusions: Our results demonstrate the importance of venous compression at the FCZ in determining CBF when ICP is elevated. The CRI may provide an indication of when compression of venous outflow becomes the dominant factor in limiting CBF following brain injury. This insight sheds light on venous control mechanisms, opening new opportunities for understanding cerebral edema's impact and developing targeted therapeutic strategies.

P001 Restoration of the cerebrovasculature to promote tau elimination from the brain following head trauma

Maxwell Eisenbaum,¹ Nolan Carothers,¹ Milica Koprivica,¹ Arianna Cembran,¹ Mike Mullan,¹ Fiona Crawford,^1,3 Benoit Mouzon,^1,3 Corbin Bachmeier ^1,2

¹Roskamp Institute, Sarasota, United States, ²Bay Pines VA Healthcare System, Bay Pines, United States, ³James A. Haley Veterans' Hospital, Tampa, United States

Introduction: Previously, we demonstrated brain pericytes have a role in the elimination of extracellular tau from the brain. The current studies examined the influence of traumatic brain injury (TBI) on cerebrovascular tau processing in the brain, in addition to evaluating the effect of pericyte stimulation on these processes post-injury.

Methods: Wild-type mice (3 months of age) were administered a brain injury 5 times per week for 1 month. One month later, mice were administered phenytoin (45 mg/kg) in mouse chow for 4 weeks, as our prior work showed phenytoin can stimulate brain pericyte function. To investigate tau uptake, fresh brain vessels were isolated from TBI mice (3 months post-first injury) and exposed to tau oligomers and fibrils for 1 hour and subsequently analyzed for tau content. For tau elimination, mice were administered the TBI paradigm and phenytoin treatment as above, and stereotaxically injected with tau oligomers or fibrils into the brain and analyzed for tau after 1 hour.

Results: For tau uptake, we observed a substantial decrease in the uptake of tau oligomers and fibrils in the TBI cerebrovessels (4-fold and 2-fold, respectively) compared to sham animals. Treatment with phenytoin improved the vascular uptake of both tau oligomers and fibrils. With respect to tau elimination, we observed elevated levels of oligomeric tau in the brain of the TBI animals compared to sham (2-fold), which was substantially mitigated with phenytoin treatment. Alternatively, we did not observe an effect of injury nor phenytoin treatment on the residence of tau fibrils in the brain.

Conclusions: The brain vasculature contributes to the elimination of pathogenic tau species from the brain, which is compromised following TBI. Our studies showed that pericyte stimulation can improve cerebrovascular tau uptake and elimination from the brain, which may be a viable treatment strategy to mitigate neurodegeneration following TBI.

P002 Glymphatic Dysfunction Following Adolescent Severe Traumatic Brain Injury is Exacerbated by Hemicraniectomy

Sydney Bennett ,¹ Dr. Zoe Teton,¹ Dr. Mayumi Prins¹

¹Ucla, Los Angeles, United States

Traumatic brain injury (TBI) remains the leading cause of death among adolescents in the US. The most definitive treatment option for intracranial hypertension remains the hemicraniectomy (HC, removal of half of the skull). The recently discovered cerebral glymphatic system describes a brain-wide pathway of perivascular CSF channels that promote a system of CSF-ISF exchange during sleep that cleanses the brain of metabolic waste products. Anterograde CSF movement through this system is partially dependent on the pulsatility of the brain against the closed cranial vault, pulsatility that is lost following HC. Glymphatic dysfunction has been demonstrated in many neurological pathologies, but has yet to be investigated in severe adolescent TBI. We hypothesized that severe TBI in adolescent rats will impair glymphatic influx and that this will be exacerbated by HC. Adolescent male rats (PND 40) received severe controlled cortical impact (CCI) injuries (n=6), CCI followed by a HC (n=5) or sham injuries (n=7). At 5 and 10 days post-injury, fluorescent dye was injected into the cisterna magna and allowed to circulate for 30 minutes. The brain was sliced and imaged to analyze mean fluorescence as a representation of glymphatic inflow.

Results: 5 days post-injury fluorescence intensity in the sham group was more than twice that of the TBI group (p=.024) and more than triple the HC group (p=.008). 10 days post-injury fluorescence intensity in the sham group was largely similar to the TBI group, however, was more than quadruple the HC group (p<.001).

Conclusion: Severe TBI causes significant glymphatic dysfunction in the adolescent rat though this has largely resolved by post-injury day 10. Hemicraniectomy results in a severe and persistent exacerbation of this dysfunction well past this point of recovery. This work was supported by NIH NS110757, the UCLA BIRC, BrainSPORT, and Easton Labs

P003 Investigation of the impact of experimental Traumatic Brain Injury on acute changes in immuno-metabolism gene expression in the mouse cortex and hippocampus

Ms Susan Burke ,¹ Professor Yvonne M. Nolan,^1,2 Dr Harriet Schellekens,^1,2 Dr Rebecca J. Henry¹

¹Department of Anatomy and Neuroscience, School of Medicine, University College Cork, Cork, Ireland, ²APC Microbiome Ireland, University College Cork, Cork, Ireland

Introduction: Neuroinflammation is a key secondary injury mechanism following exposure to traumatic brain injury (TBI). Evidence suggests that shifts in intracellular metabolic processes (oxidative phosphorylation → glycolysis) drives pro-inflammatory responses in the brain microenvironment. Thus, the aim of the present study was to examine acute changes in expression of inflammatory and metabolic genes in the cortex and hippocampus following exposure to TBI.

Methodology: Adult male C57BL/6 mice (n=5 per group) underwent either moderate level controlled cortical impact (CCI) or Sham surgery. Mice were euthanised at 7 days post injury (dpi) and ipsilateral cortical and hippocampal tissue was processed for mRNA analysis. Statistical analysis was preformed using an unpaired student’s t-test.

Results: TBI resulted in a significant increase in cortical expression of the pro-inflammatory mediators NLRP3 (p<0.01), IL-1β (p<0.01), TNF-α (p<0.001), NOX-2 (p<0.05), and p22phox (p<0.01), when compared to Sham operated counterparts. Similar effects were reported in the hippocampus including significant TBI-induced increases in NLRP3 (p<0.01), IL-1β (p<0.001), TNF-α (p<0.0001), NOX-2 (p<0.05), and p22phox (p<0.01). In addition, TBI resulted in a significant increase in both cortical (p<0.05) and hippocampal (p<0.001) expression of the microglial marker, CD11b, when compared to Sham counterparts. TBI resulted in a significant increase in cortical (p<0.001) and hippocampal (p<0.0001) expression of the master regulator of glycolysis, namely PFKFB3. Furthermore, TBI resulted in a significant increase in hippocampal (p<0.01), but not cortical expression of PFKFB1. Finally, TBI did not significantly alter gene expression of markers associated with oxidative phosphorylation including PGC-1α and SIRT-1 in the cortex or hippocampus.

Conclusions: Exposure to TBI results in acute increases in cortical and hippocampal pro-inflammatory gene expression and select markers of glycolytic metabolism. Developing our understanding of immuno-metabolic responses following exposure to TBI may offer novel insights into therapeutic targets.

P004 Blast Injury Alters Nociceptive Fiber Tracts of the Parabrachial Nucleus Impacting Motor Performance

Amanda Simon,¹ Keith Jamison,² Iryna Popovich,¹ Amy Kuceyeski,² D. Paola Calderon ¹

¹Weill Cornell Medical College, ²Cornell University

Abstract not published

P005 The effects of sex and nanoparticle treatment on traumatic brain injury

Connor Gee ,¹ Dr Aria Tarudji,¹ Dr Forrest Kievit¹

¹University of Nebraska - Lincoln, Lincoln, United States

Introduction: Nanoparticles are emerging as a promising treatment for traumatic brain injury (TBI). Previous work from our group has shown the potential for antioxidant nanoparticles (ANPs) in the treatment of TBI. However, sex-dependent responses to ANP treatment are still unresolved. Therefore, this work seeks to detangle true sex-based treatment differences from possible confounding differences in spatial learning and memory (SLM) paradigms typically optimized for male mice.

Methods: CCI mice were injected intravenously immediately following impact with ANPs or saline as a control. The Barnes Maze (BM) was utilized to assess SLM according to our previously established protocols in CCI mice. The novel object recognition test was also used to compare sex-based outcomes. Western blotting was used to measure α-spectrin breakdown products (SBDP) to compare molecular changes following ANP treatment.

Results: Male CCI mice showed a significant (p<0.01) deficit in SLM (36±7.074 seconds, probe primary latency) which was partially protected by ANP treatment (14.1±1.882). This response to ANP treatment seemed to be sex dependent as female CCI mice (25.35±7.003) and ANP treated mice (20.27±4.865) showed similar performance in SLM. However, female CCI mice did not show the significant deficit in SLM compared to control mice (25.35±7.003 vs 13.83±3.575, respectively, p=0.1391) as male mice did (36±7.074 vs 6.633±3.442, respectively, p<0.01). Interestingly, female control mice trended toward a worse performance than male control mice, but female CCI mice performed better than male CCI and ANP treated mice. Levels of necrosis from SBDP analyses partially supported sex-dependent responses to TBI and ANP treatment.

Conclusion: These results suggest a sex-based effect to both TBI and ANP treatment. Female mice seem to see less benefit from ANP treatment and reduced SLM deficits following CCI. This work underscores the importance of sex-based optimizations of behavioral paradigms for measuring outcomes in CCI mice.

P006 Brain Connectivity of the Awake, Freely-Moving Mouse before and after Repeat Concussions using Functional Ultrasound Imaging

Samuel Vander Dussen,¹ Jackson Alga-Sheriff,¹ Stefano Lepore,¹ Afshin Paydar,¹ Dr Neil Harris ¹

¹UCLA Brain Research Center & Department of Neurosurgery, UCLA, Los Angeles, United States

Introduction: Acquisition of functional connectivity (FC) using preclinical MRI typically requires sedation, which is either a significant confound, or at least a complication for making inferences about FC and behavior. Recent technological advancements in functional ultrasound imaging (FUSI) now enables the acquisition of blood volume and flow-weighted images at 100-micron resolution every 400ms through the intact skull of a mouse.

Methodology: In this work we investigated the utility of FUSI to monitor regional brain FC changes before and after multiple concussions (1/day) in the same mouse. We combined FUSI with the Mobile Homecage® to acquire FC data from awake, head-fixed, moving, adult mice for 30min, pre-injury and at 30min after 1-concussion, 5-concussions and 3wks later. We injured by either downward-direct-force onto the foam-restrained skull (DDF-RS, n=7), or side-impact into the unrestrained skull (SI-US, n=10) and 8 shams. Data were co-registered into Allen brain atlas space, and processed for FC during resting or moving conditions using time-locked Neurotar-tracking data, and assessed with network-based statistical correction over 34-brain regions (P<0.05).

Results: Concussion by DDF-RS resulted in immediate hyperconnectivity after 1-concussion compared to pre-injury, followed by hypoconnectivity after 5-concussions, which persisted to 3wks, regardless of mouse movement status. SI-US concussion resulted in subcortical hyperconnectivity and cortical hypoconnectivity after 1-concussion during resting conditions, but global hypoconnectivity during movement. After 5-concussions hyperconnectivity predominated, and this persisted and increased significantly in strength by 3wks.

Conclusion: These data highlight the utility of FUSI, the effect of different types of concussive impact, and of mouse movement status on brain functional connectivity.

P007 Neuronal Tafazzin is Essential for Cardiolipin Diversity in the Brain and for Normal Cognitive Function

Dr. Aybike Korkmaz ,^1,2 Dr. Tuana Demir Islak,¹ Yuan Gao,³ Dr. Svietlana Samovich,¹ Julie Ann Scott,⁴ Bharathi Sivakama,⁵ Carla Johnson,⁴ Dr. Eric Goetzman,⁵ Dr. Valerian E. Kagan,⁶ Dr. Hülya Bayir¹

¹Redox Health Center, Dept. of Pediatrics, Columbia University, New York, United States, ²Graduate School of Health Sciences, Ankara University, Ankara, Turkiye, ³Biomedical Science, University of Miami, United States, ⁴Children's Neuroscience Institute, University of Pittsburgh, Pittsburgh, United States, ⁵Division of Genetic and Genomic Medicine, University of Pittsburgh, Pittsburgh, United States, ⁶Department of Environmental and Occupational Health, University of Pittsburgh, Pittsburgh, United States

Cardiolipin (CL) is an unusual mitochondria-specific phospholipid featuring four fatty acyls (FA). Brain CLs exhibit a distinct composition enriched with long-chain polyunsaturated FA (PUFA). The origin of brain CL diversity and the role of PUFA-CL in cognitive functions are unknown. We hypothesized that neuronal CL remodeling mediated by Tafazzin (Taz) is a critical driver of brain CL diversity. To experimentally test this hypothesis, we developed a neuronal Taz (nTAZ) Knock-Out (KO) mouse model using CamKIIa and Cre-Lox system. Expectedly, nTAZ KO mice had decreased PUFA-CL and increased monolyso-CL (MLCL) levels in cortex vs wild-type and Cre-carrier controls. They also displayed diminished social interaction and worse spatial memory vs controls. Under normal conditions, nTAZ deficiency did not affect cortical oxidative phosphorylation. Our biochemical experiments demonstrated that MLCL can form a peroxidase complex with a mitochondrial intermembrane space hemoprotein, cytochrome c (cyt c), leading to PUFA-CL oxidation and caspase-3 activation. Exposure to controlled cortical impact (CCI) further elevated the level of MLCL and caused PUFA-CL oxidation and caspase 3/7 activation in the pericontusional cortex of nTazKO vs Cre-carrier mice. Furthermore, nTaz KO mice had larger lesion volume with impaired well-being and spatial memory vs. Cre-Carrier controls after CCI. Previously, we showed the presence of brain-specific PUFA-CLs in plasma serves as a diagnostic and prognostic marker of acute brain injury. Overall, nTAZ-mediated CL remodeling is crucial for shaping the unique profile of the brain cardiolipinome. nTAZ KO-induced changes in CL/MLCL lead to functional deficits, which are exacerbated by CCI. Support: NS076511; NS061817; NS117000.

P008 Traumatic brain injury-induced disruption of the circadian clock in rats

Dr. Lu-Ting Kuo ¹

¹National Taiwan University Hospital, TaipeI, Taiwan

Disturbances in the circadian rhythm have been reported in patients following traumatic brain injury (TBI). However, the rhythmic expression of circadian genes in peripheral blood leukocytes (PBL) following TBI has not yet been studied.

In this study, The messenger ribonucleic acid (mRNA) expression of period 1 (Per1), Per2, Per3, cryptochrome 1 (Cry1), Cry2, brain and muscle aryl hydrocarbon receptor nuclear translocator-like 1 (Bmal1), and circadian locomotor output cycles kaput (Clock) was quantified in PBLs from sham-operated rats and rats with acute subdural hematoma (ASDH) over a 48-h period. The rectal temperature of the animals was measured every 4 h over 2 days. The mesor, rhythm, amplitude, and acrophase were estimated using cosinor analysis.

Cosinor analysis revealed that Per2, Cry1, and Bmal1 mRNAs were rhythmically expressed in the PBLs of sham-operated rats. In contrast, fluctuations in rhythmic expression were not observed following ASDH. The rectal temperature of sham-operated rats also exhibited rhythmicity. ASDH rats had a disrupted rectal temperature rhythm, a diminished amplitude, and an acrophase shift.

In conclusion, TBI with ASDH results in dysregulated expression of some circadian genes and changes in body temperature rhythm. Further research is required to understand the pathophysiology of altered circadian networks following TBI.

P009 Spatially resolved astrocyte morphological and transcriptional changes in postmortem human tissue from traumatic brain injury cases

Radina Lilova ,¹ Dr. Declan McGuone,^2,3 Dr. Peter Hamilton,¹ Sean Regan,¹ Rana Ansari,¹ Dr. Audrey Lafrenaye¹

¹Virginia Commonwealth University, Richmond, United States, ²Yale School of Medicine, New Haven, United States, ³State of Connecticut Office of the Chief Medical Examiner, Farmington, United States

Introduction: The prevalence of traumatic brain injury (TBI) is increasing globally. Most TBIs are classified as mild, with resultant pathologies being predominantly diffuse as opposed to focal. Following primary insult, neuroinflammation and edema contribute to secondary injuries. Astrocytes are the most abundant cells in the CNS and are vital for structural, metabolic, and homeostatic maintenance. Astrocytes are also implicated in neuroprotection and are reactive to CNS stress. Our rat studies demonstrate morphological changes in thalamic astrocytes following TBI, as well as changes in levels of major astrocyte proteins following TBI. The current study aims to further our rat findings following diffuse injury in thalami of postmortem human tissue as well as substantially expand on the investigation of astrocyte changes, incorporating molecular as well as transcriptomic approaches.

Methodology: For the rat studies, we used the central fluid percussion injury (cFPI) model of diffuse TBI in adult male rats. Thalamic astrocyte changes were assessed using western blotting and immunohistochemical analysis at 4w post-injury. For the human studies, we used postmortem thalamic tissue from a population-based sample of critically investigated cases of lethal TBI from the State of Connecticut Office of the Chief Medical Examiner (OCME). We performed immunohistochemical labelling against GFAP to assess astrocyte morphological changes across injured and non-injured human thalami using ImageJ and are correlating these changes to proximity to pathological features within postmortem tissue. To expand on this information, we are investigating the transcriptomics of astrocytes in postmortem human thalami through a digital spatial profiling approaching using the GeoMx platform.

Results: We found morphological changes, but not bulk protein changes, in rat thalamic astrocytes following cFPI. Transcriptomic studies on injured and non-injured postmortem human thalami are actively ongoing.

Conclusions: These studies will greatly expand on the role of astrocytes in the thalamus following TBI.

Funded: NINDS R56NS128104, R01NS128104, R01NS096143.

P012 Thiol-based antioxidant nanoparticles restore redox metabolism in traumatic brain injury

Brandon McDonald ,¹ Dr. Forrest Kievit¹

¹University of Nebraska — Lincoln, Lincoln, United States

Introduction: Redox stress is a significant contributor to the metabolic dysfunction present in traumatic brain injury (TBI). Indeed, peroxides and lipid peroxidation products augment endogenous antioxidant systems, depleting NADPH and ATP. However, redox stress also activates mTOR, and inhibits AMPK, increasing ATP catabolism. Therefore, alleviating redox stress may restore metabolic function and provide a therapeutic effect in TBI. We have previously shown that thiol-based neuroprotective copolymers (NPCs) accumulate within the damaged brain and alleviate redox stress in TBI. Thus, our goal is to examine the effect of NPCs on restoring redox metabolism in TBI.

Methodology: We employed the CCI model (4.0 m/s velocity, 2.5 mm depth) to investigate the therapeutic efficacy of NPCs in TBI. 8-week old male C57BL/6J mice were divided into three groups (Control, CCI, and CCI+NPC), and NPCs were injected intravenously (8 mg/kg) immediately post-CCI. Brains were harvested at 1, 3, and 7 days post-CCI to examine changes in peroxidase expression via CAT and Prx6, Nrf2 activation via HO1, and metabolism via mTOR and AMPK phosphorylation.

Results: NPCs reduced CAT expression in the ipsilateral cortex (IC) at 1 day post-CCI (p<0.05) and Prx6 expression in the ipsilateral hippocampus at 1 and 7 days post-CCI (p<0.01). HO1 and mTOR phosphorylation (Ser2448) peaked in the IC at 1 day post-CCI (14.49 ± 1.97, p<0.0001; 59.35 ± 9.87, p<0.0001), which was reduced following NPC treatment (p<0.0001). Additionally, NPCs increased AMPK phosphorylation (Thr172) in the IC at 1 and 3 days (6.53 ± 2.35, 5.33 ± 2.63, p<0.05).

Conclusions: NPC treatment restored peroxidase expression and Nrf2 activation, suggesting an amelioration in redox stress and energy-dependent activity. Additionally, NPCs activated AMPK, shifting metabolism towards ATP anabolism. Therefore, alleviating redox stress via exogenous antioxidants may restore ATP, preserving neuronal function and activity. Thus, thiol-based antioxidant treatments may provide a neuroprotective effect in TBI.

P013 MAPping the axon degeneration and repair following traumatic axonal injury in vitro

Dr. Hailong Song ,¹ Mr. Frank Rauscher,¹ Dr. Jean-Pierre Dolle,¹ Dr. Douglas Smith¹

¹University of Pennsylvania, Philadelphia, United States

Introduction: Concussion is a major health concern, with increasing recognition that traumatic axonal injury (TAI) is a key pathologic substrate. Indeed, dynamic deformation of axons induces immediate mechanical breaking of axonal microtubules (MTs), thereby inducing axonal transport interruption and accumulation of proteins in hallmark periodic swellings. While axonal swellings and degeneration are thought to be the primary pathological feature of TAI, we have shown that this occurs in only a subset of axons in each tract. Indeed, we have shown that adjacent axons in the same tract appear to follow a pathway of repair and recovery. Here, we explored potential divergent mechanistic pathways of axon degeneration and repair through selective regulation of MT associated proteins (MAPs).

Methodology: We used our well-characterized in vitro axon stretch injury model, based on parameters of dynamic white matter tissue deformation in human concussion. Following acute axonal injury, we examined the expression and modifications of major MAPs, including tau and MAP6, that are known to bind axonal MTs. We further explored their association, or lack thereof, with the accumulation of a proteolytic fragment of spectrin, SNTF, indicating axon degeneration.

Results: Acutely after TAI, in axons destined to degenerate, we observed that disruption of axonal MTs induced phosphorylation of tau (p-tau) and production of SNTF, both of which accumulated in swellings. Conversely, in remaining axons, endogenous MAP6 expression and distribution significantly increased, signaling enhanced MT stabilization. These axons did not display swellings or other degenerative morphologies.

Conclusions: These data suggest that the immediate mechanical damage to axonal MTs due TAI in concussion appears to trigger a selection process that drives some axons towards repair/recovery and others towards degeneration. These observation warrants further examinations of how MAPs regulate axon fate and possibly reveal targets to promote axon recovery.

