Faculty Bibliography

The 2024 revised McDonald criteria for multiple sclerosis recognize the optic nerve as a topography for dissemination in space. Optical coherence tomography-derived inter-eye differences in peri-papillary retinal nerve fiber layer or ganglion cell-inner plexiform layer thicknesses (≥6μm or ≥4μm, respectively) are proposed for identifying unilateral optic nerve involvement. However, the value of combining inter-eye difference measures and optimal temporal-quadrant peri-papillary retinal nerve fiber layer inter-eye differences remains unclear. We investigated the diagnostic performance of combined inter-eye differences, optimal temporal-quadrant peri-papillary retinal nerve fiber layer inter-eye differences, and examined the effects of time, prior optic neuritis frequency, sex, and race on inter-eye differences. Retinal optical coherence tomography images from all study participants underwent rigorous quality control. Receiver operating characteristic analyses and area under the receiver operating characteristic curves (AUC) were used to determine optimal inter-eye differences of individual and combined measures to distinguish eyes with, from without, prior optic neuritis in people with multiple sclerosis. Mixed-effects models were used to assess impact of time, prior optic neuritis events, sex, and race on inter-eye differences. An independent multiple sclerosis cohort from a second center was examined for external validation. Among 1854 people with multiple sclerosis, optimal inter-eye difference thresholds for identifying unilateral optic nerve involvement were 6μm for peri-papillary retinal nerve fiber layer (AUC=0.80), 4μm for ganglion cell-inner plexiform layer (AUC=0.83), and 8μm for temporal-quadrant peri-papillary retinal nerve fiber layer (AUC=0.71) thicknesses. Peri-papillary retinal nerve fiber layer inter-eye differences ≥6μm or ganglion cell-inner plexiform layer inter-eye differences ≥4μm yielded 87.6% sensitivity, 70.0% specificity, and 64.0% positive predictive value. Concurrent inter-eye differences at lower thresholds (≥5μm peri-papillary retinal nerve fiber layer, ≥3μm ganglion cell-inner plexiform layer) reduced sensitivity to 72.5%, but improved specificity (86.6%) and positive predictive value (76.7%), while maintaining accuracy and negative predictive value. Temporal-quadrant peri-papillary retinal nerve fiber layer inter-eye differences did not improve diagnostic performance. Over a median of 5.1 years, ganglion cell-inner plexiform layer and peri-papillary retinal nerve fiber layer inter-eye differences remained stable. Prior optic neuritis counts and sex did not affect inter-eye differences. Although Black Americans had higher inter-eye differences than White Americans, optimal thresholds were comparable across races. The validation cohort comprising 254 people with multiple sclerosis confirmed these findings. In conclusion, concurrent peri-papillary retinal nerve fiber layer (≥5μm) and ganglion cell-inner plexiform layer inter-eye differences (≥3μm) improve unilateral optic nerve involvement detection versus either alone (≥6μm or ≥4μm, respectively), while temporal-quadrant peri-papillary retinal nerve fiber layer inter-eye differences offer limited benefit. Inter-eye differences remain stable longitudinally and unaffected by prior optic neuritis frequency.

Consciousness Science is entering an age of unprecedented opportunity, thanks to recent empirical and theoretical advances, increasing interest in the topic, and technological advances in neuroscience. The role theories will play in a maturing science of consciousness deserves a closer look.

Hippocampal sharp-wave ripples (SPW-Rs) are high-frequency oscillations critical for memory consolidation. Despite extensive characterization in rodents, their detection in humans is limited by coarse spatial sampling, interictal epileptiform discharges (IEDs), and a lack of consensus on human ripple localization and morphology. Here, we demonstrate that mouse and human hippocampal ripples share spatial, spectral and temporal features, which are clearly distinct from IEDs. In recordings from male APP/PS1 mice, SPW-Rs were distinguishable from IEDs by multiple criteria. Hippocampal ripples recorded during NREM sleep in female and male surgical epilepsy patients exhibited similar narrowband frequency peaks and multiple ripple cycles in the CA1 and subiculum regions. Conversely, IEDs showed a broad spatial extent and wide-band frequency power. We developed a semi-automated, ripple curation toolbox (ripmap) to separate event waveforms by low-dimensional embedding to reduce false-positive rate in selected ripple channels. Our approach improves ripple detection and provides a firm foundation for future human memory research.

