Abstracts

Rationale: The mechanisms of sudden unexpected death in epilepsy (SUDEP) are not well understood but may involve abnormalities in brain networks regulating respiration and autonomic function. Recent analysis of structural MRI showed increased volume of the right amygdala in patients who subsequently had SUDEP and patients at high risk for SUDEP compared to low risk epilepsy patients (1). We sought to confirm this finding and extend oour analysis to other subcortical structures by comparing confirmed SUDEP cases, living epilepsy patients and healthy volunteers to identify potential neuroimaging biomarkers for SUDEP risk. Methods: We queried a database of an IRB-approved study of structural and functional brain imaging in patients with epilepsy and healthy volunteers at the NYU Comprehensive Epilepsy Center acquired from 2006 to 2016 (237 patients with epilepsy [53% female] and 110 healthy volunteers [52% female]. All scans were performed on the same 3 T MRI scanner using the same T1-weighted MRI sequence optimized for gray–white matter contrast. FreeSurfer v5.1 (http://surfer.nmr.mgh.harvard.edu) was used to segment subcortical and brainstem structures and measure structure volumes  (see (2) for details). We identified 4 epilepsy patients that subsequently died of probable/definite SUDEP who had good-quality scans. Living epilepsy patients were further stratified into low and high SUDEP risk based on the clinical classification scheme proposed by (1). For each SUDEP case, we identified age (±10 years) and sex matched living epilepsy and healthy volunteer controls and examined group level differences in gray-matter volume between SUDEP, living epilepsy patients, healthy volunteers and low vs high risk epilepsy patients using ANOVA and Bonferroni correction. Z scores were calculated for individual SUDEP patients by comparing them to al healthy controls and then to all epilepsy controls. Table 1 shows the demographics of subjects included in analysis. Results: When compared to matched healthy and epilepsy controls, individual SUDEP patients had decreased volumes in the caudate, putamen and nucleus accumbens (p0.05). There was a trend for higher right amygdala volume compared to both living epilepsy patients and healthy controls, but no significant group differences. There was also a trend for increased right amygdala volumes in epilepsy subjects at high risk for SUDEP compared to low risk epilepsy patients but again no significant group differences. There were significant differences between healthy and epilepsy controls groups in the thalamus, caudate, putamen, accumbens and left pallidum (p Conclusions: These results suggest basal ganglia structures may be abnormal in patients who subsequently die from SUDEP and further studies are needed to support earlier reports of increased right amygdala volumes in SUDEP cases. Future studies with larger SUDEP cohorts are needed to determine if these imaging features could serve as a biomarker of SUDEP risk in patients with epilepsy. Funding: This study was supported by FACES (Finding a CURE for Epilepsy/Seizures).