P014 Acute changes in neurosteroid levels following mild traumatic brain injury

Kosisochukwu Emmanuel Umeasalugo ,^1,2,3 Dr. Igor Khalin,^1,4 Dr. Burcu Seker,^1,4 Dr. Carina Exner,¹ Dr. Philippe Liere,⁵ Prof. Michael Schumacher,⁵ Prof. Dr. Inga Koerte,^2,6,7 Prof. Dr. Nikolaus Plesnila^1,4

¹Institute for Stroke and Dementia Research (ISD), LMU University Hospital, LMU Munich, Munich, Germany, ²cBRAIN, Department of Child and Adolescent Psychiatry, Psychosomatics and Psychotherapy, LMU University Hospital, LMU Munich, Munich, Germany, ³Graduate School of Systemic Neurosciences (GSN), LMU Munich, Munich, Germany, ⁴Munich Cluster for Systems Neurology (Synergy), Munich, Germany, ⁵U1195 INSERM and University Paris Saclay, Le Kremlin Bicetre, Paris, France, ⁶Psychiatry Neuroimaging Laboratory, Brigham and Women's Hospital, Harvard Medical School, Boston, USA, ⁷Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School, Boston, USA

Introduction: More than 30% of patients show persisting symptoms after mild traumatic brain injury (mTBI), but no biomarkers are available to identify these individuals. Our study evaluated whether endogenous neurosteroids may serve as biomarkers. Neurosteroids include allopregnanolone and 3α,5α-tetrahydrodeoxycorticosterone (THDOC), and their 3β-enantiomers isopregnanolone (ISOPREG) and 3β,5α-tetrahydrodeoxycorticosterone (ISODOC). We thus hypothesized that mTBI would elicit changes in neurosteroid levels in the brain which may also be reflected in plasma.

Methodology: C57BL/6 mice were subjected to a modified model of weight drop mTBI which combines impact with rotational acceleration and were evaluated for locomotion and balance up to 24 hours later. Brains and plasma were harvested 6 and 24 hours after mTBI for neurosteroid analysis by gas chromatography-tandem mass spectrometry.

Results: Weight drop mTBI significantly prolonged the wake-up time from anesthesia, mimicking acute loss of consciousness, but did not cause any fractures, intracranial hemorrhages, or mortality. There were also no deficits in locomotion or balance following mTBI. However, the concentrations of ISOPREG and ISODOC were significantly decreased at 6- and 24 hours after mTBI in the brain, while ISODOC concentrations in plasma were significantly reduced at 6- and 24 hours. The affected neurosteroids were also more abundant in the brain than in the plasma, depicting greater local neurosteroid activity and less systemic influence. Other neurosteroids did not show any consistent changes after mTBI.

Conclusions: Our results show that although mTBI did not cause gross damage such as fractures, bleeding, or locomotor deficits, it led to significant changes in neurosteroid levels. Because the reduced neurosteroids are catalyzed by a single enzyme, 3β-hydroxysteroid oxidoreductase (3β-HSOR), mTBI seems to specifically affect brain 3β-HSOR activity which can further be detected in plasma. Neurosteroids, especially ISODOC, may thus be biomarker candidates for mTBI.

P015 Tau prions mediate long-term cognitive impairment and neurodegeneration in traumatic brain injury patients

Gloria Vegliante ,^1,2 Ilaria Bertani,³ Ilaria Lisi,¹ Ilaria Raimondi,³ Elena Restelli,³ Fabrizio Ortolano,⁴ Marco Carbonara,⁴ David J Loane,² Roberto Chiesa,³ Elisa Zanier¹

¹Mario Negri Institute for Pharmacological Research IRCCS, Department of Acute Brain and Cardiovascular Injury, Milan, Italy, ²School of Biochemistry and Immunology, Trinity Biomedical Sciences Institute, Trinity College Dublin, Dublin, Ireland, ³Mario Negri Institute for Pharmacological Research IRCCS, Department of Neuroscience, Milan, Italy, ⁴Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Department of Anaesthesia and Critical Care, Milan, Italy

Introduction: We have previously demonstrated in a wild-type mouse model that a single severe traumatic brain injury (TBI) induces a self-propagating tau pathology that progressively spreads in the brain and can be horizontally transmitted to naïve mice, causing synaptic degeneration and cognitive impairment¹.

Objective: To test whether transmissible tau pathology is generated after TBI in patients.

Methods: Fresh-frozen human brain tissue (hTBI, GCS at admission <7) was characterized for the presence of hyperphosphorylated tau (P-tau), homogenized (10% w/v in PBS) and inoculated bilaterally in the hippocampus and overlaying cortex of male C57BL/6J (WT) and tauKO mice. A glioma brain specimen was used as control (hCT). The cognitive function was assessed by the novel object recognition (NOR), and Y-maze tests up to 16 months post-inoculation (mpi). Pre- and post-synaptic markers and P-tau were evaluated by immunohistochemical analysis. mRNA signatures were examined using a nCounter Mouse Neuropathology panel (NanoString).

Results: An early and persistent memory deficit was observed in mice inoculated with hTBI homogenate as assessed by NOR at 4, 8 and 12 mpi (*p<0.05), associated with widespread tau pathology and synaptic loss (V-GLUT1 and drebrin, *p<0.05). The ability to induce cognitive impairment and P-tau deposition was retained when serially transmitted in naïve WT recipient mice (*p<0.05). In contrast, hTBI-inoculated tauKO mice showed preserved memory function as compared to hTBI-inoculated WT mice (*p<0.05), consistent with a key role of endogenous tau in the spreading of tau pathology and mediating its amnestic effect. Gene expression analysis showed a reduced myelination, synaptic dysfunction, and a pathological microglia phenotype in hTBI-inoculated mice (*p<0.05). The inflammatory phenotype was confirmed by pathway analysis and immunocytochemical data.

Conclusion: Self-templating pathological tau is induced by a single TBI in humans, plays a key role in exacerbating post-traumatic pathology, and may explain how an acute biomechanical insult predisposes to neurodegeneration.

P016 Systematic characterization of preclinical models of mild, moderate and severe TBI utilizing controlled cortical impact

Dr. Johannes Walter ,¹ Anastasiia Balandina,¹ Noah Kruck,¹ Dr. Obada T. Alhalabi,¹ Prof. Thomas Skutella,² Prof. Sandro M. Krieg,¹ Prof. Andreas W. Unterberg,¹ PD Dr. Alexander Younsi¹

¹Department for Neurosurgery, Heidelberg University Hospital, Heidelberg, Germany, ²Institute for Anatomy and Cell Biology, University of Heidelberg, Heidelberg, Germany

Introduction: The translation of preclinical data into clinical practice is significantly hampered by an insufficient standardization of traumatic brain injury (TBI) models. Therefore, we evaluated the effect of different adjustable parameters of the Controlled Cortical Impact (CCI) model on trauma intensity to produce a distinct mild (mCCI), moderate (moCCI) or severe (sCCI) injury.

Methodology: A total of 128 wildtype C57Bl/6 mice were subjected to mCCI, moCCI and sCCI using three different sets of parameters (tip diameter/impact depth/velocity of 1mm/1mm/4m/s; 2mm/2mm/6m/s; 3mm/3mm/8m/s, respectively) or to sham surgery (craniotomy only). Structural damage (Nissl staining), neurological function (Hole Board, Video Open Field and CatWalkXT® gait analysis) as well as blood serum concentrations of 13 pro-inflammatory cytokines (flow cytometry-based immunoassay) were assessed at 1-, 3-, 7- and 28-days post injury (dpi).

Results: 1-day mortality (2.6%, 2.5% and 35.0% after mCCI, moCCI and sCCI, respectively) and lesion volumes (e.g., 1.8 ± 0.2 mm³, 14.1 ± 1.6 mm³ and 33.8 ± 3.4 mm³ for mTBI, moTBI and sTBI 7 dpi, respectively) increased with rising trauma intensity. Several domains of the neurological function (e.g., CatWalkXT® average run speed of 23.1 ± 4.5 m/s, 19.5 ± 3.0 m/s and 14.8 ± 3.8 m/s for mCCI, moCCI and sCCI 3 dpi, respectively) were impaired throughout the observation period. Interestingly, higher trauma intensities were not associated with increased blood concentrations of pro-inflammatory cytokines; in contrast, in many cases, the highest cytokine concentration could be detected in the mCCI group (e.g., Interferon γ: 3.0 ± 0.7 pg/ml vs. 0.2 ± 0.2 pg/ml, p=0.002 for mCCI vs. sCCI 3 dpi, respectively)

Conclusions: While the three evaluated sets of CCI parameters lead to progressive structural and functional damage that resemble the impairments seen in patients, increasing trauma intensities did not correlate with a more pronounced systemic inflammatory response.

P017 Neurofilament light chain deletion is associated with increased neuropathology following diffuse traumatic brain injury

Dr Yasmine Doust,¹ Mr Ross Langley,¹ Ms Emily Garratt,¹ Prof Anna King,¹ Dr Jenna Ziebell ¹

¹University of Tasmania, Hobart, Australia

Breakdown of the axonal cytoskeleton is a key component in traumatic brain injury (TBI)-induced neuropathology. We hypothesised that an altered neuronal cytoskeleton (neurofilament knock-out; NFL-KO) would attenuate neuropathology after diffuse TBI, evidenced through APP accumulation. Diffuse TBI was modelled in male and female NFL-KO and wildtype (C57/Bl6; WT) mice by midline fluid percussion injury. Immunohistochemical analysis was conducted on 132 brains from naïve and injured mice at 3 hours, 1- and 3-days post-injury and matched for age, sex and genotype (n = 5–6/group). TBI-induced neuropathology (APP-positive axons) was evident in the corpus callosum (p < 0.0001), primary sensory barrel field (p < 0.0001), dentate gyrus (DG) of the hippocampus (p < 0.001), and ventral posteromedial nucleus (VPM) of the thalamus (p < 0.001), regardless of time post-injury, genotype or biological sex. Surprisingly, APP+ axonal injury was more extensive in all brain regions of injured NFL-KO mice compared with WT (corpus callosum: p < 0.0001; primary sensory barrel field: p < 0.0001; DG: p < 0.001; VPM: p < 0.001), regardless of time post-injury or biological sex. Biological sex differences were evident also across time post-injury where APP+ neuropathology was more robust in male mice compared with females that occurred in a region-specific manner (corpus callosum: p < 0.05; primary sensory barrel field: p < 0.01; VPM: p < 0.05). These results show that in the initial stages following TBI, the structure of the axonal cytoskeleton is important in the sequelae of diffuse TBI pathophysiology in both biological sexes.

P018 Beyond Mild, Moderate, and Severe: Classifying TBI Subgroups with Machine Learning

Marika Abe ,^1,8 Harri Merisaari,^2,3 Jussi Posti,^4,5 Olli Tenovuo,^4,6 Peter Ngum^4,6,7,8

¹Johns Hopkins University, Whiting School of Engineering, Baltimore, USA, ²Turku Brain and Mind Center, University of Turku, Finland, ³Department of Diagnostic Radiology, University of Turku, Finland, ⁴Turku Brain Injury Center, Turku University Hospital, Finland, ⁵Neurocenter, Department of Neurosurgery, Turku University Hospital, Finland, ⁶Department of Clinical Neurosciences, University of Turku, Finland, ⁷Johns Hopkins University, Carey Business School, Baltimore, USA, ⁸CerebraNetics, Baltimore, USA

Introduction: Traumatic Brain Injury (TBI) presents a spectrum of challenges due to its heterogeneous nature. Traditional mild, moderate, and severe classifications fail to capture its multifaceted impacts. Recent efforts underscore the urgency of refining our understanding of TBI. This study employs an unsupervised machine learning technique to integrate age, Glasgow Coma Scale scores (GCS), and pupil reactivity with FDA-approved GFAP (Glial Fibrillary Acidic Protein) and UCHL1 (Ubiquitin Carboxy-terminal Hydrolase L1) blood-based biomarkers (BBM), to delineate distinct TBI subgroups across all severities.

Methodology: We conducted a retrospective analysis using the EU-funded TBIcare dataset comprising 181 patients after excluding cases with missing GOSE. Our analysis included age, GCS scores, specifically the worst motor response score in patients not requiring intubation, and FDA-approved GFAP and UCHL1 BBM admission levels. Through comprehensive preprocessing, we applied K-means clustering for its efficiency in revealing natural patient groupings without predetermined outcomes. Subsequently, we analyzed cluster characteristics to uncover patterns linking clinical and biomarker profiles with recovery trajectories.

Results: Our study identified four TBI patient clusters, distinguished by demographic differences, injury severity, and biomarker levels. These differences are associated with variations in recovery outcomes, as measured by GOSE scores. Specifically, a cluster with younger patients (n=70, avg. age 30.03, GCS 13.6) with lower biomarker levels (UCHL1: 170.04 pg./mL, GFAP: 6035.76 pg./mL) demonstrated more favorable recovery (avg. GOSE: 6.87), while a cluster of older patients (n=80, avg. age 65.09, GCS 13.30) with higher biomarker levels showed poorer outcomes (avg. GOSE: 5.32). ANOVA confirmed significant differences in age, GCS, and biomarkers across clusters (p < 0.0001), underlining their distinctiveness.

Conclusion: Our analysis revealed four TBI patient clusters, showing significant differences in demographics, injury severity, and biomarker levels associated with distinct functional outcomes. Future work aims to validate outcome prediction capacities of these clusters and their utility in customizing therapeutic strategies.

P019 Enhancing individualised traumatic brain injury patient care with unsupervised models for arterial blood pressure signal artefact removal

Xuhang Chen ,¹ Stefan Yu Bögli,¹ Erta Beqiri,¹ Ihsane Olakorede,¹ Marina Sandra Cherchi,¹ Masumi Tanaka Gutiez,¹ Tommaso Rochat,¹ Cameron Smith,¹ Ari Ercole,² Peter Smielewski¹

¹Brain Physics Laboratory, Department of Clinical Neuroscience, Division of Neurosurgery, University of Cambridge, CAMBRIDGE, United Kingdom, ²Division of Anaesthesia, University of Cambridge, CAMBRIDGE, United Kingdom

Introduction: In the intensive care unit, continuously monitored signals, such as arterial blood pressure (ABP), are susceptible to noise contamination, e.g., related to arterial line clotting and flushing. These artefacts may produce significant, misleading, alterations in waveform-derived metrics. Given emerging individualised management protocols based on physiological monitoring, it is essential to address these issues in the pre-processing stages with automated algorithms. This project aimed to explore a neural network approach to this challenge, building on an architecture previously proposed by our group [1], and validating it on a large traumatic brain injury (TBI) cohort data.

Methodology: We implemented an unsupervised Variational Autoencoder (base) model and trained it on data from 150 TBI patients. By applying physiological heuristic filters (thresholds-based), we initially separated the data into normal and abnormal 10s segments. The model was trained on the normal segments to learn the essential characteristics of ABP signals, enabling it to identify artefacts by comparing the reconstructed output with the original input data via mean absolute error. This approach was subsequently fine-tuned for individual patients.

Results: Our base model achieved an accuracy of 85.9±5.2% on a balanced test dataset with 1,000 samples from 10 patients. After individual fine-tuning, the accuracy increased to 86.9±4.3%. In addition, qualitative assessment of reconstructed waveforms showed better improvement. The average processing speed of the model for 1-hour data was 0.054s, which demonstrates the real-time processing ability at the bedside monitor.

Conclusion: This study demonstrates the potential of our unsupervised, generative, model for automated ABP signal artefact removal, highlighting the benefits of individualised model fine-tuning to help the model adjust to individual patients. Importantly, the model can reproduce normal traces, which can substitute noise-contaminated sections. Our findings suggest a scalable and efficient solution for enhancing the quality of critical care monitoring with the potential for real-time bedside application.

P020 Defining an AI model’s reality: an analysis of ‘ground truth’ data for papilloedema detection as a marker of raised intracranial pressure

Dr Lekaashree Rambabu ,^1,2,3 Mr Thomas Edmiston,^3,4 Dr Brandon G. Smith,^2,3 Dr Stasa Tumpa,^3,5 Dr Katharina Kohler,^2,3,7 Dr Angelos G. Kolias,^3,6 Professor Peter J. Hutchinson,^3,6 Dr Tom Bashford^2,3,7

¹University of Leicester, Leicester, United Kingdom, ²International Health Systems Group, Department of Engineering, University of Cambridge, Cambridge, United Kingdom, ³NIHR Global Health Research Group on Acquired Brain and Spine Injury, University of Cambridge, Cambridge, United Kingdom, ⁴School of Clinical Medicine, University of Cambridge, Cambridge, United Kingdom, ⁵West Suffolk NHS Foundation Trust, Bury Saint Edmunds, United Kingdom, ⁶Division of Academic Neurosurgery, Addenbrooke's Hospital, Cambridge, United Kingdom, ⁷Division of Anaesthesia, Addenbrooke's Hospital, Cambridge, United Kingdom

Introduction: Papilloedema is a condition where swelling of the optic disc occurs secondary to raised intracranial pressure (ICP). The ‘ground truth’ or ‘reference standard’, is ‘the best available method for establishing the presence or absence of the target condition'[1]. The quality of the methods used to establish ground truth in deep learning systems (DLS) can introduce bias, and hugely impact model performance and safety.

Methods: We conducted a systematic review of studies that used artificial intelligence (AI) to detect papilloedema. We searched Ovid Medline, Embase, Web of Science and IEEE Xplore databases. We assessed reporting quality using the Checklist for Artificial Intelligence in Medical Imaging, and critically appraised with a 5-domain rubric, ‘SMART[2]. In this abstract, we share results of the analysis of ground truth in a subset of included studies.

Results: Of the 11 DLS models included, the ground truth for 7 models was set by ophthalmologists or ‘expert clinicians’. Only one study use ‘neuroimaging evidence of intracranial cause for raised ICP or lumbar puncture with opening pressure >28 cmH20’ as ground truth. Two studies do not report reference standard. One paper partially reports use of a local dataset through the use of a ‘combination of detailed clinical history, clinical examination, ophthalmic imaging, and/or neurological imaging.

Discussion: The gold standard method for establishing ground truth in AI for papilloedema detection as a marker of raised ICP is yet to be known. Establishing objectivity, reproducibility and clarity in the ground truth will be the defining point of an AI model’s instruction for diagnosis. Majority of the included studies used expert clinicians for diagnosis of papilloedema at an image level, which risks misdiagnosis with its mimics such as pseudopapillodema, introduces bias and limits model’s ability to only provide diagnosis of papilloedema at an image level, without any clinical correlation to ICP.

P021 Unsupervised Clustering Analysis in Neurocritical Care Patients: A Systematic Review

Jeanette Tas ,^1,2 Verena Rass,³ Melanie Bergmann,^1,2 Raimund Helbok^1,2,3

¹Department of Neurology, Kepler University Hospital, Johannes Kepler University Linz, Austria, Linz, Austria, ²Research Institute for Neuroscience, Johannes Kepler University Linz, Austria, Linz, Austria, ³Department of Neurology, Medical University of Innsbruck, Innsbruck, Austria, Innsbruck, Austria

Background: Managing patients with acute brain injury in the neurocritical care unit (NCC) has become increasingly complex due to technological advances and information derived from multimodal neuromonitoring. Diverse data streams necessitate innovative approaches for clinicians to understand complex interactions between i.e., physiologic variables. Unsupervised clustering has been applied in NCC. However, a systematic review dedicated to this field is lacking. We aim therefore to provide insights in this field.

Methods: In this study we systematically review articles using clustering in acutely brain injured patients and aim for a guidance/recommendation for future research. The primary objective is to provide an overview of clustering applications in NCC studies, including data types and identified clusters. As secondary objective we explore on different study design, settings, and challenges in NCC clustering. Databases (Medline, Scopus, Web of Science) were searched for English studies involving NCC patients; traumatic brain injury, subarachnoid hemorrhage, intracerebral hemorrhage, acute ischemic stroke, hypoxic ischemic brain injury, that focused on unsupervised clustering analysis. Abstracts were screened and study variables (publication, objectives, study-, patient- and clustering characteristics, and relationship with clinical outcome) extracted and summarized.

Results: So far, we identified 18 articles using clustering in NCC. Predominantly, studies focused on TBI patients. The studies covered a variety of clustering methods both traditional (i.e., k-means) and advanced like Gaussian Mixture Model-Based Clustering, and explored diverse data types, with intracranial pressure mostly used. Furthermore, among the studies that examined clinical outcomes, the majority reported a significant difference among the clusters.

Conclusions: Unsupervised clustering will help us to better phenotype diseases and disease stages of individualized neurocritical care patients. It reveals so far hidden associations between distinct resource data including demographics, biomarkers, monitoring, and neuroimaging. In the future, this may help to identify and delineate novel NCC management principles and treatment approaches.

P022 Prevention is Better Then Cure: The Effect of Helmet Usage Among Cyclists Sustaining Traumatic Brain Injury in Europe: a prospective observational study within CENTER-TBI

Inge Van Erp ,¹ Hugo den Boogert,¹ Godard de Ruiter,¹ Wilco Peul,¹ Thomas van Essen,¹ Crispijn van den Brand,³ Bram Jacobs,² Joukje van der Naalt²

¹University Neurosurgical Center Holland, Leiden/Den Haag, Netherlands, ²University Medical Center Groningen, Groningen, Netherlands, ³Erasmus Medical Center, Rotterdam, Netherlands

Objectives: To study differences in demographics and outcomes according to helmet use in cyclists sustaining traumatic brain injury (TBI) across Europe.

Methods: In this prospective cohort study within CENTER-TBI, patients with TBI after a bicycle accident were included. Patients that wore helmets were compared to those that did not at the time of the injury. Primary outcome was the Glasgow Outcome Scale Extended (GOSE) at six months, which was analysed with multivariable logistic regression. Secondary outcomes included in-hospital mortality, discharge from the emergency department to home, hospital length of stay and the need for a neurosurgical intervention.

Results: A total of 511 TBI patients with a cycling accident were included, of whom 131 (26%) were wearing a helmet and 380 (74%) were not. Patients wearing a helmet had a higher Glasgow Coma Scale on admission (median 15 [IQR, 10 – 15] vs 15 [14 – 15]) and less intracranial hematomas on their first head CT scan (acute subdural hematoma 38 vs 12%) The lowest percentage of helmet use was reported in the Netherlands (5%), whereas Norway and the United Kingdom had the highest percentage of helmet use (64 and 51%). Helmet use was associated with better functional outcome (GOSE >5 83 vs. 72%, adjusted odds ratio (aOR) 2.2 [95% CI, 1.2 – 3.3]), lower in-hospital mortality (2 vs. 7%, aOR 0.5 [95% CI, 0.3 – 0.7]) and shorter hospital length of stay (median 2 days [1, 12] vs. 4 [1, 4], beta 2.6 [95% CI, 0.0 – 5.2]).