Carotid webs are increasingly recognized as an underdiagnosed etiology of ischemic stroke, especially in young, otherwise healthy patients. These fibrous intimal protrusions create regions of flow stasis within the internal carotid artery, predisposing to thromboembolism. Diagnosis remains challenging due to their subtle radiographic appearance and underappreciation in clinical practice. While antiplatelet therapy or anticoagulation used to be the cornerstone of management, medical therapy alone has been found to be insufficient for stroke prevention in symptomatic patients. Definitive intervention includes carotid artery stenting or carotid endarterectomy; both have demonstrated excellent safety and efficacy. Risk stratification for symptomatic and asymptomatic carotid webs remains an area of active research, with emerging evidence suggesting that specific anatomic features, termed the carotid web angioarchitecture, may help predict stroke risk. Further studies are needed to determine the role of preventative intervention. A deeper understanding of carotid web pathogenesis, natural history, and hemodynamic impact is critical for guiding clinical decision-making.

Large Language Models (LLMs) offer a framework for understanding language processing in the human brain. Unlike traditional models, LLMs represent words and context through layered numerical embeddings. Here, we demonstrate that LLMs' layer hierarchy aligns with the temporal dynamics of language comprehension in the brain. Using electrocorticography (ECoG) data from participants listening to a 30-minute narrative, we show that deeper LLM layers correspond to later brain activity, particularly in Broca's area and other language-related regions. We extract contextual embeddings from GPT-2 XL and Llama-2 and use linear models to predict neural responses across time. Our results reveal a strong correlation between model depth and the brain's temporal receptive window during comprehension. We also compare LLM-based predictions with symbolic approaches, highlighting the advantages of deep learning models in capturing brain dynamics. We release our aligned neural and linguistic dataset as a public benchmark to test competing theories of language processing.

BACKGROUND AND OBJECTIVES/OBJECTIVE:Vascular dysfunction contributes to Alzheimer disease (AD) and related dementias (ADRDs), but the underlying mechanisms remain unclear. Previous studies link midlife hemostasis and platelet aggregation measures to late-life dementia risk. We aimed to determine whether platelet aggregation in midlife is associated with imaging markers of AD pathology. METHODS:F-flortaucipir) PET uptake in dementia-free, middle-aged adults from the Framingham Heart Study. Co-primary outcomes included amyloid and tau uptake in AD-vulnerable regions. We also examined an MRI-based cortical thickness signature of AD risk as a secondary outcome. We used multivariable regression models adjusted for demographic and clinical factors, considering potential nonlinear associations. RESULTS:< 0.035), consistent with a neurodegenerative pattern. DISCUSSION/CONCLUSIONS:Our findings indicate that platelet aggregation is linked to PET and MRI markers of AD pathology as early as midlife. These findings support further investigation of platelet-mediated mechanisms in AD pathogenesis.

This position statement provides updated member guidance from the American Academy of Neurology (AAN) regarding (1) communication with surrogate decision makers about brain death/death by neurologic criteria (BD/DNC), (2) management of surrogate decision-maker objections to BD/DNC, (3) the ethical considerations associated with BD/DNC determination in a pregnant person, and (4) enhancing public trust in BD/DNC. This position statement is intended to complement recommendations in the 2023 "Pediatric and Adult Brain Death/Death by Neurologic Criteria Consensus Guideline" published by the AAN, American Academy of Pediatrics, Child Neurology Society, and Society of Critical Care Medicine, as well as the 2021 AAN Code of Professional Conduct. It replaces the 2019 AAN position statement, "Brain death, the determination of brain death, and member guidance for brain death accommodation requests."