Conclusions: There is a large difference in helmet use across Europe. Helmet use leads to better functional outcomes among cyclists sustaining a TBI. Our findings highlight the importance of promoting helmet use to mitigate the debilitating impact of TBI.

P023 Age-stratified treatment variations in intracranial surgical interventions for traumatic brain injury in Europe: a prospective observational study within CENTER-TBI

MSc Rick Vreeburg ,^1,2 MSc Ranjit Singh,^1,2 Dr. Jeroen van Dijck,^1,2 MSc Hugo den Boogert,^1,2 Dr. John Yue,³ Prof. dr. Alfonso Lagares,⁴ Dr. Alexander Younsi,⁵ Dr. Godard de Ruiter,^1,2 MSc Inge van Erp,^1,2 Prof. dr. Andrew Maas,^6,7 Prof. dr. Wilco Peul,^1,2 Dr. Thomas van Essen^1,2,8

¹Leiden University Medical Center, Leiden, Netherlands, ²University Neurosurgical Center Holland, Leiden & the Hague, Netherlands, ³Department of Neurosurgery, University of California, San Francisco, United States of America, ⁴Neurosurgery department, University Hospital “12 de Octubre”, Madrid, Spain, ⁵Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Germany, ⁶Department of Neurosurgery, Antwerp University Hospital, Edegem, Belgium, ⁷Department of Translational Neuroscience, Faculty of Medicine and Health Science, University of Antwerp, Antwerp, Belgium, ⁸Department of Surgery, Division of Neurosurgery, QEII Health Sciences Center and Dalhousie University, Halifax, Canada

Background: Traumatic brain injury (TBI) is a multifaceted disease causing 1.5 million hospital admissions in Europe each year. Patient age is a prodigious predictor of outcome and important factor in surgical decision-making. We aimed to provide insights in the age-stratified treatment variations in neurosurgical interventions in patients sustaining TBI.

Methods: We selected all patients with TBI from the prospective, observational Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI) study. Differences in intracranial surgery, surgery type, timing, between-center variation in surgical likelihood were examined across age categories <15 years old, 15–24 years, 25–44 years, 45–64 years, 65–79 years and ≥80 years old. Random-effects logistic and ordinal regression with appropriate adjustment for confounding were used to describe the associations.

Results: CENTER-TBI included 4509 patients of which 1245 patients underwent 2014 intracranial procedures. Decompressive craniectomy and craniotomy for hematoma evacuation or contusion were the most prevalent acute neurosurgical interventions in all age groups (pooled 14% for <15 years, 25% for 15–24 years, 29% for 25–44 years, 33% for 45–64 years, 39% for 65–79 years and 46% for ≥80 years old). Significant between-center variations in intracranial surgery were observed for all age categories (median odds ratios 2.5 for pediatric, 3.1 for adolescents, 2.2 for young adults, 2.1 older adults, 2 elderly patients and 2.9 for octa- and nonagenarians. The odds of acute surgery for epidural hematoma, for acute subdural hematoma or contusions respectively decreased with older age (per 10-year increment; aOR 0.78, 95% confidence interval [CI] 0.69–0.88; aOR 0.93, 95%CI 0.86–1 and aOR 0.93, 95%CI 0.84–1, respectively).

Conclusions: There exist large age-stratified differences regarding intracranial surgery type and between-center variability in surgical aggressiveness. These findings offer a foundation encouraging further age-specific research and initiate discussion regarding the substantial discrepancies in clinical practices throughout Europe between younger and older patients with a TBI.

P024 Microbiological Profile of Septic Complications in Open Skull Fractures secondary to assault

Dr. Ruan Grobler ,¹ Dr. Iain Walker,¹ Prof Ian Vlok,¹ Mr. Philip Hoffman¹

¹Stellenbosch University, CAPE TOWN, South Africa

Background: Assault-related open skull fractures pose a heightened risk of septic complications, necessitating careful antibiotic selection to balance risk of septic complications and antibiotic resistance. Determining microbial profiles and influencing factors can inform antibiotic selection.

Methods: A retrospective analysis at Tygerberg Academic Hospital (2018–2021) included adults with open skull fractures from assaults who developed septic complications. Parameters such as demographics, injury details, imaging findings, surgical notes, and microbial cultures were evaluated.

Results: Out of 569 patients, 108 developed septic complications, with 60.2% yielding positive cultures. Common organisms included Staphylococcus aureus (40.0%) and Streptococcus spp. (29.2%), Enterobacter spp. (16.9%), Acinetobacter baumanii (13.8%) and Enterococcus spp (13.8%). Tissue specimens had highest culture yield (84%) and CSF the lowest (50%). Superficial wound sepsis was associated with S. aureus and Streptococcus spp., while systemic signs and brain abscesses were associated with S. aureus. (p<0.05). Staphylococcus aureus showed resistance to Amoxicillin/Clavulanic Acid, Ampicillin/Amoxicillin, and Cefuroxime and susceptibility to fluoroquinolones and trimethoprim-sulfamethoxazole was observed. Streptococcus spp. showed susceptibility to all tested antibiotics. Enterobacter spp. displayed resistance to several antibiotics but susceptibility to gentamycin, while Enterococcus spp. showed susceptibility to all tested antibiotics. Acinetobacter spp. isolates were resistant to several antibiotics, including Amikacin, Tobramycin, Imipenem, and Meropenem.

Conclusion: Compound skull fractures from assaults lead to septic complications with organisms like S. aureus, Streptococcus spp., Enterobacter spp., Enterococcus spp., and Acinetobacter baumanii. Tissue specimens had higher culture rates than CSF. Superficial wound sepsis was commonly primarily due to S. aureus and Streptococcus spp. Systemic signs of sepsis and brain abscesses were more likely due to S. aureus. Broad-spectrum antibiotics showed efficacy, but coverage for gram-negative anaerobic bacteria should be considered in treatment strategies. Antibiotic susceptibility profiles highlight the importance of tailored antibiotic regimens based on microbial culture results.

P025 Factors influencing septic complications in open skull fractures

Dr. Ruan Grobler ,¹ Prof Adriaan Johannes Vlok¹

¹Stellenbosch University, CAPE TOWN, South Africa

Background: Open skull fractures have the risk of developing life-threatening infections yet the role of prophylactic antibiotics and timing of intervention remains uncertain. The largest European series focused on early-presenting patients predominantly from traffic accidents, Glasgow 1972 and Rotterdam 1972 cited low sepsis rates of 3.5 and 4.9%. However, our study addresses a distinct population where assault, often involving contaminated weapons, is the primary cause of injury, presenting a different risk profile. The study aim was to analyze factors influencing septic complications and assess the impact of timely interventions in the largest series of open skull fractures to date.

Methods: We conducted a retrospective analysis of adults admitted with open skull fractures between January 2018 and December 2021. We examined demographic data, clinical and imaging findings, and treatment timelines. In-hospital outcomes were assessed using the extended Glasgow Outcome Score.

Results: Out of 605 open skull fractures 569 were included with an overall mortality rate of 4% (n=23). Septic complications accounted for 7 deaths, 16 were due to severity of primary injury. Assault was the leading cause of injury 95% (n=540), 4% traffic related incidents and 1% due to falls. The overall sepsis rate was 22.3%. Prophylactic antibiotic administration significantly reduced septic complications (p<0.001). CT evidence of wound contamination, air sinus involvement and time to primary wound closure are significant risk factors for developing septic complications in both uni- and multivariate analysis. Hammers was the weapon with the lowest (9.5%), and golf clubs the highest risk of developing sepsis (36.4) Time to definitive surgery >24 hours increased the sepsis rate from 2.3% to 26.4%.

Conclusions: Time to wound closure and definitive surgery > 24 hours, CT evidence of wound contamination and air sinus involvement, and lack of administration of prophylactic antibiotics contribute significantly to developing septic complications in open skull fractures.

P026 Outcomes of patients with GCS less than 12 discharged to the community following TBI

Dr Sai Phaneendra Gugamsetti ,¹ Dr Srinivas Dwarakanath¹

¹National Institute of Mental Health and Neurosciences, BANGALORE, India

Introduction: The recovery of the patients after severe TBI may follow various trajectories depending on the rate and extent of the recovery. In view of prolonged time for recovery and associated risk of secondary complications, these patients require specialized care. Unfortunately, such a care is neither uniformly available nor affordable by many in low resource countries. Objective of this study is to determine the outcomes of the patients who are in need of such specialized care discharged to the community.

Methodology: All the patients discharged from the head injury facility from July 2016 to June 2021 were retrospectively studied using hospital records and telephonic interviews. Patients with age less than 16 or with aphasia were excluded from the study. Telephonic GOSE was used as an outcome measure with GOSE of 1–3 considered unfavourable and 4–8 considered favourable.

Results: Of 1441 patients discharged from the head injury facility during the study period, 311 patients were included. 238(76.5%) patients had an unfavourable GOSE with GOSE of 1 in 222(71%) patients. The median duration of survival in patients with GOSE 1 was 30 days (Q1,Q3:3,90 days). 84%(188) of the patients with GOSE 1 were in vegetative state(GOSE 2) at the time of death. While GCS at discharge and change in GCS during hospitalization correlated positively with GOSE. GCS at admission, age, gender, socioeconomic status and hospitalization post discharge did not correlate with GOSE. However, socioeconomic status had a positive association with the duration of survival in patients with GOSE 1.

Conclusion: The outcomes of patients with low GCS depend on smooth transition to the community. Trained personnel in managing patients with head injury and specialized nursing facilities can improve the outcomes of these patients.

P027 Negative-pressure hydrocephalus following the base of skull fracture: an illustrative case and literature review

Dr. Supachai Srichantha ¹

¹Kalasin Hospital, Kalasin, Thailand

Background: Negative-pressure hydrocephalus (NegPH) is a rare condition and a phenotype of hydrocephalus that is characterized by ventriculomegaly and symptoms consistent with increased ICP in the setting of negative intracranial pressure. This case report aims to describe this patient's clinical symptoms and treatment associated with the literature.

Case Description: A Thai male patient, 26 years old, was admitted to our hospital with a TBI. The CT brain showed a thin EDH with a right mastoid fracture. The patients had right facial palsy (LMN) and CSF otorrhea. The lumbar drain was inserted. He had a fever and was diagnosed with meningitis. The antibiotics were given. Then the patient was confused and agitated, so the lumbar drain was removed. The patient’s consciousness was worse. The CT showed the progression of hydrocephalus. He underwent urgent insertion of external ventricular drainage (EVD) with a drainage height of 10 cm above the EAC (external auditory canal). After the infection was cured and the GCS was 10T, He underwent a medium-pressure VP shunt. After 2 days, the patient's consciousness was worse with a GCS of 6T. The CT showed the progression of hydrocephalus. He was suspected of having been diagnosed with NegPH. Then he underwent EVD again with a drainage height below EAC 20 cm (subatmospheric drainage). The patient was clinically improved with a GCS of 10T. The CT showed a reduction in hydrocephalus. Afterward, he underwent mastoid packing with a revision of the low-pressure VP shunt. After surgery, he was alert and had a successful weaning of the ventilator. After 3 months, he was able to walk himself to follow-up without tracheostomy, and there was no evidence of hydrocephalus.

Conclusion: NegPH is difficult to recognize. It is challenging to manage with standard strategies of CSF shunting. NegPH can be effectively treated with appropriate management.

P028 Epidemiology of Pediatric Traumatic Brain Injury in Indonesia: a Systematic Review

Tamara Tango ,¹ Alexander Yosua Santoso,² Dimas Rahman Setiawan³

¹Faculty of Medicine Universitas Indonesia, Cipto Mangunkusumo Hospital, Jakarta, Indonesia, ²Kalideres General Hospital, Jakarta, Indonesia, ³Medistra Hospital, Jakarta, Indonesia

Introduction: Traumatic brain injury (TBI) encompasses significant morbidity and mortality. Globally, the incidence of pediatric TBI ranges from 47 to 280 per 100,000 children. Nevertheless, pediatric TBI has not been extensively researched in Indonesia. This study aimed to depict the profile of pediatric TBI in Indonesia.

Methods: A systematic literature search was conducted until March 2024 using three different databases (PubMed, MEDLINE, Scopus), and grey literature, also manually searched the reference lists of the included studies with relevant keywords. Two independent authors screened and selected the literature. Information regarding the cause of injury, severity, types of intracranial injury, types of management, and mortality were extracted. The quality of the included studies was assessed using the Joanna Briggs Institute (JBI) tool.

Results: A total of 1,414 pediatric TBI patients from eight studies were included in this study, with an overall mean age of 8.7 years old. Gender proportions were mentioned only in five included studies, with males dominating at 64.6%. The most common cause of injury was road traffic accidents (n = 545), followed by others (n = 338) and falls (n = 300). About 70.3% of cases were categorized as mild TBI. Of the six included studies that detailed the types of intracranial injury, the most prevalent were epidural hematoma (n = 166), no intracranial lesions (n = 118), and cerebral concussion (n = 107). Surgery was required for approximately 23.5% of cases (222/946). From five included studies, the overall mortality rate was 2.4%.

Conclusion: By understanding the epidemiology of pediatric TBI in Indonesia, efforts should prioritize implementing prevention strategies targeting males, approximately 8.7 years old, and road traffic accidents as the primary cause. Moreover, further studies should involve a larger sample size in various locations across Indonesia to enhance representation.

P029 Elevated Skull Fractures: An Institutional Experience and Individual Participant Data Meta-Analysis

Dr Sina Zoghi ,¹ Dr. Adrina Habibzadeh,² Dr. Ali Ansari,¹ Dr. Megan E. H. Still,³ Dr. Victor M. Lu,⁴ Dr. Zahra Tabesh,⁵ Dr. Mohammad Sadegh Masoudi,⁶ Dr. Reza Taheri⁶

¹Student Research Committee, Shiraz University Of Medical Sciences, Shiraz, Iran, ²Student Research Committee, Fasa University of Medical Sciences, Fasa, Iran, ³Lillian S. Wells Department of Neurosurgery, University of Florida, Gainesville, USA, ⁴Department of Neurological Surgery, University of Miami, Jackson Memorial Hospital, Miami, USA, ⁵Noncommunicable Diseases Research Center, Fasa University of Medical Sciences, Fasa, Iran, ⁶Department of Neurosurgery, Shiraz University Of Medical Sciences, Shiraz, Iran

Background: Elevated skull fracture (ESF) is a rare but potentially life-threatening type of skull fracture. The literature on this topic is relatively sparse. Herein, we conducted a meta-analysis of all the patients reported in the literature with ESFs with respect to their clinical management to better inform practice.

Methods: On 20th of January 2023, we conducted a systematic search of literature to find all published cases of ESF. We also conducted a retrospective review of ESF cases from our institution. The data collection and analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

Results: After screening, 28 studies met the inclusion criteria. A total of 104 individual patients were included in the meta-analysis, with a median age of 24 years and 85.7% of whom were males. 11 patients (11.2%) had an unfavorable outcome while 37 (35.2%) had one or more complications. We found that GCS on admission is an independent predictor of poor outcome in ESF (odds ratio (95% confidence interval) = 1.605 (1.110–2.315), p value = 0.012). Regarding complications, dural injury (odds ratio (95% confidence interval) = 66.667 (7.407–500.00), p value < 0.001) and multiple bone involvement (odds ratio (95% confidence interval) = 6.849 (2.127–22.222), p value = 0.001) were independent predictors of complication.

Conclusion: ESFs represent a rare yet consequential form of cranial injury, carrying potentially life-threatening implications if not promptly addressed. In this study, we present the meta-analysis of outcomes and complications within this patient cohort, offering a comprehensive synthesis of existing literature on this pathology. However, further investigation is imperative to provide higher-quality evidence and address lingering uncertainties in the classification and management of ESFs.

P030 Neuromotor recovery is associated with gut dysbiosis following surgical decompression for Degenerative Cervical Myelopathy

Pía Vidal ,¹ PhD Candidate Sydney Brockie,² PhD Carlos Farkas,¹ PhD Ariel Avila,¹ PhD, MD Michael Fehlings³

¹UCSC, Concepcion, Chile, ²University of Toronto, Toronto, Canada, ³University of Toronto and University Health Network - TWH, Toronto, Canada

Introduction: Degenerative cervical myelopathy (DCM) describes a spectrum of disorders that cause progressive and chronic cervical spinal cord compression. The clinical presentation can be complex and can include locomotor impairment, hand and upper extremity dysfunction, pain, loss of bladder and bowel function, as well as gastrointestinal dysfunction. Once diagnosed, surgical decompression is the recommended treatment for DCM patients with moderate to severe impairment.

Our body is composed of a large community of microorganisms, known as the microbiota. Traumatic and non-traumatic spinal cord injuries (SCIs) can induce changes in the gut microbiota and gut microbiota derived metabolites. These changes have been reported as important disease-modifying factors after injury. However, whether gut dysbiosis is associated with functional neurological recovery after surgical decompression has not been examined today.

Methodology: DCM was induced in C57BL/6 mice by implanting an aromatic polyether material underneath the C5–6 laminae. The extent of gut dysbiosis was assessed by gas chromatography and 16S rRNA sequencing from fecal samples before and after decompression. Neuromotor activity was assessed using the Catwalk test.

Results: DCM pre- and post- surgical decompression is associated with gut dysbiosis, without altering short chain fatty acids (SCFAs) levels. Significant differences in Clostridia, Verrumicrobiae, Lachnospiracea, Firmicutes, Bacteroidales, and Clostridiaceae were observed between the DCM group (before decompression) and after surgical decompression (2 and 5 weeks). The changes in gut microbiota composition correlated with locomotor features of the Catwalk. For example, a wider basal support and dysfunctional step cycle were positively correlated with gut dysbiosis.

Conclusions: Our results show that gut dysbiosis is associated with functional neurological recovery after surgical decompression.

P031 Selective Cranial-Brain Cooling System for Improved Outcomes in Traumatic Brain Injury Patients

Ing. Bernardo Yelicich ,¹ Dr Alberto Biestro,¹ Dr Leandro Moraes,¹ Dra. Corina Puppo,¹ Ing. Héctor Gómez,¹ Dr. Federico Salles,¹ Dra Matilde Lisarrague,¹ Dr Mauricio Mercado¹

¹Universidad De La República. Hospital de Clínicas, Montevideo, Uruguay

Introduction: Traumatic brain injuries (TBI) continue to be a major cause of death necessitating urgent attention and thorough care in the Intensive Care Unit (ICU) for improved patient outcomes and reduced long-term complications. Recent interest lies in decompressive craniectomy (DC) and hypothermia as possible therapeutic approaches. Nevertheless, systemic hypothermia can yield unwanted side effects. Hence, a novel concept for a selective cranial-brain cooling system seeks to regulate elevated intracranial pressure (ICP) without instigating systemic hypothermia.

Keywords: selective hypothermia, head injury, temperature, brain cooling.

Methodology: The research centers around building a selective cranial-brain cooling system equipped with a helmet in which circulates chilled fluid coolant. An external monitoring unit controls temperatures across the whole apparatus. Major components incorporate a portable freezer, copper coils, a peristaltic pump, a water reservoir, and a monitoring and control system. Initial experiments utilized a stainless-steel container surrounded by a thermal insulation box lined with water to simulate the skull

Results: Preliminary testing displayed positive results, with the system delivering adequate cooling power accounting for inherent operational losses. Additional advancements included crafting a lifelike head phantom for forthcoming stages, including a craniectomized skull, synthetic outer layer, authentic dura matter steeped in formaldehyde, and collagen hydrogel mimicking neural material. Experiments affirmed consistent cooling abilities, implying the viability of employing this selective cranial-brain cooling strategy in prospective analyses.

Conclusions: Our experimental prototype has demonstrated the potential for effectively cooling target areas in the brain, offering promise for mitigating ICP in TBI patients undergoing DC procedures. Non-invasive selective cranial-brain cooling presents itself as a viable supplementary therapy worthy of further investigation. While this research addresses a pertinent concern surrounding TBI management, several challenges persist, including optimization of cooling duration and intensity, ensuring minimal invasiveness, and maintaining compatibility with existing neurocritical care paradigms.

P032 Completion of Massive Open Online Course Improved Global Distribution of Knowledge about Traumatic Brain Injury

Dr Yasmine Doust,¹ Dr Hannah Fair,¹ Dr Claire Eccleston,¹ Dr Christine Padgett,¹ Dr. Kathleen Doherty,¹ Dr Peta Cook,¹ Dr Jenna Ziebell ¹

¹University of Tasmania, Hobart, Australia

The understanding traumatic brain injury massive open online course (TBI MOOC) was developed to improve awareness and reduce misconceptions of TBI in the community. This study assessed the impact of the TBI MOOC on participant's misconceptions of TBI. Participants (n=1,267) were from 45 different countries that represented all 6 continents except for Antarctica where 91.7% were from high income countries. Most of the study population were women (82.6%) with a mean age of 45.8 years (18 – 83 years) and have completed tertiary education or above (65.8%). In the March 2023 iteration of the TBI MOOC, 10.5% of the respondents indicated that that they had sustained a TBI whilst 45.7% responded that a family member is living with a TBI. Additionally, 48.9% of all study participants stated that they are currently providing paid care for individuals living with a TBI. Participants who exhibited the lowest scores on the misconceptions about TBI survey, and thus lower TBI knowledge, prior to completing the MOOC had either lower educational levels, never provided paid care, are currently providing unpaid or voluntary care, have never personally sustained a TBI or had a family member who has sustained a TBI. However, after completing the MOOC all participants, regardless of group, had similar scores indicating that the MOOC is an effective tool to reduce misconceptions of TBI in the community.

P033 Decompressive Craniectomy Requirement in Traumatic Brain Injury Patients, Predicting Factors, and Short- and Long-Term Outcomes: A Retrospective Analysis of the Shiraz TBI Database

Dr Sina Zoghi ,¹ Dr. Ali Ansari,¹ Dr. Victor M. Lu,² Dr. Omid Yousefi,³ Dr. Reza Taheri,³ Dr. Hosseinali Khalili³

¹Student Research Committee, Shiraz University of Medical Sciences, Shiraz, Iran, ²Department of Neurological Surgery, University of Miami, Jackson Memorial Hospital, Miami, USA, ³Department of Neurosurgery, Shiraz University of Medical Sciences, Shiraz, Iran

Background: Brain edema following Traumatic Brain Injury (TBI) can raise intracranial pressure. Decompressive Craniectomy (DC) is the surgical modality of choice for managing high intracranial pressure not responsive to medical intervention. The objective of this study was to survey a large institutional database to determine the trends of DC, clinical characteristics of patients who require DC, and the outcome of patients who underwent DC.

Methods: We reviewed TBI patients admitted to our center from 2015 to 2021 from our prospectively maintained registry. Demographic data, mechanism of injury, findings on admission, neuroimaging findings, DC requirement, procedures needed during hospitalization, and functional outcome at discharge and six-month follow-up were gathered.

Results: A total of 4011 patients were included in this study, with 506 patients undergoing a primary DC. International Normalized Ratio, activated Partial Thromboplastin Time, subdural hematoma, midline shift, epidural hematoma, intracerebral hemorrhage, compressed or absent basal cisterns, Rotterdam score, and Marshall score were independently associated with DC requirement. Furthermore, DC requirement was independently associated with the need for tracheostomy. In terms of outcome for patients who required DC, older age, lower Hb levels, higher Rotterdam scores, the presence of compressed and absent basal cisterns, and the need for tracheostomy were independently associated with unfavorable outcome in individuals with mild to moderate TBI. On the other hand, lower Glasgow Coma Scale score and fixed pupills on examination were independently associated with unfavorable outcome in patients with severe TBI. Interestingly, the requirement for DC in cases of subdural hematoma showed an inverse independent association with unfavorable outcome.

Conclusion: This study is one of the most comprehensive comparative analyses conducted to date, examining the characteristics of patients requiring DC and the outcome of those who underwent this intervention.

P034 A machine learning algorithm to predict the recurrence of chronic subdural hematoma using computed tomography images

Dr Daisu Abe ,¹ Dr Motoki Inaji,¹ Mr Takeshi Hase,¹ Dr Taketoshi Maehara¹

¹Tokyo Medical And Dental University, Bukyo-ku, Japan

Background: Chronic subdural hematoma (CSDH) has been reported to recur in 10–20% of surgical cases. In this study, we developed a machine learning model to predict the recurrence of CSDH.

Subjects/Methods: 200 patients who underwent surgical treatment for CSDH at our institution were enrolled. We extracted preoperative computed tomography (CT) image features using a convolutional autoencoder and created a prediction model.

Results: Recurrence was observed in 36 of 200 cases. As a result, we built a prediction model with an area under the Receiver Operating Characteristics curve (ROC-AUC) of 0.82. Prediction results for test data were an ROC-AUC of 0.67, a sensitivity of 82%, and a specificity of 49%.

Discussion: By using machine learning techniques, quantitative and objective prediction of the recurrence of CSDH may be possible from preoperative CT images. In the future, it is necessary to increase the number of cases and build a more versatile model.

P035 Magnetic resonance imaging in traumatic brain injury: A survey of clinical practitioners’ experiences and views on current practice and obstacles

Professor Karen Caeyenberghs ,¹ Dr Mervyn Singh,¹ Annalee Cobden,¹ Elizabeth Ellis,^1,2 Liam Graeme,¹ Dr Priscilla Gates,^1,3 Dr Alex Burmester,¹ Jade Guarnera,¹ Jake Burnett,^1,4 Evelyn Deutscher,¹ Lyndon Firman-Sadler,¹ Bec Joyce,¹ Jacqueline Notarianni,¹ Christian Flores,¹ Dr Juan Dominguez¹

¹Deakin University, Melbourne, Australia, ²University of Turku, Turku, Finland, ³Peter MacCallum Cancer Centre, Melbourne, Australia, ⁴St Vincent’s Hospital, Melbourne, Australia

Introduction: Magnetic resonance imaging (MRI) has revolutionized our capacity to examine alterations in brain structure and function in patients with traumatic brain injury (TBI). However, there is little knowledge about the current level of implementation of MRI techniques in the clinical practice in TBI patients and the obstacles that clinicians experience to apply MRI. Here we report on results from a survey seeking to examine clinical practitioners’ experiences and views on the use of MRI in TBI.

Methods: The survey included 19 multiple choice and free text questions grouped in four sections: (1) Demographics and MRI Experience; (2) MRI Scan Administration; (3) Application of MRI in TBI patients; and (4) Factors influencing the use of MRI in clinical translation of TBI. The survey was intended for a diverse set of health professionals.

Results: Among 81 respondents, 73.4% indicated that they acquire or order MRI scans in TBI patients and 66% indicated that they would prefer if MRI was more often used with this cohort. The greatest impediment for MRI usage in TBI was the availability of the MRI scanner (57.1%). MRI was reportedly performed most often during the sub-acute stage (57.4%), followed by the acute stage (38.3%), and the chronic stage (36.2%). Only 27.8% of respondents indicated that a clinical protocol was used for imaging acute TBI patients. Less than half of respondents (42.1%) indicated that they perform advanced MRI analysis. Factors, such as dedicated experts within the team (44.4%), user-friendly tools to analyse the MRI scans (40.7%), were listed as potentially helpful to implement advanced MRI analyses in clinical practice.

Conclusions: Results suggest a wide variability in the purpose, timing, and composition of the scanning protocol of clinical MRI after TBI. Addressing obstacles could catalyse a broader implementation of MRI in the clinical practice in TBI patients.

P036 MRI and EMG technologies in assessing the effectiveness of surgical treatment of cerebrospinal fluid dynamics disorders in the craniovertebral region

M.D. OLEG DULUB ,¹ Ph.D. Inessa Ilyasevich,¹ M.D. Elena Soshnikova,¹ Doctor Sergey Korchevsky¹

¹The Republican Scientific And Practical Centre Of Traumatology And Orthopedics, Minsk, Minsk, Belarus

Introduction: Disturbances of cerebrospinal fluid dynamics in the craniovertebral region are serious complications of a number of congenital diseases, CNS, and consequences of vertebrospinal injury.

Methodology: A prospective cohort study includes 59 operated patients with craniovertebral stenosis, Arnold-Chiari types I and II malformations with or without associated syringohydromyelia, 12 patients with an extended and total forms of posttraumatic syringohydromyelia. Neurological deficits ranged from pyramid symptoms to severe bulbar and cerebellar disturbances. The MRI algorithm for postsurgical examination includes flow programs, functional MRI, contrast MRI. The EMG study was based on motor evoked potentials (MEP), somatosensory evoked potentials (SEP), and blink reflex evoked potentials.

Results: Qualitative and quantitative assessment of MRI is particularly valuable in evaluating regression or increasing of structural changes, transient compression of the brain, and excluding instability. The positive dynamics of MEP indicators, depending on the group of patients, was characterized by normalization of amplitude to control values (6.6±2.9 mV or its increase and stabilization at the lower limits of control (from 1.4±0.8 mV to 3.6±2. 2 mV) from 6–12 months after surgery. Positive dynamics of SEP time parameters (N9, N13, central conduction time N20-N13) was determined by 12–24 months after surgery. The use of the original blink reflex technique made it possible to verify lesions in the C0 - C4 spinal cord segments, and the effectiveness of cisterno-peritoneal shunting in syringohydromyelia.

Conclusions: Modern MRI techniques, evoked potentials EMG made it possible to achieve an optimal clinical result; to perform revision surgeries timely.

P037 Association of diffuse axonal injury with outcome in moderate to severe traumatic brain injury patients

Radina Lilova ,¹ Owen P. Leary,² Dr. Tyler J. Harder,³ Dr. David W. Wright,⁴ Dr. Lisa H. Merck^1,2

¹Virginia Commonwealth University, Richmond, United States, ²Brown University, Providence, United States, ³Beth Israel Deaconess Medical Center, Boston, United States, ⁴Emory University, Atlanta, United States

Introduction: Diffuse axonal injury (DAI) contributes to neurologic deficits after traumatic brain injury (TBI) however, DAI remains frequently underdiagnosed. Co-occurrence of TBI phenotypes is examined across patients from a prospective, multicenter clinical trial and correlated with outcomes.

Methodology: Hyperacute CT imaging, completed within 4 hours of injury, was examined from moderate-to-severe TBI patients enrolled in the ProTECTIII multicenter clinical trial and reviewed by neuroradiology (n=881). Identified imaging phenotypes included: DAI, epidural hematoma (EDH), subdural hematoma (SDH), subarachnoid hemorrhage (SAH), intraventricular hemorrhage (IVH), intraparenchymal hemorrhage (IPH), and contusion. Patient cohorts were defined: no injury (n=154), isolated DAI (n=26), DAI + other hemorrhage (n=221), and isolated other hemorrhage (n=480). Affected anatomical regions were identified. Phenotypes were correlated with 6-month outcome data (n=833); poor outcome was defined by sliding dichotomy (Disability Rating Scale/Extended Glasgow Coma Scale) per original study. Multiple linear regression analyses were used to define statistical relationships across groups p<0.05 (RStudio,1.4).

Results: DAI in isolation was not associated with increased risk of mortality/poor outcome. However, co-occurrence of DAI and other injury phenotypes, was associated with lower index GCS (iGCS) (6.33±0.127 SEM, p=0.019). Significant events of co-occurrence were identified between DAI and: EDH (p<0.001), SDH (p<0.001), IVH (p<0.001), contusion (p=0.012). Co-occurrence of intra-axial lesions and DAI was more frequently associated with poor outcome (+DAI: OR=2.50, 95% CI 1.32–4.74; - DAI: OR=1.99, CI 1.15–3.43).

Conclusions: Patients with co-occurrence of DAI and other bleed phenotypes on hyperacute CT presented with lower iGCS. Isolated DAI was not associated with increased morbidity/mortality, when compared to other phenotypes of injury. Differences in frequency of co-occurring bleed phenotypes were found across DAI+ and DAI- cohorts, with highest co-occurrence between DAI, SDH, and IVH. Identifying imaging-based predictors of outcome after TBI provides insight to effective classification, patient triage, and management.

P038 Optimization and multicentre external validation of a deep learning model for segmentation and quantification of traumatic brain injury lesions on head CT

Dr Francois Mathieu ,¹ Carolina Picarra,² Armaan Malhotra,¹ Miguel Monteiro,² Stefan Winzeck,² Christopher Smith,⁴ Dr. Virginia Newcombe,³ Dr. David Menon,³ Dr. Christopher Witiw,^1,4 Dr. Damon Scales,⁵ Dr. Farhad Pirouzmand,¹ Dr. Ben Glocker²

¹Division of Neurosurgery, Department of Surgery, University of Toronto, Toronto, Canada, ²Biomedical Image Analysis Group, Department of Computing, Imperial College London, London, United Kingdom, ³Division of Anaesthesia, Department of Medicine, University of Cambridge, Cambridge, United Kingdom, ⁴Li Ka Shing Knowledge Institute, St. Michael's Hospiital, Toronto, Canada, ⁵Interdepartmental Division of Critical Care, Department of Medicine, University of Toronto, Toronto, Canada

Introduction: We previously proposed BLAST-CT, a convolutional neural network able to perform multiclass segmentation and quantification of TBI lesions on CT. In this study, we 1) optimize the performance of our model using additional training data from CENTER-TBI 2) externally validate our model using an independent imaging dataset from the PROTEST randomized trial.

Methods: 680 scans from CENTER-TBI were annotated by experts and used to retrain the CNN, creating BLAST version 2.0. Traumatic lesions were subdivided into 4 classes: intraparenchymal (IPH), extra-axial (EAH) and intraventricular hemorrhage (IVH) and perilesional edema. 51 scans from PROTEST were randomly selected and manually annotated to obtain ground-truth lesion labels on an independent dataset. The same PROTEST scans were then contemporaneously run through version 1.0 and 2.0 of BLAST-CT to evaluate the performance change resulting from the optimization procedure while calculating segmentation accuracy metrics on an external validation dataset.

Results: The additional training implemented for version 2.0 yielded an overall mean Dice similarity coefficient improvement of 4%. Mean absolute volume errors between automated and ground-truth segmentations also improved for most lesion types (2.94 vs 1.55mL for IPH, 18.44 vs 16.33mL for EAH, 0.74 vs 0.80 for IVH and 1.56 vs 0.27 for perilesional edema using version 1.0 vs 2.0 respectively). Overall, the performance of BLAST-CT on the PROTEST external validation dataset was comparable or better to the results obtained on our internal development sample (median DSC was 0.60 [IQR 0.0–0.94] on PROTEST images vs 0.36 [IQR 0.0–63.4] on the CENTER-TBI development dataset).

Conclusion: We propose one of the first models capable of automated multiclass volumetric lesion segmentation in TBI to be trained and externally validated in a multicentre fashion. After optimizing our model using a large additional training sample from CENTER-TBI, we were able to achieve a level of performance comparable to other state-of-the-art methods.

P039 TimesNet in TBI Research: Leveraging DTI Radiomics and Multivariate Time Series Biomarkers for Ordinal GOSE Classification

Eugenia Akpo,^1,2 Lisa Inyange,^1,2 Patience Bwire,^1,2 Harri Merisaari,^3,4 Jussi Posti,^5,8 Olli Tenovuo,^6,8 Peter Ngum ^1,6,7,8

¹Carnegie Mellon University Africa, Kigali, Rwanda, ²Pan African AI Health Initiative, United States, ³Turku Brain and Mind Center, University of Turku, Turku, Finland, ⁴Department of Diagnostic Radiology, University of Turku, Finland, ⁵Neurocenter, Department of Neurosurgery, Turku University Hospital, Turku, Finland, ⁶Department of Clinical Neurosciences, University of Turku, Turku, Finland, ⁷Johns Hopkins University, Carey Business School, Baltimore, United States, ⁸Turku Brain Injury Center,Turku University Hospital, Turku, Finland

Introduction: Traumatic brain injury (TBI) presents complex diagnostic and prognostic challenges that necessitate an integrated approach for optimal patient management. Reflecting insights from the January 2024 NINDS reclassification meeting, this study adopted a multimodal strategy combining clinical assessment, neuroimaging, and blood-based biomarkers (BBMs) to predict functional recovery on the ordinal Glasgow Outcome Scale Extended (GOSE). This approach aligns with the consensus that merging clinical observations, biomarker data, and imaging results offers a comprehensive framework for understanding TBI and its trajectory.

Methodology: The methodology analyzed data from 200 TBI patients and 40 controls using the Finnish TBIcare dataset. Predictive models were developed with TimesNet, a state-of-the-art time series analysis framework that transforms complex 1D biomarker and diffusion tensor imaging (DTI) radiomic features into an interpretable 2D space. DTI radiomics, extracted via Pyradiomics software, quantified post-injury tissue microstructure changes. These features and GFAP and UCHL1 BBMs predicted outcomes across four frameworks: ordinal GOSE prediction, ordinal with orthopedic control inclusion, multiclass GOSE categorization, and multiclass including controls. Model performance was evaluated using F1 scores.

Results: Incorporating DTI significantly improved TimesNet's performance across all frameworks. The fused models combining clinical assessments, BBMs, and imaging achieved impressive F1 scores of 0.96, 0.96, 0.94, and 0.94 for the respective GOSE categories. In contrast, models utilizing only clinical assessments and BBMs yielded lower F1 scores of 0.43, 0.65, 0.72, and 0.85. This enhancement underscores the efficacy of integrating multiple data modalities.

Conclusion: While the results support a multimodal approach's utility in improving TBI recovery trajectory understanding, the exceptionally high F1 scores warrant cautious interpretation, considering potential overfitting and small sample constraints.

P040 An Automated Algorithm for Intracranial Hematoma Volume Calculation

Dr Sina Zoghi ,¹ Mr. Sepehr Khademolhosseini,² Dr. Adrina Habibzadeh,³ Dr. Reza Taheri,² Dr. Amin Niakan,² Dr. HosseinAli Khalili²

¹Student Research Committee, Shiraz University Of Medical Sciences, Shiraz, Iran, ²Shiraz Trauma Research Center, Shiraz, Iran, ³Student Research Committee, Fasa University of Medical Sciences, Fasa, Iran

Background: Intracranial hemorrhage (ICH) is a severe condition that requires rapid diagnosis and treatment. Automated methods for calculating ICH volumes can reduce human error and improve clinical decisioPlease provide professional degrees (e.g., PhD, MD) for the corresponding author.n-making. A novel automated method has been developed that is comparable to the ABC/2 method in terms of speed and accuracy while providing more accurate volumetric data.

Methods: We developed a novel automated algorithm for calculating intracranial blood volume from computed tomography (CT) scans. The algorithm consists of a Python script that processes Digital Imaging and Communications in Medicine images and determines the blood volume and ratio. The algorithm was validated against manual calculations performed by neurosurgeons.

Results: Our novel automated algorithm for calculating intracranial blood volume from CT scans demonstrated excellent agreement with the ABC/2 method, with a median overall difference of just 1.46 mL. The algorithm was also validated in patient groups with ICH, epidural hematoma (EDH), and SDH, with agreement coefficients of 0.992, 0.983, and 0.997, respectively.

Conclusions: The study introduces a novel automated algorithm for calculating the volumes of various ICHs (EDH, and SDH) within CT scans. The algorithm showed excellent agreement with manual calculations and outperformed the commonly used ABC/2 method, which tends to overestimate ICH volume. The automated algorithm offers a more accurate, efficient, and time-saving approach to quantifying ICH, EDH, and SDH volumes, making it a valuable tool for clinical evaluation and decision-making.

P042 NEUROPROTECTION AGAINST SROKE MEDIATED BY REMOTE CONDITIONING DEPENDS ON THE INFLAMMATORY CONDITION PREVALENCE IN THE RAT MODEL OF COVID-19

Jana Koncekova ,¹ Klaudia Kotorova, Miroslava Nemethova, Petra Bonova

¹Biomedical Research Centre of SAS, Institute of Neurobiology, Kosice, Slovakia

Introduction: Infection by SARS-CoV-2 causes the release of pro-inflammatory cytokines responsible for abnormal levels of coagulation factors and thrombus formation. Despite increased risk of stroke, treatment with t-PA is less effective in these patients. Remote ischemic conditioning (RIC) stimulates endogenous protective mechanisms ensuring higher resistance to stroke and may be effective therapy for COVID-19 positive patients with stroke.

Methodology: COVID-19-inflammation was simulated by intratracheal administration of lipopolysaccharide (LPS). Cerebral ischemia was induced in two intervals; 6 hours (acute phase) and 24 hours after LPS administration (development of hyper-inflammatory reaction). RIC, in the form of hind-limb ischemia, was applied at one hour of reperfusion. Evolution of inflammation was characterized by monitoring of plasma cytokines level, sedimentation rate, hematocrit and clotting time. Efficacy of RIC treatment was evaluated by infarct size measuring.

Results: The results showed neuroprotective potential of RIC 24 hours, but not 6 hours following the inflammation induction. Early stages of LPS intoxications results in the increase of chemokines level (CINC-1, ICAM-1, LIX, CXCL7). During acute phase, RIC treatment of brain injury was not effective. On the other hand, RIC applied after 24 hours of inflammation reduced infarct size about 50%. At this time LPS-inflammation was developed, resulting in the decrease of hematocrit by about 15%, 2.5 fold boost in sedimentation rate and decline of the clotting time by half. However, plasma chemokines level (CINC-1, LIX, RANTES) significantly dropped compared to acute phase.

Conclusions: We can conclude that proper timing of RIC treatment is effective in reduction of infarction despite the presence of inflammation. However, based on our results it could be assumed, that plasma level of chemokines could affect the efficacy of RIC. This observation would be helpful in arrangement of proper timing of RIC treatment in patients exhibiting hyper-inflammatory reaction, like COVID-19.

Supported by APVV-21-0069 and VEGA 2/0096/22.

P043 Increased incidence and mortality of civilian penetrating traumatic brain injury in Sweden

Dr. Robert Lilford ¹

¹Karolinska Institutet / Stockholm South General Hospital, Stochkolm, Sweden

Background: Penetrating trauma to the head and neck (HN) has increased during the past decade in Sweden. This aim of this study was to characterize these injuries and to evaluate the outcome in patients treated at a tertiary trauma centre.

Methods: Swedish trauma registry (SweTrau) data was extracted on patients with HN injuries admitted to Karolinska University Hospital, Stockholm, between 2011 and 2019. Outcome information was extracted from hospital records, with primary endpoints focusing of physiological outcome measures and secondary endpoints on surgical and radiological outcomes.

Results: Of 1436 penetrating trauma patients, 329 patients with penetrating HN injuries were identified. 20% (n=66) suffered a gunshot wound (GSW), 73% (n=240) a stab wound (SW), and 7% (n=23) other trauma mechanisms (OTM). The median age for GSW, SW and OTM were 25, 33, and 21, respectively. Assault was primary intent, GSW (81.8%, n=54) and SW (65.8%, n=158). Patients suffering GSWs had severer injuries, worse admission GCS-M and higher intubation rate at the injury site. Most GSW patients underwent major surgery (59.1%) as an initial procedure and more likely had intracranial haemorrhage (ICH) (21.2%). The 30-day mortality: 45.5% (n=30) for GSWs, 5.4% (n=13) for SWs and 0% (n=0) for OTMs. There was a yearly increase in incidence and mortality for GSW and SW.

Conclusion: Between 2011 and 2019, there was an increasing yearly trend of incidence and mortality from penetrating HN trauma in Stockholm, Sweden. GSW patients suffered more severe injuries, ICHs, and underwent more surgical interventions compared to SW and OTM.

P044 Bilateral epidural hematoma in the posterior fossa after decompressive craniectomy in severe traumatic brain injury (TBI)

Aleksandra Maiwald ,¹ Prof. Dr. med. Oliver Schnell, Dr. med. Sven-Martin Schlaffer

¹Erlangen University Hospital, Germany, Erlangen, Germany

Background: We present a case of bilateral EDH of the posterior fossa due to brain herniation in a 34-year-old female patient who sustained severe TBI following a high-speed car collision while jogging. On arrival at the hospital, the patient had a GCS score of less than 8 and required intubation. Imaging revealed multiple skull fractures and a generalized cerebral edema, so an EVD was placed to monitor intracranial pressure. Despite maximal conservative measures, the patient’s ICP continued to rise. A further CT-scan showed a progressive acute subdural hematoma leading to an instantly surgical hematoma evacuation and decompressive hemicraniectomy on the left side. However, during the procedure, a severe brain herniation occurred. After temporary skin closure with artificial materials a CT-scan showed in a bioccipital epidural hematoma.