BACKGROUND AND OBJECTIVES/OBJECTIVE:Epilepsy is a common neurologic disorder. Although antiseizure medications (ASMs) are the first-line treatment, identifying the most effective ASM for each individual remains a trial-and-error process. Genetic variation may influence treatment response. We aimed to develop and validate a multimodal deep learning model that integrates clinical and genomic features to predict response to the initial ASM in people with newly diagnosed epilepsy. METHODS:We used data from individuals with newly diagnosed epilepsy in Australia as the development cohort and participants from the Human Epilepsy Project 1 (recruited in the United States, Europe, and Australia) as the external validation cohort. All participants initiated ASM treatment and were followed prospectively for at least 1 year. We included 16 clinical factors and constructed 4 genomic feature types related to epilepsy and ASM pharmacogenomics, with and without functional impact annotations. We evaluated various machine learning architectures and multimodal fusion strategies to predict seizure freedom while taking the initial ASM at 1 year. RESULTS:< 0.05). Applying this model to the development cohort, if all participants took the highest ranked ASMs, the mean predicted seizure-free probability would be 68.05% (95% CI 65.79%-70.35%) compared with the observed seizure-free rate of 47.2% (95% CI 41.3%-53.2%). DISCUSSION/CONCLUSIONS:Integrating genomic data with clinical features enhances the ability of deep learning models in predicting ASM response in newly diagnosed epilepsy. This approach may support personalized treatment selection and improve clinical outcomes.

OnabotulinumtoxinA (onabotA) is approved for the treatment of various therapeutic indications, which require retreatment. In clinical practice, many patients receive onabotA for multiple therapeutic indications concomitantly over extended time periods; however, there is limited long-term utilization and safety data for treating comorbid indications. SYNCHRONIZE, a 2-year, multicenter, retrospective observational chart review study in 10 US clinics, describes onabotA real-world utilization and safety in adults treated for ≥2 therapeutic indications within repeating 3-month periods for up to 7 treatments. This analysis assessed the long-term onabotA safety profile for multiple therapeutic indications by analyzing the incidence of treatment-emergent adverse events (TEAEs). Of 279 patients treated for ≥2 different therapeutic indications across all treatment combination groups in Period1, there was a gradual decrease to 80 patients at the last treatment period. The overall mean onabotA treatments over the study period was 9.3 (range: 2-48). Across treatment periods, most patients had a treatment interval between different indications of ≤24 h (range: 62-98 %) and received ≥200-<400U of cumulative 3-month dosages for multiple indications (range: 43 %-50 %) with a mean 3-month dose from 231.8 to 287.0 U. In total, 28.7 % of patients reported ≥1 TEAE after Period1; this proportion remained broadly constant across treatments (range: 28.3-31.8 %). Overall, the most common TEAEs across treatments were UTIs (range: 0.7-5.7 %), neck pain (range: 3.7-9.1 %), headache (range: 2.9-6.5 %), and migraine (range: 2.5-6.4 %). There was no apparent trend between TEAE incidence and treatment intervals nor cumulative 3-month dose categories for multiple indications. No patients were determined to have lack of effect based on clinical objective measurement. OnabotA showed a safety profile with no new signals in patients treated concomitantly for ≥2 therapeutic indications over repeat treatments up to 2 years. TEAEs across treatment periods were commonly related to the site of injection and were consistent with those previously reported for individual indications.

Clinical decision-making often simplifies continuous risk data into discrete levels using round-number thresholds. These simplifications can distort risk assessments. To systematically uncover these distortions, we develop an interpretable machine learning model that identifies anomalies caused by threshold-based practices. Through simulations, real-world data, and longitudinal studies, we demonstrate how round-number thresholds can lead to discontinuities and counter-causal paradoxes in mortality risk. Despite advances in medicine, these anomalies persist, underscoring the need for dynamic and nuanced risk assessment methods in healthcare. Our findings suggest continuous reassessment of clinical protocols, especially in critical settings like intensive care, to improve patient outcomes by aligning thresholds with the continuous nature of risk.