Immediate surgical intervention was performed to remove the haematoma and extend the decompressive surgery. Postoperatively, the patient suffered an ischaemic and haemorrhagic infarction in the left hemisphere. The patient eventually stabilised and was transferred to a rehabilitation facility with significant deficits, including global aphasia and severe hemiparesis.

Over the course of two years, the patient showed a clear improvement in motor functions and speech comprehension, although residual deficits remained.

Results: Despite the complex and life-threatening nature of the initial injury and subsequent complications, the patient's overall outcome was considered more favourable than expected.

Methods: Case report.

Conclusion: This case illustrates the rare occurrence of epidural haematoma of the posterior fossa following decompressive craniectomy for severe traumatic brain injury and highlights the challenges in treating such cases. The lack of a clear source of haemorrhage in this case underlines the unpredictability of such complications in patients with traumatic brain injury. Early recognition and prompt surgical intervention were critical to the management of this difficult case and a relatively positive outcome for the patient.

P045 Long-term histopathological changes after traumatic brain injury in swine that did or did not develop post-traumatic epilepsy

Linda Kwakman,¹ Benjamin Baskin,¹ Dr. Kevin Staley,^1,2 Dr Declan McGuone ,³ Dr. Beth Costine-Bartell^1,2

¹Massachusetts General Hospital, Boston, United States, ²Harvard Medical School, Boston, United States, ³Yale School of Medicine, New Haven, United States

It is not known why some individuals with TBI go on to develop post-traumatic epilepsy (PTE) and some do not. Histopathological analysis of acute and chronic changes was performed to find clues as to the reason for the development of PTE. We examined the brains of castrated, male swine 4 - 12 months after a single, bilateral cortical impact (n = 10) or sham surgery (n = 3) at 5 - 6 months of age. Half of the swine developed PTE an average of 5.2 ± 1.6 months after cortical impact. The extent of neuronal drop out, gliosis, hemosiderin containing macrophages, chronic white matter degeneration, and “hypoxic-ischemic” red neurons was greater in pigs that did not develop PTE vs. those pigs that did develop PTE. A single instance of perivascular albumin extravasation was observed in a single pig (cortical impact with PTE). Atrophic changes as measured by wet brain weight, ventricle size, cortical thickness, gray matter area, and white matter area were not different among cortical impact pigs with or without PTE. A subset of swine with and without PTE demonstrated subpial gliosis presumed to be related to screws installed into the skull used to attach the EEG leads. Work is ongoing to quantitate macrophages (CD68), activated microglia (Iba1; activated morphology) and T-cells (CD3) in the remodeled cortex as well as determining the expression of phosphorylated tau (PHF-1). In conclusion, there was no gross nor microscopic structural pathologic changes unique to swine that developed PTE though there might be subtle changes in chronically activated immune cells and phosphorylated tau.

This work is supported by Citizens United for Research in Epilepsy and P01 NS127769.

P046 A reflection on the value of data collection for a Spinal Cord Injury Hub service designed to improve services while patients await a specialist Spinal Cord Injury Centre

R Rowlingson,² K Alexander,² J Ferreira,² Dr F Anwar,² Dr Harry Mee ^1,2

¹University of Cambridge, Cambridge, United Kingdom, ²Cambridge University Hospital NHS Foundation Trust, Cambridge, United Kingdom

The Cambridge University Hospital (CUH) Spinal Cord Injury (SCI) Hub is one of three designated hubs in the region established following the East of England SCI transformation project. It is designed to improve patient care by providing early access to specialist services in an acute hospital setting.

The service has been running for 2 years and comprises a Specialist Clinical Nurse, a Physiotherapist, and an Occupational Therapist. It also has access to a Clinical Psychologist who covers the region and support from rehabilitation medicine. It has four rehabilitation beds that provide input while patients await transfer to a Spinal Cord Injury Centre (SCIC). Patients receive intensive rehabilitation and education regarding their SCI.

Methods: Data has been routinely collected on patients under the care of the SCI Hub.

Information on the referral times, response times, outcomes, and wait times for SCICs has been collected.

Initial Asia scores are collected, as are the incidence of any pressure ulcers.

The Spinal cord Independence Measure (SCIM) was completed upon admission and discharge from the hub for all patients.

Results: Data from before the service was minimal. Now, we have a range of information that gives a clear picture of the current situation and challenges.

The SCIM and discharge destinations have shown that we are able to demonstrate functional improvements and discharge some patients to their homes without requiring inpatient rehab at a specialist centre.

Conclusions: Data collection is invaluable in establishing the current situation of the rehabilitation pathway for the SCI patient, the challenges faced by services and patients, and the effectiveness of the SCI Hub service.

P047 A Core Outcome Set for CrAnioplasty Following Stroke or Traumatic Brain Injury - COAST Study

Dr Harry Mee ,^1,3 Dr A Castano Leon,² Dr F Anwar,³ Ms K Grieve,³ Mrs N Owen,³ Mrs C Turner,¹ Ms G Whiting,¹ Mr E Viaroli,^1,3 Mr I Timofeev,³ Mr A Helmy,^1,3 Professor P Hutchinson,^1,3 Mr A Kolias^1,3

¹University of Cambridge, Cambridge, United Kingdom, ²Department of Neurosurgery, Instituto de Investigación Sanitaria Hospital 12 de Octubre, Hospital Universitario 12 de Octubre, Madrid, Madrid, Spain, ³Cambridge University Hospital NHS Foundation Trust, Cambridge, United Kingdom

Introduction: There is significant heterogeneity in the reporting of outcomes in cranioplasty literature. This study aimed to establish a core outcome set (COS) for cranioplasty after decompressive craniectomy for stroke or traumatic brain injury.

Methodology: The scope was defined according to the criteria recommended by the Core Outcome Measures in Effectiveness Trials (COMET) Initiative. Phase 1 focused on outcome gathering through a systematic review and qualitative study. Phase 2 focused on consolidation and consensus of outcomes through a two-round Delphi survey and a consensus meeting. Participants from the four stakeholder groups individually scored all the outcomes on a 9-point Likert scale. Variables that did not reach the predefined consensus threshold for COS inclusion or exclusion were voted upon in the final consensus meeting.

Results: 208 verbatim outcomes were consolidated into 56 domains. A total of 153 participants completed round 1, with 45 additional outcomes suggested for inclusion. Four were included in Round 2 following rationalisation. 109/153 participants (71%) from 16 countries completed Round 2 and re-scored all 60 outcomes (56 original + 4 additional). Nine outcomes were voted in, and 12 were voted out from Delphi. The remaining 39 were discussed at a consensus meeting, with 11 voted in. The final COS included 20 outcomes (9 + 11) across four domains: life impact, pathophysiological manifestations, resource use/economic impact, and mortality.

Conclusion: COAST COS covers key cranioplasty outcomes, as assessed by international stakeholders consisting of surgical, medical, rehabilitation, and nursing professionals and patients and their relatives. Future implementation will aid in the standardisation of outcomes and facilitate the development of cranioplasty-specific outcome measures, aiding between-study comparisons and improving the relevance of trial findings to healthcare professionals and patients.

P048 Intranasal Delivery of miR-9-5p Mimics Attenuates Inflammatory Response After a Repetitive Concussive Traumatic Brain Injury

Dr. Manish Bhomia ,¹ Piper Deleon,¹ Yanru Feng,¹ Barbara Knollmann-Ritschel¹

¹Uniformed Services University of The Health Sciences, Bethesda, United States

Introduction: There are no effective therapeutics for traumatic brain injury (TBI), and therefore many TBI patients experience long-term neurological impairments. A novel therapeutic is needed to reduce mortality and long-term morbidity associated with TBI. MiR-9 is a microRNA primarily expressed in cells of the central nervous system (CNS) and is important for CNS development and maintenance. Our studies have shown increased miR-9 expression in blood and brain after a TBI. The central hypothesis for this study is that delivering miR-9-5p intranasally after a TBI will reduce neuroinflammation and improve outcome.

Methods: C57BL/6J mice, aged 8–10 weeks, were categorized into four treatment groups: uninjured control treated with lipid nanoparticles (LNPs) and 1X PBS, injured control treated with only LNPs, injured mice treated with LNP-miR-9-5p inhibitor, and injured mice treated with LNP-miR-9-5p mimic. Repetitive concussive brain injury (RCBI) was induced using a controlled cortical impact (CCI) device three times over three days. LNP-conjugates were administered intranasally via pipette. Blood samples were collected on days 4 and 7 of the study. Physical parameters and pro-inflammatory cytokine analysis were performed.

Results: A significant weight gain in mice treated with miR-9-5p mimic compared to sham. Mice treated with miR-9-5p mimics also gained weight compared to the miR-9-5p inhibitor group. To evaluate the effect of treatment with miR-9-5p mimics, multiplex pro-inflammatory cytokine analysis was performed on serum (days 4 and 7) and brain lysates (day 7). A significant decrease in pro-inflammatory cytokine concentration, including TNF-alpha and IL-6, at day 4 in serum samples was observed. A significant reduction in both IL-5 and IL-6 was observed in brain lysates.

Conclusion: Results show that administration of miR-9-5p mimics leads to a reduction of serum and brain cytokine levels after a TBI and therefore should be explored further as a potential therapeutic strategy for TBI treatment.

P049 Using Hexb-tdTomato reporter mice to track the response of microglia after spinal cord injury

Miss Ruth Colbert ,^1,2 Dr James Reynolds,^1,2 Ciara Walsh,^1,2 Irem Isik,¹ Eoghan Hyde,¹ Dr Dearbhaile Dooley^1,2

¹School of Medicine, University College Dublin, Belfield, Ireland, ²UCD Conway Institute of Biomolecular & Biomedical Research, University College Dublin, Belfield, Ireland

Introduction: Traumatic spinal cord injury (SCI) is a complex and debilitating condition, with no curative therapy. Trauma to the spinal cord initiates a primary injury phase, which is followed and exacerbated by a secondary injury phase. Largely inflammatory in nature, this secondary phase is principally coordinated by microglia; the resident immune cells of the central nervous system parenchyma. Microglia have the potential to incur harmful or neuroprotective effects on the injured spinal cord, depending on their reactive state, as dictated by their microenvironment. As potential therapeutic targets, further research into microglial responses post-injury is required prior to clinical translation. However, discriminating between resident microglia and macrophages (immune cells that infiltrate from the periphery after SCI) remains a challenge.

Methodology: We aim to construct a detailed spatiotemporal map of microglial responses after SCI, utilising Hexb-tdTomato reporter mice; a novel genetic tool whereby microglia selectively express tdTomato¹, enabling the study of these immune cells within the spinal cord post-injury, in the clinically-relevant rodent model; contusion SCI. Using fluorescence microscopy and RNA-seq, we are tracking microglial behaviour and phenotype in the lesion, at key timepoints post-injury.

Results: Preliminary data show an increase in microglial number and morphological circularity in the lesion following SCI, accompanied with significant transcriptional changes; all of which are time-dependent.

Conclusions: Further to characterising their response to traumatic injury, future work will assess their responsiveness to interleukin (IL)-13, an immunomodulatory cytokine which we have previously demonstrated improves functional and histopathological recovery after SCI. This project aims to further our understanding of how microglia respond to pro-inflammatory (SCI) and anti-inflammatory (IL-13) stimuli in an in vivo trauma environment. These data will ultimately unveil details of spinal pathology and microglial mechanisms not yet described, forming the basis for future immune-based therapeutic interventions exceeding previous approaches.

P050 Pro-inflammatory signalling along the P2X7R/NOX2 inflammatory axis in microglia

Ms Carly Douglas ,¹ Ms Shannon McGreevy,¹ Ms Janna Grimm,¹ Ms Janeen Laabei,¹ Dr Tobias Engel,² Dr David Loane¹

¹Trinity College Dublin, Dublin, Ireland, ²Royal College of Surgeons in Ireland, Dublin, Ireland

Microglia, the resident innate immune cells of the central nervous system play both protective and damaging roles in secondary injury following traumatic brain injury (TBI). The P2X7 receptor (P2X7R) is a ligand gated ion channel that is upregulated in microglia and brain infiltrating macrophages after TBI. P2X7R is activated by high concentrations of the DAMP, ATP, which is released by injured cells/tissues. ATP-dependent microglial activation promotes the release of cytotoxic pro-inflammatory mediators, interleukin-1β (IL-1β) and tumour necrosis factor-α (TNF-α). Furthermore, P2X7R activation can enhance NOX2 activity, leading to extracellular ROS production that can act downstream to promote NLRP3 inflammasome assembly and IL-1β release. The goal of this project was to investigate pro-inflammatory signalling along the P2X7R/NOX2 inflammatory axis in microglia.

Immortalized microglial (IMGs) cells or primary microglia from wildtype (WT) and P2X7R knockout murine pups were stimulated with lipopolysaccharide (LPS) in combination with benzoylbenzoyl ATP (BzATP), an ATP homologue specific to P2X7R. IMGs were treated with inhibitors JNJ-47965567 (P2X7R) or GSK2795039 (NOX2). Cellular outputs, including ROS, nitric oxide (NO), IL-1β, TNF-α, and NOX activity were measured. IMGs cells were also assessed for protein expression of P2X7R, NLRP3 inflammasome components, iNOS, GSDMD, using Western blotting.

We found that stimulating P2X7R induces pro-inflammatory microglial activation, shown by increased levels of NO, ROS, TNF-α, and IL-1β. The P2X7 inhibitor (JNJ-47965567) attenuated NOX2 activation, iNOS expression and NLRP3 inflammasome activity, resulting in significant reductions in downstream pro-inflammatory mediators (ROS, NO, TNF-α, IL-1β and GSDMD). In primary microglia, WT cells react similarly to IMGs with increased levels of IL-1β and TNF-α. However, P2X7R knockout microglia show a complete eradication of the IL-1β response.

Thus, the P2X7R/NOX2 inflammatory axis may be an important mechanism of microglial activation that can be therapeutically targeted in TBI to mitigate damaging neuroinflammation after acute neuronal injury.

P051 CSF cells in the chronic phase of traumatic brain injury share similarities with neurodegenerative conditions

MD Susanna Friberg ,^1,2 PhD Katrine Dahl Bjørnholm,³ MD, PhD Caroline Lindblad,^2,4 MD, Professor Fredrik Piehl,^1,2 Professor Henrik Zetterberg,^5,6,7,8,9,10 MD, Associate Professor Caroline Ingre,^1,2 MD, Associate Professor Eric Peter Thelin^1,2

¹Department of Clinical Neuroscience, Karolinska Institutet, Solna, Sweden, ²Medical Unit Neurology, Karolinska University Hospital, Solna, Sweden, ³Department of Neurobiology, Care Sciences, and Society, Division of Neurogeriatrics, Center for Alzheimer Research, Karolinska Institutet, Stockholm, Sweden, ⁴Medical Unit Neurosurgery, Karolinska University Hospital, Solna, Sweden, ⁵United Kingdom Dementia Research Institute at University College London, London, United Kingdom, ⁶Department of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, The Sahlgrenska Academy at the University of Gothenburg, Mölndal, Sweden, ⁷Department of Neurodegenerative Disease, University College London Queen Square Institute of Neurology, London, United Kingdom, ⁸Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden, ⁹Hong Kong Center for Neurodegenerative Diseases, Clear Water Bay, Hong Kong, China, ¹⁰Wisconsin Alzheimer’s Disease Research Center, University of Wisconsin School of Medicine and Public Health, University of Wisconsin-Madison, Madison, USA

Introduction: Traumatic brain injury (TBI) results in neurodegeneration and neuroinflammation in the acute stage, but to what degree such disease processes remain long-term is less clear. The objective of this study was to obtain single-cell RNA sequencing (scRNAseq) data of cerebrospinal fluid (CSF) cells in chronic TBI and compare with other disease entities and controls.

Methodology: Cells from CSF was captured using 10X Genomics chemistry from patients diagnosed with early onset Alzheimer’s disease (AD) (n=3), subjective cognitive decline with no clinical manifestation (n=3) and chronic (c) TBI 13–15 years after injury (n=7). Additional data was obtained from Yazdani et al1 (n=4 amyotrophic lateral sclerosis (ALS) and n=3 controls), resulting in a dataset of 45,445 cells after initial quality control.

Results: Using the standard Seurat workflow and Celltypist for annotation we identified the largest cluster as T-cells including CD4 (46.6%), CD8+ T-cells (17.9%), and mucosal-associated invariant T-cells (0.81%). Other cells of lymphoid origin were NK cells (2.45%) and a small subset of B-cells (0.35%). The second most prevalent cell type was myeloid cells including intermediate monocytes (15.3%), classical monocytes (3.35%), dendritic cells (3.33%), macrophages (1.08%), and plasmacytoid DCs (0.69%). Preliminary data does not show any major shift in cell populations between AD, cTBI, and ALS. Results from differentially expressed gene analysis is pending.

Conclusions: cTBI patients present with a majority T-cells in CSF. Further comparisons between cTBI and AD, ALS, and controls will be included in the final presentation.

P052 Acute and chronic responses of microglia following traumatic brain injury

Dr Shannon Gilchrist ,¹ Dr Kamar E. Ameen-Ali,² Professor Douglas S. Smith,³ Professor William Stewart^1,4

¹School of Psychology and Neuroscience, University Of Glasgow, Glasgow, United Kingdom, ²School of Health and Life Sciences, Teesside University, Teesside, United Kingdom, ³Department of Neurosurgery, Penn Centre for Brain Injury and Repair, Perelman School of Medicine, University of Pennsylvania, Philadelphia, United States of America, ⁴Department of Neuropathology, Laboratory Medicine Building, Queen Elizabeth University Hospital, Glasgow, United Kingdom

Introduction: Traumatic brain injury (TBI) is a recognised risk factor for neurodegenerative disease. Nevertheless, the biological process driving an acute biomechanical injury towards lifelong neurodegenerative pathology remains unknown. Pre-clinical research has identified neuroinflammation as a significant driver of TBI-related neurodegeneration (TReND). However, the contribution of neuroinflammation to pathologies following survival from TBI in human patients remains uncertain. Here, we aim to study the temporal and anatomical profiles of the microglial response to TBI, leveraging human post-mortem samples from the Glasgow TBI archive.

Methods: Cases were selected from individuals with history of acute (n=27; <2 weeks survival) or chronic TBI (n=32; >1-year survival), or as age-matched, non-injured controls (n=19). Sections from multiple brain regions (corpus callosum, sub-cortical white matter, grey matter of cingulate gyrus, grey matter of cingulate sulcus, internal capsule and grey matter of the thalamus) were then stained with a panel of microglial markers (Iba1, CR3/43, CD14, CD163) chosen for their associated microglial phenotypes. Sections were assessed for the extent and distribution of pathology.

Results: In contrast to age-matched controls, activated microglia were the predominant phenotype observed in the corpus callosum, cingulate sulcus and thalamus of acute TBI cases. In addition to this, levels of CR3/43, a marker of activated microglia, were increased in the corpus callosum and cingulate white matter of chronic TBI cases. Furthermore, CD163+ microglia found in brain parenchyma displayed distinct differences between grey and white matter regions.

Conclusions: These results support a dynamic microglial response to TBI, largely localised to white matter and sparing grey matter cortical and sub-cortical regions. Further work is required to document the complex interaction between post-TBI neuroinflammation and TReND. This must include characterisation of specific microglial subtypes involved, particularly focusing on perivascular microglia which are susceptible to damage following blood-brain barrier disruption.

P053 The Potential of Gene Delivery for the Treatment of Traumatic Brain Injury

James Dooley,¹ Jasmine Hughes ,¹ Edward Needham,¹ Katerina Palios,¹ Adrian Liston¹

¹University of Cambridge, Cambridge, United Kingdom

Therapeutics for traumatic brain injuries constitute a global unmet medical need. Despite the advances in neurocritical care, which have dramatically improved the survival rate for the ∼70 million patients annually, few treatments have been developed to counter the long-term neuroinflammatory processes and accompanying cognitive impairments, frequent among patients. This review looks at gene delivery as a potential therapeutic development avenue for traumatic brain injury. We discuss the capacity of gene delivery to function in traumatic brain injury, by producing beneficial biologics within the brain. Gene delivery modalities, promising vectors and key delivery routes are discussed, along with biological cargos with the potential to improve long-term outcomes for patients. Finally, we review the health economics of traumatic brain injury, and whether future gene delivery approaches will be economically viable for health care systems.

P054 Novel immune-matrix interactions after traumatic brain injury

Miss Katy Palios ,¹ Dr Smaranda Badea,¹ Dr Raquel Oliveira,¹ Miss Hannah McConchie,¹ Mr Aminul Ahmed,¹ Dr Claire Troakes,¹ Professor Elizabeth Bradbury¹

¹King's College London, London, United Kingdom

Introduction: Chronic inflammation appears a shared pathological mechanism between traumatic brain injury (TBI), ageing, and dementia. Novel evidence suggests specific components of the extracellular matrix (ECM) play a major role in modulating the immune response after spinal cord injury. Here, we hypothesised that ECM-immune interactions also contribute to chronic inflammatory pathology in TBI.

Methods: First, we sought to characterise spatiotemporal changes in ECM components with putative immunomodulatory functions in both human and rat post-mortem TBI tissue via immunohistochemistry. Human cases included control (non-TBI), acute (< 30 mins – 5 days survival), and subacute/chronic (19 days – 1 year survival) groups. We also employed a preclinical rat model of controlled cortical impact (CCI) and collected tissue at 7- and 28-days post-injury. Next, we investigated the effects of therapeutic modification of the ECM in TBI. At the time of CCI injury, rats received intracortical injections of lentiviral-ChABC (ChondroitinaseABC: an enzyme which remodels the ECM), or control lentiviral-GFP. Tissue was collected at 28 days post-injury and processed for immunohistochemistry or qPCR.

Results: In cross-translational studies of human and rat TBI post-mortem tissue, we identified ECM protein accumulation over acute-chronic injury phases, with a predominant glial-like expression. Therapeutic modulation of the ECM generated a dramatic attenuation of pro-inflammatory gene expression in the perilesional cortex of rats. This was apparent by a significant reduction in mRNA expression of pro-inflammatory cytokines and complement factors, including “synapse-tagging” opsonins C1q and C3. Interestingly, there was no change in the total number of glial cells, suggesting a beneficial change in glial phenotypes after ECM-modulation.

Conclusions: These data build on novel research indicating the ECM has pleiotropic functions after neurotrauma, including a potent immunomodulatory role. The accumulation of ECM components in chronic TBI suggests this could be a shared pathological mechanism between TBI and neurodegenerative disease.

(King’s College London&UKRI/MRC funding)

P055 Inflammatory biomarkers Differentiate the Stage of Maturation in Chronic Subdural Hematomas

Dr Teodor Svedung Wettervik ¹

¹Uppsala University, Uppsala, Sweden

Objective: Inflammation is a major pathophysiological driver of the development of chronic subdural hematomas (CSDH), but there is limited knowledge on the key molecular processes and corresponding biomarkers involved. The aim of this study was to investigate a subset of inflammatory biomarkers and their relation to the clinical status of the patient and the radiological characteristics of the CSDH.

Methods: In this observational study, 58 patients who were operated on with CSDH evacuation, Uppsala, Sweden, between 2019 and 2021, were prospectively included. The CSDH fluid was collected peri-operatively and was later analyzed with PEA technique (Olink) for a panel of 92 inflammatory biomarkers. Demographic, neurological (Markwalder), radiological (general (Nakaguchi classification) and focal (septa below the burr holes)), and outcome variables were collected.

Results: In 84 of the 92 inflammatory biomarkers, the concentration was above the detection limit in more than 50% of the patients. There was a significant difference in GDNF, NT-3, and IL-8 depending on the Nakaguchi class, with higher values in the trabeculated CSDH subtype. In addition, those with septa at the focal area of CSDH collection, had higher levels of GDNF, MCP-3, NT-3, CXCL1, CXCL5, IL8, and OSM. There was no association between Markwalder grade and the inflammatory biomarkers.

Conclusions: Our findings support the presence of local inflammation in the CSDH, a shift in biomarker pattern as the CSDH matures towards the trabeculated state, potentially differences in biomarker patterns within the CSDH depending on the focal environment with presence of septa, and that the brain might develop protective mechanisms (GDNF and NT-3) in case of mature and long-standing CSDHs.

P056 Interleukin-13 delivery via PLGA microparticle-embedded GelMA hydrogel significantly improves functional and histopathological recovery in a mouse contusion SCI model

Ciara Walsh ,^1,2 Ruth Colbert,^1,2 Dr James P Reynolds,^1,2 Emily Dunne,¹ Emmanuelle D Aiyegbusi,^1,2 Dr Ross O’Carroll,^1,2 Dr Jacek K Wychowaniec,^3,4 Dr Dermot F Brougham,³ Dr Dearbhaile Dooley^1,2

¹School of Medicine, University College Dublin, Ireland, ²UCD Conway Institute of Biomolecular & Biomedical Research, University College Dublin, Ireland, ³School of Chemistry, University College Dublin, Ireland, ⁴AO Research Institute, Switzerland

Introduction: Spinal cord injury (SCI) is a devastating condition with limited regeneration, and no curative therapy is currently available. We have previously demonstrated that cell-based delivery of the immunomodulatory cytokine interleukin(IL)-13 drives alternative immune cell activation and improves functional and histopathological recovery after SCI in mice. However, cell-based delivery approaches carry several translational limitations and thus a more clinically relevant delivery platform is necessary. To address this challenge, we have developed an injectable hydrogel-based system for localized and sustained delivery of IL-13 in preclinical SCI.

Methods: IL-13-encapsulated poly(lactic-co-glycolic acid) (PLGA) microparticles were synthesized using the double emulsion method and embedded in a gelatin methacrylate (GelMA) hydrogel to form our injectable IL-13 delivery system (HGIL13). Cumulative IL-13 release was measured in vitro and IL-13 bioactivity was assessed using BV2 microglia cells. The therapeutic efficacy of HGIL13 was then determined in vivo using a mouse contusion SCI model. Functional recovery was measured using the gold standard Basso Mouse Scale for locomotion, while the effect of HGIL13 on lesion size, demyelinated area and glial cell reactivity was measured by histopathological staining and analysis.

Results: HGIL13 released IL-13 for up to 6 weeks in vitro with a final cumulative dose of 185.7 ± 17.9 pg IL-13/µL. Released IL-13 retained its bioactivity in vitro as demonstrated by a significant reduction in LPS-induced TNF-α and iNOS expression, and increased Arg-1 expression in BV2 microglia. In a mouse contusion SCI model, BMS scores of HGIL13-treated animals were significantly higher than vehicle controls over 28 days post-injury. Furthermore, we found that HGIL13 significantly improved various measures of histopathological recovery including lesion size, demyelinated area and astrogliosis in vivo.

Conclusions: Our HGIL13 delivery system improves both functional and histopathological recovery in a mouse contusion SCI model, indicating significant translational potential as an immunomodulatory therapy for SCI repair.

P058 Glial Responses after TBI in Aging htau Mice: An mRNA Analysis with Nanostring Glia Platform

MD., PhD Jiepei Zhu ,¹ Dr Firas Kobeissy,¹ Dr. Mojtaba Golpich,¹ Zo-Yu Wu,¹ Guangzheng Cai,¹ Dr. Kevin Wang¹

¹Morehouse School of Medicine, Atlanta, United States

Introduction: Traumatic brain injury (TBI) is a known risk factor for neurodegenerative diseases, such as Alzheimer's Disease-Related Dementias (ADRD) and Chronic Traumatic Encephalopathy (CTE). The specific mechanisms by which TBI may precipitate these conditions remain unclear, but recent studies suggest neuroinflammation and changes in glial cells (astrocytes and microglia) play a role in the development of tau-related pathologies post-TBI.

Methods: An hTau mouse model (C57BL6/HuMAPT Mut) aged 18 months and subjected them to controlled cortical impact (CCI, a preclinical TBI model) to analyze glial responses associated with TBI-induced neurodegeneration in the context the potentially pathogenic tau protein. Using the NanoString glial profiling panel, we measured 770 glial-related transcripts in 21 hTau mice divided into a sham group (n=4) and TBI groups with survivals of 24 hours (n=6), 3 days (n=5), and 7 days (n=6) post-CCI, analyzing cortex and hippocampus tissues.

Results: Gene expression differences among the groups were identified using Student’s t-test (p-value < 0.05) and presented in a volcano plot. Data from the cortex and hippocampus showed that both astrocyte A2 (neuroprotective phenotype) and A1 (neurotoxic phenotype) genes were increased in the injured site and the contralateral site with an early rise of the A2 at early time points (D1) followed by A1 (D3) which matches the microglial M2 polarization elevation at similar time points.

Conclusion: Our targeted NeuroProteomoics Platform identified several novel specific gene targets that are related to astrocytic (A1) activation including (C3, GFAP, TSPO), M1 microglial phenotype (Fcgr1, Fcgr4 and CD68) that were showing temporal changes across time. Systems Biology and Gene-Related Neuroinflammatory and Neuropathological pathways are being analyzed to understand the contribution of tau protein and altered glial phenotypes. Taken together, Our data point to the utility of multi-omic approaches to understand glial dysregulation in the context of TBI leading to tauopathic neurodegenerative conditions.

P059 Effect of low fibrinogen level on in-hospital mortality and 6-month functional outcome of TBI patients, a single center experience

Dr Sina Zoghi ,¹ Dr. Omid Yousefi,² Dr, Amirmohammad Farrokhi,² Dr. Reza Taheri,² Dr. Hosseinali Khalili²

¹Student Research Committee, Shiraz University of Medical Sciences, Shiraz, Iran, ²Department of Neurosurgery, Shiraz University of Medical Sciences, Shiraz, Iran

Background: In patients affected by traumatic brain injury (TBI), hypofibrinogenemia within the initial hours of trauma can be expected due to vascular and inflammatory changes.

Methods: This study included all TBI patients admitted to our center who had no prior history of coagulopathy or any systemic disease, were alive on arrival, and had not received any blood product before admission. On admission, hospitalization, imaging, and 6-month follow-up information of included patients were extracted from the TBI registry database. The baseline characteristics of patients with fibrinogen levels of less than 150 mg/dL were compared with the cases with higher levels.

Results: A total of 3049 patients (84.3% male, 15.7% female), with a mean age of 39.25 ± 18.87, met the eligibility criteria of this study. 494 patients had fibrinogen levels < 150 mg/dl, who were mostly younger and had lower average GCS scores in comparison to cases with higher fibrinogen levels. By comparison of the patients who died during hospitalization and survivors, it was shown that fibrinogen < 150 mg/dl is among the prognostic factors for in-hospital mortality (OR:1.75, CI: 1.32:2.34, P-value < 0.001), while the comparison between patients with the favorable and unfavorable functional outcome at 6-month follow-up, was not in favor of prognostic effect of low fibrinogen level (OR: 0.80, CI: 0.58: 1.11, P-value: 0.19).

Conclusion: Hypofibrinogenemia is associated with in-hospital mortality of TBI patients, along with known factors such as higher age and lower initial GCS score. However, it is not among the prognostic factors of midterm functional outcome.

P060 Development of a Novel Neurological Score Combining GCS and FOUR Scales for Assessment of Neurosurgical Patients with Traumatic Brain Injury: GCS-FOUR Scale

Dr Sina Zoghi ,¹ Dr. Ali Ansari,¹ Dr. Amirabbas Khoshbooei,² Dr. Mohammad Amin Mosayebi,³ Dr. Maryam Feili,¹ Dr. Omid Yousefi,³ Dr. Reza Taheri,³ Dr. Hosseinali Khalil³

¹Student Research Committee, Shiraz University Of Medical Sciences, Shiraz, Iran, ²School of Medicine, Fasa University of Medical Sciences, Fasa, Iran, ³Department of Neurosurgery, Shiraz University Of Medical Sciences, Shiraz, Iran

Background: Consciousness assessment is crucial in patients with traumatic brain injury (TBI). The best coma scale is the one that is applicable for almost all medical staffs, adoptable for all patients, and predictor for patients’ outcome. we developed a novel scoring system, GCS-FOUR, regarding the mentioned features by combining GCS and FOUR score.

Methods: We retrospectively used the data of TBI patients admitted to our neurosurgical ICU in a period of 2 years. Eye and motor component of GCS score and brainstem reflex component of FOUR score was utilized for obtaining GCS-FOUR. Then, the association between the patients’ GCS score, FOUR score, and GCS-FOUR and their outcome in 6 months was assessed.

Results: A total of 140 patients were included in our study. Their mean age was 30.6 and 89.3% them were male. In the logistic regression, all three scales were able to predict ICU length of stay, mortality, persistent vegetative states, and good recovery. GCS was superior to others in predicting mortality while GCS-FOUR best predicted good recovery.

Conclusion: We found our new coma scale, GCS-FOUR, to be predictor for outcome of patients with traumatic brain injury. This coma scale is adoptable for almost all patients and is also easily applicable for every medical staff.

P061 Blood-borne protein biomarkers of human epileptogenesis after severe traumatic brain injury

Professor Denes Agoston ,¹ Mr. Jesse McCullough,¹ Ms. Roxanne Aniceto,¹ Ms. I-Hsuan Lin,¹ Mr. Michael Eklund,¹ Ms. Christina Kwak, Professor Paul Vespa²

¹USUHS, Bethesda, United States, ²UCLA, Los Angeles, United States

Traumatic brain injury (TBI) is one of the most significant causes of acquired epilepsy. Post traumatic epilepsy (PTE) develops after a latent period post injury thus early anti-epileptogenic therapy can prevent PTE [1]. The epileptogenic process is currently poorly understood and there is no efficient clinically validated anti-epileptogenic (AEG) treatment. The Epilepsy Bioinformatics Study for Antiepileptogenic Therapy (EpiBioS4Rx), a large, international, multicenter Center without Walls (CWOW) study was designed to address this pressing need [2].

The clinical arm, Project 3 of (EpiBioS4Rx) has enrolled 255 subjects who have been highly characterized using early MRI, EEG, and clinical biomarkers and patients were assigned to various groups defined by long-term clinical outcomes. Blood samples were collected at various post-injury time points at a high temporal resolution and plasma samples have been analyzed for protein biomarkers indicative of neuron, axon and astroglia damage/loss; general tissue/cell damage; inflammatory process(es); hippocampal damage/atrophy and epileptic activity using Reverse Phase Protein Microarray (RPPM) screening followed by sandwich ELISA verification.

We then have utilized a supervised decision tree machine learning algorithm to determine a combination of time point and biomarker rules that best separates out the various patient groups defined by clinical outcomes. The training data set was used to train the model, which we then tested on the test data set using R programming language. We have generated decision trees PER post-injury timepoints that have shown a) time dependent evolution of pathobiological responses to TBI involving tissue and hippocampal damage and inflammatory processes.

Combined with clinical outcome measures, especially EEG, the biomarker data will enable to develop an algorithm to be used as a research specific model that is human readable, which could elucidate a mechanistic/systems biology understanding of the pathobiological differences between the clinical groups and guide therapeutic interventions.

P062 The relation between NfL and GFAP in CSF after a severe traumatic brain injury (sTBI) and one-year and 10–15 years outcome

Emma Andersson ,¹ MD, PhD Martin Öst,¹ MD, PhD Keti Dalla,¹ MD, PhD Henrik Zetterberg,^2,3,4,5,6 MD, PhD Kaj Blennow,^2,3 MD, PhD Bengt Nellgård¹

¹University of Gothenburg, Mölndal, Sweden, ²Department of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, the Sahlgrenska Academy at the University of Gothenburg, Mölndal, Sweden, ³Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden, ⁴Department of Neurodegenerative Disease, UCL Institute of Neurology, Queens Square, London, United Kingdom, ⁵UK Dementia Research Institute at UCL, London, United Kingdom, ⁶Hong Kong Center for Neurodegenerative Diseases, China

Introduction: Neurofilament light (NfL) is a marker of axonal damage and Glial fibrillary acidic protein (GFAP) a marker of reactive astrocytes. This study investigates the concentration dynamics of NfL and GFAP in Cerebrospinal fluid (CSF), in near proximity to the injury, during the acute phase after a sTBI. Furthermore, the relation between concentrations of NfL and GFAP in CSF and long-term outcome are investigated.

Method: Patients from the neurointensive care unit, Sahlgrenska university hospital, Gothenburg, Sweden, between 2000–2005 were included if having: a Glasgow coma scale of <9, need for invasive mechanical ventilation, need for a ventricular catheter, and residing in Sweden. One-year Glasgow outcome scale (GOS) was completed the following year (2001–2006) and GOS 10–15 year after trauma in 2015. NfL and GFAP in CSF, was analysed with ELISA in 2006.

Results: Forty-seven patients were included with a mean age of 45 and the majority were male (83%). The concentration dynamics over time are described of Nfl and GFAP, collected intermittently in CSF on day 0–2, 3–4, 6–8 and 10–18. The relation between the concentration of NfL and GFAP, in CSF, and the outcome after a sTBI one-year and 10–15 years after trauma is also explored, considering age and a pre-trauma history of neurological disease as confounders.

Conclusion: Our results indicates that NfL and GFAP are relevant biomarkers, associated with both short-term and long-term outcome after a severe traumatic brain injury.

P063 The relation between total-tau in serum after a severe traumatic brain injury (sTBI) and outcome one-year and 10–15 years after trauma

Emma Andersson ,¹ MD, PhD Keti Dalla,¹ MD, PhD Fredrik Olsen,¹ MD, PhD Henrik Zetterberg,^2,3,4,5,6 MD, PhD Kaj Blennow,^2,3 MD, PhD Bengt Nellgård¹

¹Department of Anesthesiology and Intensive Care Medicine, Institution of Clinical Sciences, Gothenburg University, Gothenburg, Sweden, ²Department of Psychiatry and Neurochemistry, Institute of Neuroscience and Physiology, the Sahlgrenska Academy at the University of Gothenburg, Mölndal, Sweden, ³Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, Mölndal, Sweden, ⁴Department of Neurodegenerative Disease, UCL Institute of Neurology, Queen Square, London, United Kingdom, ⁵UK Dementia Research Institute at UCL, London, United Kingdom, ⁶Hong Kong Center for Neurodegenerative Diseases, China

Introduction: Normally found in the microtubules of axons, Tau is a biomarker indicating axonal injury. It is also a pathologic hallmark for Alzheimer disease and chronic traumatic encephalopathy. Here we explore the concentration trajectory of serum total-tau the first two weeks after a sTBI and relate these levels to long-term outcome.

Method: During 2000–2005 patients were included from the neurointesive care at Sahlgrenska university hospital, Gothenburg, Sweden, if they had: a Glasgow coma scale <9, need for invasive mechanical ventilator, need for a ventricular catheter and residing in Sweden. Glasgow outcome scale (GOS) was assessed one-year (2001–2006) and 10–15 years (2015) after trauma. Patients were included from the above-described population based on one-year GOS. Afterwards, serum total-tau was analysed with SIMOA.

Results: Thirty patients were included. The mean age was 34 and the majority were male n=22 (73%). Here we describe the temporal profile of total-tau in serum collected on day 0–2, 3–4, 6–8 and 10–12 after a sTBI and its relation to outcome one-year and 10–15 years after trauma. Furthermore, in regard to outcome one-year and 10–15 years after trauma, the temporal profile of total-tau is studied to identify optimal time windows for analyses.

Conclusion: The preliminary results indicate that total-tau in serum has potential as a biomarker in regard to long-term outcome and the optimal time to analyse total-tau in serum might not be when the temporal profile peaks.

P064 Serum biomarker concentrations upon admission in acute traumatic brain injury: associations with TBI severity, Toxoplasma gondii infection, and outcome in a referral hospital setting in Cameroon

Dr FRANKLIN CHU BUH ,¹ Prof Germain Sotoing Taiwe,¹ Dr Firas H. Kobeissy,¹ Prof Kevin W. Wang,¹ Prof Andrew I. R. Mas,¹ Prof Mathieu Motah,¹ Mr Basil Kum Meh,¹ Mr Eric Youm,¹ Prof Peter J. A. Hutchinson,¹ Prof Irene Ule Ngole Sumbele¹

¹University of Buea, Buea, Cameroon

Background: traumatic brain injury (TBI) is extremely common and associated with complex biological changes. Therefore, understanding the underlying pathomechanisms of TBI and achieving an accurate diagnosis is necessary for effective and patient-oriented treatment. Despite the available literature on TBI biomarkers elsewhere, data are limited or non-existent in sub-Saharan Africa (SSA).

Aim: The study aimed to analyze associations in acute TBI between the admission serum biomarker concentrations and TBI severity, CT-scan findings, and outcome, as well as to explore the influence of concurrent Toxoplasma gondii infection on TBI biomarker concentrations.

Methods: The concentrations of serum biomarkers (GFAP, NFL Tau, UCH-L1, and S100B) were measured and Toxoplasma gondii were detected in the samples obtained <24 h post-injury. GFAP, UCH-L1, t-tau, and NFL was analysed at the University of Florida using single molecule arrays (SiMoA). T. gondii and S100B were measured using ELISA. GOSE was used to evaluate the 6-month outcome.

Results: All of the biomarker levels increased with the severity of TBI, but this increase was significant only for NFL (p = 0.01). The GFAP values significantly increased (p = 0.026) in those with an unfavourable outcome. The Tau levels were higher in those who died (p = 0.017). GFAP and NFL were sensitive to CT-scan intracranial abnormality (p values of 0.004 and 0.002, respectively). The S100B levels were higher (p < 0.001) in TBI patients seropositive to Toxoplasma gondii.

Conclusion: NFL was found to be sensitive to TBI severity, while NFL and GFAP were predictive of CT intracranial abnormalities. Increased levels of GFAP and Tau were associated with poorer outcomes 6 months after TBI, and the S100B levels were significantly affected by concurrent T. gondii infection in TBI patients compared with the seronegative patients.

P065 The Brain-Immune-Genome (BIG) Clinical Profile of Human Neuroglobin Following Stroke – A systematic Review and Meta-Analysis

Ph.D., EMBA, MSN Pei-Ying Chuang ,¹ PhD, MSN Shih-Ni Chen,² MS Yunting Fu¹

¹University of Maryland, Baltimore, United States, ²Hsin-Chu Long-Term Department in Taiwan, Hsin-Chu, Taiwan

Background/Introduction: Ischemic or hemorrhagic strokes can damage neurons due to insufficient oxygen and blood flow in the brain. Neuroglobin (NGB) is a novel neuroprotector (chromosome 14q24.3, 151 amino acid, 17 kDa) found in the central and peripheral nervous systems. NGB has two oxygen-binding sites [His64 and His96] that prevent oxidative stress under cerebral hypoxia/ischemia. The study AIMS to establish a Brain-Immune-Genome (BIG) clinical profile between NGB and stroke and provide early predictions of individuals’ neurological health outcomes.

Methods: Six public databases (PubMed, EMBASE, Medline, Cochrane library, CINAHL via EBSCO, and Scopus) were utilized. The MeSH search terms used brain ischemia (hypoxia), infarction, stroke, transient, emboli, or thrombosis, and neuroglobin from the years 2000 to 2023. The inclusion was human, stroke, neuroglobin, and experimental research design, while exclusion included no full-text articles, non-English, and animal studies. The PRISMA flow diagram and Bayesian network meta-analysis (SPSS 29.0) was used to examine and compare various BIG clinical profiles.

Results: Twenty-eight studies out of 323 published were included. The biosample, which comprised peripheral blood, cerebrospinal fluid, and brain tissues/neurons. The results showed that NGB protein and mRNA expression increased after acute stroke. The 5'-flanking region of human NGB identified a transcription start site located at -306bp relative to the translation start site ATG. The various markers were identified: neuronal markers (NSE, NGB, H-FABP, and NSE), inflammatory markers (IL6, TNF, CRP, VEGF, AIF, and IGF), blood-brain-barrier marker (MMP-9), hemostatic markers (D-dimer and PAI-1), and astroglia (GFAP and S100B), including apoptotic signaling pathway in mitochondria. Finally, the neurological outcomes were assessed by the Glasgow Coma Scale, Glasgow Outcomes Scale, the National Institute of Health Stroke Scale, Hunt and Hess scale, the World Federation of Neurological Surgeons scale, and modified Fisher Scale, and the Modified Ranking Scale.

Future Direciton: NGB shows promise as a neuroprotective candidate.

P066 IGF-1 deficiency is not involved in autoregulatory dysfunction after severe traumatic brain injury

Andras Czigler ,^1,2 Dominika Lendvai-Emmert,^1,2 Emőke Hegedüs,^1,3,4 Zsofia Dina Magyar-Sumegi,^1,4,5 Levente Stankovics,¹ Ivan Domink,¹ Virag Magvasi, Andras Buki,^1,6 Peter Toth^1,2,7,8

¹Department of Neurosurgery, Medical School, University of Pecs, Pecs, Hungary, ²Institute for Translational Medicine, Medical School, University of Pecs, Pecs, Hungary, ³Department of Anaesthesiology and Intensive Therapy, Medical School, University of Pecs, Pecs, Hungary, ⁴Doctoral School of Clinical Neurosciences, Medical School, University of Pecs, Pecs, Hungary, ⁵Department of Psychiatry and Psychotherapy, Medical School, University of Pecs, Pecs, Hungary, ⁶Department of Neurosurgery, Faculty of Medicine and Health, Orebro University, Orebro, Sweden, ⁷Department of Public Health, Semmelweis University, Budapest, Hungary, ⁸Department of Neurosurgery, Oklahoma Center for Geroscience and Healthy Brain Aging, University of Oklahoma Health Sciences Center, Oklahoma City, USA

Introduction: Severe traumatic brain injury (TBI) leads to impairment of autoregulation of cerebral blood flow (CBF), determining the outcome of patents. TBI frequently leads to neuroendocrine changes, most commonly disturbances in the growth hormone/insulin-like growth factor 1 (GH/IGF-1) axis. IGF-1 deficiency was shown to impair cerebrovascular function both in laboratory animals and humans. We tested the hypothesis that TBI-induced IGF-1 deficiency is associated with autoregulatory dysfunction and aimed to determine outcome of patients.

Methods: We studied 25 patients after severe TBI in a prospective manner admitted to the Department of Neurosurgery of the University of Pecs. We assessed autoregulatory function by obtaining pressure reactivity index (PRX), as well as calculating the transcranial Doppler-based mean flow index (Mx). IGF-1 and IGFBP-3 levels from blood samples collected within the first day post-TBI were measured with ELISA. We evaluated TBI outcomes at 6 months post-injury using the modified Rankin Scale (mRS).

Results: IGF-1 decreased in patients with severe TBI compared to controls, but there was no significant correlation between IGF-1/IGFBP-3 levels and autoregulatory parameters Mx or PRx. However, we observed a negative correlation between both IGF-1 and IGFBP-3 levels and clinical outcomes assessed by the mRS (R=-0.55 and R=-0.68 respectively).

Conclusion: IGF-1/IGFBP-3 deficiency is not associated with autoregulatory dysfunction after severe TBI, however, they predict better outcome. Further studies should establish the role of IGF-1 deficiency in TBI-induced secondary injury in order to improve patients’ outcome.

P067 Dynamic Trends in Clinical Blood Labs in the First Two Weeks after Severe TBI Improves Six-Month Outcome Prediction compared to the IMPACT Model

Dr Shawn Eagle ,¹ Maxwell Wang,¹ Regan Shanahan,¹ Anna Slingerland,¹ Shovan Bhatia,¹ Michael Kann,¹ Tyler Augi,¹ Ava Puccio,¹ David Okonkwo¹

¹University of Pittsburgh, Pittsburgh, United States of America

Introduction: The International Mission for Prognosis and Analysis of Clinical Trials (IMPACT) is the most studied prognostic model for severe TBI and uses data from hospital admission to predict 6-month neurological outcome. We evaluated dynamic trends of clinical labs within the first two weeks of hospitalization to improve upon this model.

Methodology: Severe TBI patients (GCS=3-8) presenting to a level 1 trauma center were enrolled under an IRB-approved protocol to collect injury and acute care data as well as neurological outcome(n=240). IMPACT was the base model with daily rates of sodium, hemoglobin, platelets and glucose. Random forest predictor models were developed to identify the variance accounted for (R2) between predicted and 6-month neurological outcome measured by Glasgow Outcome Scale Extended (GOSE). Spearman correlations were conducted for clinical labs to six-month GOSE. Cortisol labs were collected in a subset of patients (n=100) within 48 hours of admission and correlated to daily glucose, sodium and platelet levels.

Results: The IMPACT model had an R2 of 22%, while addition of serial clinical labs increased R2 to 31%. Glucose and sodium trends over the first two weeks significantly correlated with 6-month outcome (r=-0.41 to -0.14). Higher platelet count on Days 4–9 correlated with better outcomes (r=0.15–0.29; p=0.006). Cortisol correlated with glucose on days 1–6 (r=0.27–0.34; p=0.016) and platelets on days 6–11 (r=-0.40 to -0.23; p=0.01).

Conclusions: The magnitude of sodium and glucose increase was negatively linked to patient outcomes. The magnitude of platelet increase was linked to positive recovery. Acute cortisol was associated with these glucose and platelet responses, implicating stress and endocrine physiological mechanisms.

P068 Deep Learning vs. Logistic Regression in TBI Outcome Prediction Using Ordinal and Multiclass Glasgow Outcome Scale Extended

Lisa Yvette Inyange ,^1,2 Eugenia Mawuenya Akpo,^1,2 Harri Merisaari,^3,4 Jussi Posti,^2,3 Olli Tenovuo,^2,3 Peter Ngum^2,5,7,8

¹Carnegie Mellon University - Africa, Department of Information & Communication Technology, Kigali, Rwanda, ²Pan African AI Health Initiative, ³Turku Brain and Mind Center, University of Turku, Turku, Finland, ⁴Department of Diagnostic Radiology, University of Turku, Turku, Finland, ⁵Turku Brain Injury Center, Turku University Hospital, Finland, ⁶Neurocenter, Department of Neurosurgery, Turku University Hospital, Finland, ⁷Department of Clinical Neurosciences, University of Turku, Finland, ⁸Johns Hopkins University, Carey Business School

Introduction: Early TBI outcome prognosis is crucial for patient management, especially in limited-resource settings. We evaluate TimesNet, a state-of-the-art deep learning model, against logistic regression in predicting 6-month Glasgow Outcome Scale Extended (GOSE) scores across various GOSE class stratifications utilizing admission data and blood-based biomarkers (BBM).

Methodology: Analyzing the EU-funded TBIcare dataset of 200 TBI cases and 40 orthopedic controls, we focused on demographics, worst motor response scores for non-intubated patients, and BBM levels at admission. We applied comprehensive data preprocessing—imputation, and outlier management. To enhance the reliability and validity of our results against overfitting and to ensure generalizability, we implemented a 5-fold cross-validation. The performance of TimesNet and logistic regression was evaluated across four experimental frameworks: (1) Ordinal GOSE prediction, (2) Ordinal with orthopedic control inclusion, (3) Multiclass GOSE, and (4) Multiclass with an extra control class, utilizing ordinal c-index and F1 scores.

Results: Our cohort was 69.61% male with a median age of 51 years (IQR: 34). GCS scores at admission were 13–15 for 70.72%, 3–8 for 19.34%, and 9–12 for 8.84%. TimesNet consistently outperformed logistic regression across scenarios. In traditional class assignments (8 and 3 classes), it achieved F1/ or c index scores of 0.42 and 0.72, surpassing logistic regression's 0.36 and 0.70. Its superiority was more pronounced in new class configurations (9 and 4 classes), with F1 scores of 0.65 and 0.85 against logistic regression’s 0.43 and 0.69, especially notable in the 4-class setup.

Conclusion: TimesNet demonstrates promising predictive performance for TBI outcomes, offering better recovery spectrum differentiation compared to traditional models. Its performance indicates a step forward in predicting TBI recovery, yet the full scope of its clinical utility and applicability in diverse settings requires additional study. This study underscores the importance of advanced modeling in improving TBI patient care.

P069 Serum biomarkers in the prediction of outcomes following traumatic brain injury

Dr Chirag Jain ,¹ Dr Bhagavatula Devi,² Dr Dhaval Shukla,¹ Dr Dhananjaya Bhat³

¹National Institute of Mental Health and Neurosciences, Bangalore, India, ²NIHR, UK, ³RV Aster, Bangalore, India

Introduction: Traumatic brain injury affects millions of individuals worldwide and continues to pose challenges for evaluation and management. Serum biomarkers have recently shown promise in prediction of traumatic brain injury as well as its outcomes.

Methodology: We conducted a prospective, single-center study and recruited patients with traumatic brain injury with onset <24 hours aged 18–60 years. Clinical details and routine blood investigations were recorded at time of admission. Biomarkers were analyzed on Day 1, Day 3 and Day 10 post-trauma by collecting 10 ml of venous blood. The following biomarkers were analyzed – IL-1beta, IL-6, IL-10, TNF-alpha, TGB-beta1, S100-B, GFAP, NSE, MBP, UCH-L1.

Results: 347 patients were recruited in the study. 74 control samples were collected from healthy volunteers. Out of 347 TBI patients, 166 had mild TBI, 117 had moderate TBI and 64 had severe TBI. There was no significant difference in the concentration of all biomarkers except IL-6 in patients with different severities of TBI. Serum IL-6 was significantly lower in patients on Day 10 compared to Day 1. Higher serum levels of Il-1beta, IL-6, GFAP, NSE and UCH-L1 were associated with better outcomes (moderate disability and good recovery). Higher serum levels of IL-10 and MBP were noted in patients who died as compared to those who survived.

Conclusion: A panel of biomarkers was tested in patients with traumatic brain injury and serum levels of IL—1beta, IL-6, IL-10, GFAP, NSE, MBP and UCH-Li1 were found to correlate with outcome at follow up. No association was found between serum levels of biomarkers and severity of traumatic brain injury.

P070 Biomarkers in identifying patients benefiting from acute subdural hematoma evacuation

Otto Korhonen ,¹ Rick Vreeburg,² Dr. Mehrbod Mohammadian,^1,12 Dr. Iftakher Hossain,^1,10 Dr. Teemu Luoto,⁵ Prof. Olli Tenovuo,⁶ Prof. Andrew Maas,⁷ Prof. David Menon,⁸ Prof. Andras Büki,⁹ Dr. John Yue,¹¹ Dr. Thomas van Essen,^2,3,4 Dr. Jussi Posti¹

¹Neurocenter, Department of Neurosurgery and Turku Brain Injury Center, Turku University Hospital and University of Turku, Turku, Finland, ²University Neurosurgical Centre Holland, Leiden University Medical Centre, Haaglanden Medical Centre, and Haga Teaching Hospital, Leiden-The Hague, The Netherlands, ³Dpt of Neurosurgery, Leiden University Medical Centre, The Netherlands, ⁴Department of Surgery, Division of Neurosurgery, QEII Health Sciences Center and Dalhousie University, Halifax, Canada, ⁵Department of Neurosurgery, Tampere University Hospital and Tampere University, Tampere, Finland, ⁶Division of Clinical Neurosciences, University of Turku, Turku, Finland, ⁷Department of Neurosurgery, Antwerp University Hospital and University of Antwerp, Edegem, Belgium, ⁸Division of Anaesthesia, University of Cambridge, Addenbrooke’s Hospital, Cambridge, UK, ⁹School of Medical Sciences, Örebro University, Örebro, Sweden, ¹⁰Department of Clinical Neurosciences, Neurosurgery Unit, University of Cambridge, Addenbrooke's Hospital, Cambridge, UK, ¹¹Department of Neurological Surgery, University of California – San Francisco, San Francisco, USA, ¹²Department of Radiology, Massachusetts General Hospital, Boston, USA

Introduction: Acute subdural hematoma (ASDH) is the most common space-occupying mass lesion in patients with traumatic brain injury (TBI). The decision to treat these patients either surgically or conservatively is typically made on neurological and CT findings. However, the selection of patients for acute surgery may benefit from objective decision support with blood-based biomarkers. Our aim was to examine if protein biomarkers can be used to identify those who benefit from acute surgical treatment of ASDH.

Methods: Our study cohort (CENTER-TBI) included 609 patients with traumatic ASDH and blood biomarker levels of Glial Fibrillary Acidic Protein (GFAP), Tau, Neuron-Specific Enolase (NSE), Neurofilament-Light (NF-L), S100 Calcium-Binding Protein B (S100B), Ubiquitin C-terminal Hydrolase L1 (UCH-L1) available within 24 h of injury. Outcome was assessed with Glasgow Outcome Scale Extended (GOSE) 6 months after the injury. Surgical patients were divided into favorable (GOSE≥5) vs. unfavorable outcome (GOSE≤4). Additionally, we did a subgroup analysis of surgical ASDH patients who died (GOSE=1). Comparison analyzes were done using Mann-Whitney U -test and ability to differentiate between groups was assessed with AUC, where ≥0,7 was considered clinically adequate.

Results: 184 (30%) were treated surgically and 425 (70%) conservatively. Biomarker levels did not differ significantly between the treatment groups in the whole study cohort. However, there was a significant difference in biomarker levels between the outcome groups. The AUCs in discriminating between outcome groups were 0,65 for GFAP, 0,68 for Tau, 0,62 for NSE, 0,64 for NF-L, 0,7 for S100B, 0,68 for UCH-L1. For the subgroup analysis, the AUCs were 0,73 for GFAP, 0,73 for Tau, 0,64 for NSE, 0,74 for NF-L, 0,74 for S100B, 0,7 for UCH-L1.

Conclusions: Biomarkers were not able to adequately discriminate between favorable and unfavorable outcome. However, GFAP, Tau and S100B were able to differentiate surgically treated ASDH patients surviving and dying.

P071 Identification of an Isolated Epidural Hematoma Using Blood-Based Biomarkers in Traumatic Brain Injury

BM Katja Malmi ,¹ Dana Pisica,^2,3 Inftakher Hossain,^1,9 Mehrbod Mohammadian,^1,4 Teemu Luoto,⁵ Riikka Takala,⁶ Olli Tenovuo,⁷ Thomas van Essen,^8,9,10 John Yue,¹¹ Rick Vreeburg,⁸ Andrew Maas,¹² David Menon,¹³ András Büki,¹⁴ Jussi Posti¹

¹Neurocenter, Department of Neurosurgery and Turku Brain Injury Center, Turku University Hospital and University of Turku, Turku, Finland, ²Department of Neurosurgery, Erasmus MC - University Medical Center, Rotterdam, the Netherlands, ³Center for Medical Decision Making, Department of Public Health, Erasmus MC - University Medical Center, Rotterdam, the Netherlands, ⁴Department of Radiology, Massachusetts General Hospital, Boston, USA, ⁵Department of Neurosurgery, Tampere University Hospital and Tampere University, Tampere, Finland, ⁶Perioperative Services, Intensive Care Medicine and Pain Management, Turku University Hospital and University of Turku, Turku, Finland, ⁷Division of Clinical Neurosciences, University of Turku, Turku, Finland, ⁸University Neurosurgical Center Holland, Leiden University Medical Center, Haaglanden Medical Center, HAGA, Leiden and The Hague, the Netherlands, ⁹Division of Neurosurgery, Department of Clinical Neurosciences, University of Cambridge and Addenbrooke's Hospital, Cambridge, United Kingdom, ¹⁰Department of Surgery, Division of Neurosurgery, QEII Health Sciences Centre and Dalhousie University, Halifax, Nova Scotia, Canada, ¹¹Brain and Spinal Injury Center, Department of Neurological Surgery, Zuckerberg San Francisco General Hospital, University of California San Francisco, San Francisco, USA, ¹²Department of Neurosurgery, Antwerp University Hospital and University of Antwerp, Edegem, Belgium, ¹³Division of Anaesthesia, University of Cambridge, Addenbrooke’s Hospital, Cambridge, United Kingdom, ¹⁴School of Medical Sciences, Örebro University, Örebro, Sweden

Background: Blood-based biomarkers have shown promise in identifying head injury patients with traumatic intracranial findings, however their capacity in detecting patients with isolated extra-axial hematomas remains unknown. We aimed to assess the potential of S100 calcium-binding protein B (S100B), glial fibrillary acid protein (GFAP), total tau (t-tau), neurofilament light (NF-L), ubiquitin C-terminal hydrolase L1 (UCH-L1), and neuron specific enolase (NSE) to identify mild TBI (mTBI) patients with isolated epidural hematoma (EDH).

Methodology: This prospective study included 1048 patients (median age 46, interquartile range 28–63) with mTBI (GCS ≥13) with 1) all six serum blood biomarker levels measured and 2) head computed tomography (CT) scan within 24 hours of injury. The main outcome measure was biomarkers' area under the receiver operating characteristic curve (AUC) in distinguishing between patients with CT-negative findings and with isolated EDH. AUC≥0.7 was considered clinically adequate. Distinction was made between i) isolated EDH and ii) EDH coexisting with non-surgical findings (e.g. subarachnoid haemorrhage, intraventricular haemorrhage, axonal injury). In sensitivity analyses, patients with major extracranial injury (MEI) were excluded.

Results: Of the patients, 38 (3,6%) patients had an isolated EDH, 28 (2,7%) patients had EDH with non-surgical lesions and 982 patients had CT-negative findings (94%). In discriminating between any EDH versus CT-negative, the AUCs were 0.82 for GFAP, 0.65 for NSE, 0.65 for t-tau, 0.60 for NF-L, 0.57 for S100B, and 0.66 for UCH-L1. For discriminating between isolated EDH and CT-negative, the AUCs were 0.77 for GFAP, 0.64 for NSE, 0.60 for t-tau, 0.50 for NF-L, 0.50 for S100B, and 0.60 for UCH-L1. When excluding 227 patients with MEI, discrimination increased for all 6 biomarkerks. The GFAP cut-off with 100% negative predictive value for identifying isolated EDH was 0.19 ng/ml regardless of the presence of MEI.

Conclusion: Blood levels of GFAP adequately discriminated mTBI patients with isolated EDH.

P072 A novel rodent model of spinal cord injury reveals radiological biomarkers in long bones for the detection of concomitant mild traumatic brain injury

Mr Paul Marciano ,^1,2 Miss Keziah Skein,^2,3 Miss Rachel Hollyoak,³ Miss Reeya Patel,^1,2 Miss Sandra Jenkner,^2,3 Dr Agnieszka Arthur,³ Associate Professor Frances Corrigan,³ Dr Anna Leonard,^2,3 Dr Ryan O' Hare Doig^1,2,4

¹Adelaide Medical School, Faculty of Health and Medical Science,The University of Adelaide, Adelaide, Australia, ²Neil Sachse Centre for Spinal Cord Research, Lifelong Heath Theme, South Australia Health and Medical Research Institute, Adelaide, Australia, ³School of Biomedicine, Faculty of Health and Medical Science, The University of Adelaide, Adelaide, Adelaide, Australia, ⁴Preclinical Imaging and Research Laboratories, South Australia Health and Medical Research Institute, South Australian Node, National Imaging Facility, Adelaide, Australia

Introduction: Concomitant traumatic brain injury (TBI) at the time of spinal cord injury (SCI) risks physical and cognitive impairment. Unfortunately, there are limited clinical strategies to identify concomitant injuries, which presents challenges to appropriately identify and rehabilitate patients. Current literature supports the hypothesis that both SCI and TBI alone have peripheral effects, including long-term consequences in bone remodelling. However, it is unknown whether a concomitant injury results in unique bone remodelling compared to a single neurotrauma. Therefore, in a newly developed rodent model of TBI+SCI, we explored whether specific bone tissue biomarkers could diagnose concomitant trauma.

Methods: Femurs harvested from male Sprague-Dawley rats (N= 47), 6-weeks post-injury (Naive, Sham, SCI [C5 hemi-contusion, Infinite Horizon;100kdyne or 200kdyne], mild TBI [Marmarou weight drop, 1m], TBI+SCI [mild/moderate]) underwent micro-computed tomography (35μm, 85kV, 235mA), reconstruction (NRecon software) and analysis (Dragonfly software). Within each proximal femur, cortical and trabecular bone were measured in three regions: subchondral bone; primary spongiosa; and secondary spongiosa. Bone volume fraction (BV/TV), Trabecular number (Tb.N), Trabecular thickness (Tb.Th), Trabecular separation (Tb.Sp), and Cortical thickness (Ct.Th) were analysed.

Results: mTBI and SCI alone induced bone remodelling evidenced by significant decreases in BV/TV (5.4–5.10%), Tb.Th (1.5–2.7%) and Ct.Th (7.7–9.4%). Additive effects of TBI+SCI were evident in primary spongiosa, with a observed increased Tb.Th (6.8 ± 3μm3) and Ct.Th (24.7 ± 11.2μm3). Contrasingly, TBI+SCI demonstrated significant decline in BV/TV (-0.061 ± 0.027μm3) and Tb.N (-599.2 ± 148μm3) of primary spongiosa, in comparison to SCI.

Conclusion: Limited strategies exist to detect concomitant mTBI and SCI. Using a clinically relevant imaging modality, we demonstrated for the first time, unique bone remodelling outcomes post-injury. Further histological and proteomic characterisation is currently underway. Therefore, non-invasive medical imaging and quantification of biomarkers in the proximal femur may serve as a tool for diagnosis of undetected concomitant trauma.

P073 Blood biomarkers, diffusion tensor imaging, and outcome after mild traumatic brain injury

Malla Mononen ,^1,2 PhD Mehrbod Mohammadian,¹ MD, PhD Iftakher Hossain,^1,2,3 PhD Timo Roine,^1,4 MD, PhD Olli Tenovuo,^1,2 MD, PhD Kaj Blennow,^6,7 PhD, FmedSci Peter Hutchinson,³ MD Henna-Riikka Maanpää,^1,2 PhD, FmedSci David K. Menon,³ MD, PhD Virginia F. Newcombe,³ PhD Jean-Charles Sanchez,¹⁰ MD, PhD Riikka S.K. Takala,^1,2 MD Jussi Tallus,^1,2 PhD Mark van Gils,⁵ MD, PhD Henrik Zetterberg,^6,7,8,9,11 MD, PhD Jussi P. Posti^1,2

¹University of Turku, Turku, Finland, ²Turku University Hospital, Turku, Finland, ³University of Cambridge, Cambridge, United Kingdom, ⁴Aalto University School of Science, Espoo, Finland, ⁵University of Tampere, Tampere, Finland, ⁶University of Gothenburg, Gothenburg, Sweden, ⁷Sahlgrenska University Hospital, Gothenburg, Sweden, ⁸University College London, London, Finland, ⁹Hongkong Center for Neurodenegerative Diseases, Hong Kong, China, ¹⁰University of Geneva, Geneva, Switzerland, ¹¹University of Wisconsin, Madison, USA

Background: Association of blood levels of the axonal biomarker neurofilament light and brain white matter (WM) injury in mild traumatic injury (mTBI) has been reported. The purposes of this study were to assess the association between the levels of non-axonal blood-based biomarkers at admission and WM integrity, measured using post-acute diffusion tensor metrics, and to investigate their associations with outcome in patients with mTBI.

Methods: 92 patients with mTBI (Glasgow Coma Scale ≥ 13) having plasma samples for glial fibrillary acidic protein (GFAP), interleukin 10 (IL-10), heart fatty-acid binding protein (H-FABP), S100 calcium-binding protein B (S100B), total tau (T-Tau), amyloid beta 40 and 42 (amyloid B40 and B42) within 24 h of admission and diffusion-weighted magnetic resonance imaging (DW-MRI) ≥ 90 days post-injury (median = 231) were included. Patients were divided into computed tomography (CT)-positive and CT-negative subgroups. Outcome was assessed using Glasgow Outcome Scale-Extended (GOSE) at the time of imaging during the follow-up visit. Outcomes were dichotomized as complete (GOSE 8) and incomplete recovery (GOSE <8). Mean fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) were calculated from the skeletonized WM tracts of the whole brain.

Results: Admission levels of GFAP, IL-10, and T-Tau had negative correlations with FA and positive correlations with MD, RD, and AD in whole mTBI cohort, in CT-positive group (excluding IL-10), and in incomplete recovery group. In contrast, AB-42 had a positive correlation with FA in whole mTBI cohort. The levels of S100B had a positive correlation with AD in CT-positive group. CT-positive patients had significantly lower FA levels and higher levels of MD and RD than the CT-negative patients.

Conclusion: Admission levels of GFAP, IL-10, and T-Tau correlate with WM integrity measured ≥ 90 days post-injury. Higher levels of these biomarkers may be associated with diffuse axonal injury.

P074 A plasma phospholipid signature as a possible biomarker of the acute phase of traumatic brain injury

Dr Isabell Nessel ,¹ Dr Simon C Dyall,² Dr Luke Whiley,³ Prof Adina T Michael-Titus¹

¹Centre for Neuroscience, Surgery and Trauma, Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, London, United Kingdom, ²School of Life and Health Sciences, University of Roehampton, London, United Kingdom, ³Health Futures Institute, Murdoch University, Murdoch, Australia

Introduction: Traumatic brain injury (TBI) triggers complex neurochemical and metabolic changes, including membrane phospholipid (PL) breakdown. We investigated plasma PL changes in the first 3 days post-TBI, to assess whether such changes could become a marker of human acute TBI.

Methodology: Patients with TBI (Abbreviated Injury Scale (AIS)3 indicating serious injury, n=5; AIS4 indicating severe injury, n= 8), and controls (n= 13), were selected from the QMUL Trauma-Biobank. Plasma samples were analysed for PLs by LC-MS. Neurofilament light (NFL) and pro-inflammatory cytokines were measured using electro-chemiluminescence. Red blood cell omega-3 index was calculated after fatty acid analysis by gas chromatography.

Results: NFL levels were significantly increased at 24 and 72 h after injury in AIS4 TBI cases, vs. controls. IL-6 was significantly elevated 24 h after injury in AIS4 patients. LPC (18:0/0:0) and (16:0/0:0) and PC (40:8) and (36:4) were significantly decreased 24 h after TBI and were still significantly lower 72 h after injury in AIS4. Similar changes were seen for LPE, PE and SM. Furthermore, orthogonal projections to latent structures discriminant analysis revealed specific lipid patterns, separating AIS4 TBI samples at 24 and 72 h from controls. Over the first 3 days post-injury, the omega-3 index did not change significantly; notably, baseline levels were low in all patients (controls: 4.3±1.1% and TBI: 4.0±2.1%).

Conclusion: We have identified a range of changes affecting various PLs in the acute phase post-TBI. After confirmation in larger patient cohorts, this could inform the development of new lipid biomarker panels and also new therapies in TBI.

P075 Dynamics of synaptic damage in patients after traumatic brain injury

Mr. Florian Olde Heuvel ,¹ Dr. Zhenghui Li,^1,2 Dr. Patrick Oeckl,^1,3 Prof. Sandy Schulz,⁴ Prof. Cristina Morganti-Kossmann,⁵ Prof. Bridgette Semple,⁴ Prof. Francesco Roselli^1,3

¹Ulm University, Ulm, Germany, ²Zhengzhou University, Zhengzhou, China, ³DZNE Ulm, Ulm, Germany, ⁴Monash University, Melbourne, Australia, ⁵University of Arizona College of Medicine, Phoenix, United States of America

Introduction: Traumatic brain injury (TBI) is characterized by neuronal damage, vascular impairment and neuroinflammation. While a large component of the brain parenchyma is made of synapses, the extent of synaptic damage that occurs independent of neuronal loss remains unexplored. We aimed to investigate the time course of synaptic damage in cerebrospinal fluid of patients with severe TBI and determine the relationship between synaptic damage and clinical outcome.

Methodology: We used the high-sensitivity SIMOA and Ella platforms to quantify the levels of synaptic markers SNAP25, VILIP-1 together with NFL, UCH-L1, IL-6 and IL-8 in cerebrospinal fluid (CSF) samples from TBI patients between the day of injury up to discharge or catheter removal. Correlations with injury severity and outcome prediction have been performed.

Results: We found a strong elevation of the synaptic markers at D0 and D1 respectively, with a rapid decrease from D3, however a subset of patients did not return to baseline. This was followed by a secondary elevation from D7 in a subset of patients. Strong correlations of SNAP-25 and VILIP-1 were observed with the neuronal and axonal injury markers NFL and UCH-L1 until D5. The synaptic markers correlate with the neuroinflammatory markers, IL-6 and IL-8, at two time points (D1 and D5). Synaptic markers on the day-of-injury have better sensitivity and specificity for unfavourable outcome at 6 months than NFL or UCH-L1. Later elevation of synaptic markers was associated with poorer outcome.

Conclusion: Our data reveal that SNAP-25 and VILIP-1 levels in CSF are strongly affected by TBI, with temporal patterns only partially overlapping with those reflecting neuronal damage but not neuroinflammation. Thus, synaptic damage seems to display a specific pathophysiological mechanism post TBI, having a potential prognostic relevance. Thus, markers of synaptic damage constitute a new potential for assessing the diagnosis and prognosis in patients with TBI.

P077 Portable sleep monitors as digital biomarkers compared to polysomnography in patients with traumatic brain injury: systematic review and meta-analysis (PRISMA)

Thom Van Der Meer,^1,2 Sandra Abeijon,² Sarah Schiebler,³ Beatrice Nasta,¹ Joerg Steier,⁴ Katrin Rauen ^1,5,6,7

¹University of Zurich, University Hospital of Psychiatry Zurich, Zurich, Switzerland, ²Neuroscience Centre Zurich (ZNZ), Zurich, Switzerland, ³Department of Consultation-Liaison Psychiatry and Psychosomatic Medicine, University Hospital Zurich, Zurich, Switzerland, ⁴Centre for Human Applied Physiological Sciences, (CHAPS), Faculty of Life Sciences and Medicine, London, United Kingdom, ⁵Competence Centre Sleep & Health Zurich, University of Zurich, Zurich, Switzerland, ⁶Neuroscience Centre Zurich, University of Zurich, Zurich, Switzerland, ⁷Institute for Stroke and Dementia Research (ISD), University Hospital, LMU Munich, Münich, Germany

Introduction: Sleep plays a pivotal role in recovery of traumatic brain injury (TBI), up to 70% of TBI patients suffer from sleep disorders, often remaining undiagnosed. Polysomnography (PSG) is the gold standard for sleep measurements. However, PSG studies are scarce, and TBI patients often have a disturbed day-night-rhythm. Therefore, this pre-registered PROSPERO systematic review and meta-analysis reports on the clinical value of portable sleep monitors (PM’s) compared to PSG in TBI patients.

Methodology: A comprehensive literature search was performed in MEDLINE, EMBASE, APA PSYCHINFO, COCHRANE, SCOPUS. Two independent researchers reviewed based on predefined eligibility criteria using Covidence, R, and the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2). Qualitative analysis reviewed type II-IV PM’s compared to PSG in patients with TBI. Quantitative analyses reviewed the diagnostic accuracy of PM’s versus PSG in TBI (AHI/sPO2 parameters); and the overall sleep macro-architecture (13 parameters) of TBI patients versus healthy adults (expressed in forest-plots).

Results: Out of 13′764 original English peer-reviewed articles, 20 were eligible for the systematic review, and one for the meta-analysis. A total sample size of 870 TBI patients showed a mean age (± SD) of 39.4 ± 14.0 years; TBI severity and time post-injury were highly heterogenic. Qualitative synthesis showed that type II-III PM’s pose as comparable diagnostic devices to PSG in all TBI severities, and can remove the “first-night-effect”. Quantitative synthesis revealed type II-III PM’s having a negative standardized mean difference (SMD) of -0.12 compared to PSG; non-REM sleep stage 2 was at 1.37 (SMD) higher in TBI patients’ sleep macro-architecture compared to healthy controls, indicating pronounced superficial sleep post-TBI.

Conclusions: Portable sleep monitors seem to serve as beneficial digital diagnostic biomarkers compared to polysomnography (PSG) in specific subgroups of TBI patients, able to remove the “first-night-effect”, reducing personnel efforts and costs, and potentially helping minimise sleep burden post-TBI.

P078 The utility of antisaccade assessment to assess cognitive dysfunction in patients with a single TBI: systematic review and meta-analysis

Dr Georgia Symons ,^1,2 Dr Elysia Sokolenko,³ Ms Sophia Mast,^2,4 Mr Thom Van Der Meer,^2,4 Prof Sandy Shultz,^1,5 Prof Katrin Rauen^2,4,6,7

¹Department of Neuroscience, School of Translational Medicine, Monash University, Melbourne, Australia, ²Department of Geriatric Psychiatry, Psychiatric Hospital Zurich, University of Zurich, Zurich, Switzerland, ³Discipline of Anatomy and Pathology, School of Biomedicine, University of Adelaide, Adelaide, Australia, ⁴Neuroscience Center Zurich, University of Zurich, Zurich, Switzerland, ⁵Health Sciences, Vancouver Island University, Vancouver, Canada, ⁶Institute for Stroke and Dementia Research (ISD), University Hospital, LMU Munich, Munich, Germany, ⁷Neurological Rehabilitation Center Godeshöhe, Bonn, Germany

Introduction: Traumatic brain injury (TBI) results in persistent cognitive dysfunction that is often not sufficiently diagnosed. Ocular motor assessment of antisaccades might be a promising biomarker for identifying cognitive dysfunction. Still, it has not yet been established in adults with a single TBI, forming the aim of this systematic review and meta-analysis.

Methodology: A preregistered PROSPERO protocol (CRD42023444933), according to the PRISMA guidelines, was established. A literature search was conducted on 7th March 2023 (MEDLINE, EMBASE, APA PSYCHINFO, COCHRANE, SCOPUS) to identify original, English, and peer-reviewed research articles using antisaccades to assess cognition in adult single TBI patients. Two independent reviewers screened articles based on the PRISMA guidelines using the Covidence software for systematic reviews, and descriptive and preliminary quantitative analyses were applied.

Results: Overall, 11 out of 640 original articles met the criteria for systematic review criteria and only four for the meta-analysis. The qualitative analysis included 281 TBI and 202 healthy controls, and 72% of studies focused on mild TBI outcomes. Studies included a range of TBI etiologies and time points post-injury. Antisaccade protocols were highly mixed; none of the included studies referenced the internationally standardized antisaccade protocol. The two most reported outcomes were directional error (91%) and latency (81%), which reported increased errors and longer latencies in all TBI severities and timepoints. Preliminary quantitative analysis revealed that the TBI groups had significantly longer antisaccades latencies compared to age and sex-matched healthy controls (p <0.01). Directional error analysis is ongoing and will be presented.

Conclusion: Antisaccade latency was significantly longer in adults with a single TBI compared to healthy controls and might offer a useful clinical biomarker for identifying cognitive dysfunction at various severities and timepoints post-TBI. However, methodological inconsistencies are evident between studies. Therefore, a widespread standardized approach is necessary for further validation and clinical implementation.

P079 Acute plasma levels of brain specific proteins related to tryptophan and kynurenine pathway metabolites following mild traumatic brain injury

Mr Koen Visser ,¹ dr. Myrthe E. de Koning,² dr. Bram Jacobs,¹ dr. Walid Chayoua,³ dr. Arno R. Bourgonje,^4,5 Mr. Martijn van Faassen,⁶ Prof. dr. Ido Kema,⁶ Prof. dr. Harry van Goor,⁷ Prof. dr. Joukje van der Naalt,¹ dr. Harm-Jan van der Horn¹

¹Department of Neurology, University Medical Center Groningen, Groningen, Netherlands, ²Department of Neurology, Medical Spectrum Twente, Enschede, Netherlands, ³Department of Clinical Chemistry and Laboratory Medicine, Medlon BV, Enschede, Netherlands, ⁴Department of Gastroenterology and Hepatology, University Medical Center Groningen, Groningen, Netherlands, ⁵The Henry D. Janowitz Division of Gastroenterology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York., U.S.A., ⁶Department of Laboratory Medicine, University Medical Center Groningen, Groningen, Netherlands, ⁷Division of Pathology of the Department of Pathology and Medical Biology, University Medical Center Groningen, Groningen, Netherlands

Introduction: Acute degradation of the essential amino acid Tryptophan (Trp) has been associated with long term functional recovery after mild traumatic brain injury (mTBI). The kynurenine pathway (KP) is the main degradation pathway of Trp and is comprised of (neuroactive)metabolites that are implicated in the pathophysiology of mTBI. However, studies have only quantified Trp and KP metabolites in blood – leaving uncertainty about the extent to which these markers associate with brain injury. In this study, acute plasma levels of Trp and major KP metabolites are related to brain specific proteins and findings on head-CT after mTBI.

Methods: In a prospective longitudinal cohort study (AIM-TBI), 252 patients with mTBI were recruited. Blood-samples were acquired <24 hours of injury; all participants underwent acute head CT-scan. Glial Fibrillary Acidic Protein (GFAP), Ubiquitin C-terminal Hydrolase (UCH-L1), neurofilament light (NF-L), and KP metabolites were measured. The following generalized linear models were run: KP metabolite/Trp ∼ 1 + GFAP + UCH-L1 + NF-L+ age + sex + time of day. An additional model was run with the presence of CT lesions as an independent variable.

Results: Most samples were obtained within 2 hours of mTBI [median (IQR); 107 (120)]. At a significance level corrected for multiple testing (n=5, P<0.01) GFAP was associated with Quinolinic acid [Wald χ2=10.55, P=0.001, β=-0.11] and NFL with the Kynurenic acid/Quinolinic acid ratio [Wald χ2=6.89, P=0.009, β=-0.09]. Lesions on head CT were identified in 55 (22%) of participants. The presence of CT lesions was associated with Kynurenine [Wald χ2=7.32, P=0.007, β=-0.12].

Discussion: Our findings reveal that acute plasma KP metabolites are associated with brain specific protein release, indicating a trend for brain injury. Investigation of kinetic profiles is needed to determine the exact associations between Trp, KP metabolites and brain specific proteins after mTBI and the clinical relevance of these associations.

P080 Prevalence of cavum septum pellucidum in athletes

Elizabeth Hovenden ,¹ Dr. Hannah Lindsey,^1,2 Josephine Dimanche,¹ Dr. Paula Johnson,¹ Dr. Carrie Esopenko,^1,3 Dr. Michael Larson,⁴ Dr. Larry Carr,¹ Finian Keleher,¹ Dr. Emily Dennis,^1,2 Dr. Elisabeth Wilde^1,2

¹University of Utah School of Medicine, Department of Neurology, Salt Lake City, United States, ²George E. Wahlen Veterans Affairs Medical Center, Salt Lake City, United States, ³Icahn School of Medicine at Mount Sinai, New York, United States, ⁴Brigham Young University Department of Psychology and Neuroscience Center, Provo, United States

Introduction: Cavum septum pellucidum (CSP) is a cavity between the membranous leaves of the septum pellucidum. In recent years, presence of CSP has emerged as an area of increased interest in populations experiencing repetitive head acceleration events (RHAEs). Historically, CSP has been considered an incidental finding of unexplained significance; however, there is increasing evidence that disrupted CSP may have functional consequences given its proximity to the limbic system. Although some literature exists regarding this variation in boxers and retired American football players, relatively little is known about the prevalence and impact of CSP in younger athletes. We assessed the presence of CSP in three groups of contact-sport athletes and ex-athletes endorsing a history of RHAEs.

Methodology: Participants across three cohorts (n=89, 78 males and 11 females, ages 18–65) underwent magnetic resonance imaging. Cohort 1 (n=35; age M=21.8, SD=1.4) included male collegiate American football team members, cohort 2 (n=7; age M=20.7, SD=1.5) included collegiate female soccer team members, and cohort 3 (n=47; age M=45.85, SD=14.49) included current and former athletes with a history of RHAE (43M, 4F). Raw imaging data was AC-PC aligned before utilizing ITK-Snap to evaluate T1-weighted images. Axial and coronal planes were assessed to determine presence of CSP.

Results: Our findings indicate a higher frequency of CSPs than in general populations. Across cohorts, we found that 84 of 89 participants presented with CSP, with 32/35 (cohort 1), 7/7 (cohort 2), and 45/47 (cohort 3). Interestingly, all females in these cohorts presented with CSP.

Conclusion: We discovered a higher-than-expected incidence of CSP in athletes endorsing RHAE. Although implications of these findings are not yet clear, this is among the first studies to demonstrate CSP in younger, otherwise healthy athletes and warrants further investigation. Future studies will examine specific measurements and severity of CSP and its relationship to cognitive outcome.

P081 Impact of Head Injuries in University Sports Clubs

Ms Elika Karvandi ,¹ Prof Peter Hutchinson, Mr Adel Helmy

¹University of Cambridge, Cambridge, United Kingdom

Participation in sports is a major part of university life. As there is growing awareness on the effects of head injuries in sports, this study assessed the impact of head injuries on university students participating in university sports clubs to inform safety protocols and improving student wellbeing.

A survey was distributed to university sports clubs at the University of Cambridge. Questions were related to their head injury experience, protocol compliance, reporting practices, and impact on their sport participation and studies. All students participating in sports were eligible to complete the survey, with or without a history of mild TBI.

A total of 121 students completed the survey with a completion rate of 66%. Of the participants who reported a head injury, about a third said it took them longer than a month to recover and return to play. Most students did not follow return to play protocols, and only around half received medical attention for their head injury. Only 59% reported to have received education on head injuries.

Overall, some students experience disruptions to their studies and had a poor experience with the available student support after their injury. There needs to be more education on head injuries, better guidelines, and improvements to support services accessible to student athletes.

P082 Sex-specific cerebral blood flow alterations in soccer players with repetitive head impacts

Mr. Zheyuan Li ,^1,2 Mr. Alberto Villagran,^1,2 Ms. Luisa Schuhmacher,^1,2,3 Ms. Julie Joyce,^1,4 Ms. Anja Betz,^1,2 Mr. Lasse Lohmann,^1,2 Prof. Dr. Inga Koerte^1,2,3

¹cBRAIN, Department of Child and Adolescent Psychiatry, Psychosomatics, and Psychotherapy, Ludwigs-Maximilians-Universität, Munich, Germany, ²NICUM (NeuroImaging Core Unit Munich), Munich, Germany, ³Psychiatry Neuroimaging Laboratory, Department of Psychiatry, Mass General Brigham, Harvard Medical School, Boston, USA, ⁴Center for Clinical Spectroscopy, Department of Radiology, Mass General Brigham, Harvard Medical School, Boston, USA

Introduction: Repetitive head impacts (RHI) in soccer are common and raise concerns for long-term neurodegenerative disease. However, the precise pathophysiological effects on the brain remain unclear. Cerebral blood flow (CBF) alteration may serve as a biomarker. Here, we aimed to investigate sex-specific alterations in CBF among soccer players.

Methodology: We assessed 41 age-matched soccer players (24 males, 17 females) with at least 5 years of exposure to RHI, alongside 46 controls (CTL) (21 males, 25 females) with less than 5 years of exposure to RHI. Participants, aged between 20 to 25 years, completed clinical assessments and underwent 3T PRISMA MRI. CBF was assessed using pulsed arterial spin labeling (PASL). Data was preprocessed using FSL6.0.6.5. Regional analyses were conducted to compare CBF between groups and sexes across global gray matter, hemispheres, lobes, and deep/cortical gray matter. ANCOVA was employed, covarying for age, education (years), mean arterial blood pressure, and body mass index.

Results: Female CTL individuals demonstrated significantly higher CBF in global gray matter compared to male CTL (p = 0.002, Δ[95%CI] = 9.0 [3.8, 14.2] ml/100g/min), as did female RHI compared to male RHI (p < 0.001, Δ[95%CI] = 14.1 [8.2, 19.9] ml/100g/min). Additionally, female RHI exhibited higher CBF in global gray matter compared to female CTL (p = 0.018, Δ[95%CI] = 6.6 [1.4, 11.8] ml/100g/min). The male RHI group demonstrated elevated CBF in regions of the occipital lobe compared to male CTL (p = 0.046, Δ[95%CI] = 4.9 [0.3, 9.4] ml/100g/min).

Conclusions: Our findings are in line with previous research showing that females have higher CBF than males. Compared to their respective control groups, male and female athletes exposed to RHI show higher CBF with regionally-specific sex differences. Increased CBF may potentially be due to enhanced demand through neuroinflammatory processes. Future research is needed to elucidate pathophysiological mechanisms.

P083 Incorporation of a dual-task cognitive exercise does not affect unsupported seated balance in university para-sport athletes

Dr. Ryan Moran ,¹ Alexandra Curry¹

¹The University of Alabama, Tuscaloosa, United States

Introduction: Balance is a key component of sport-related concussion evaluation and rehabilitation. Athletes living with spinal cord injury and pathologies that secondarily affect the spinal cord, use a wheelchair for sport and daily living, making validated balance measures unreliable. Dual-tasks have yet to be explored during balancing in para-sport athletes. Therefore, the objective was to determine if a dual-task cognitive exercise during unsupported sitting on a stable and unstable surface induces greater balance errors in university para-sport athletes.

Methodology: 52 uninjured university para-sport athletes from across the United States completed a balance assessment consisting of single and dual-task unsupported upright seated balancing on a stable and unstable surface out of their wheelchair. Measures included errors consistent with the Wheelchair Error Scoring System during the 2 (task) x 2 (surface) battery. Wilcoxon Signed-Rank Tests were used to compare single and dual-task errors between surfaces.