Hippocampal morphometry in sudden and unexpected death in epilepsy (SUDEP)

ABSTRACT Objective To determine hippocampal morphometric parameters, including granule cell dispersion and features of malrotation, as potential biomarkers for SUDEP from an archival post-mortem series. Methods In a retrospective study of 187 archival post-mortems from three groups, SUDEP (68; 14 with hippocampal sclerosis (HS)), non-SUDEP epilepsy controls (EP-C =66; 25 with HS) and non-epilepsy controls (NEC= 53), Nissl/H&E stained sections from left and right hippocampus from five coronal levels were digitised. Image analysis was carried out for granule cell layer (GCL) thickness and measurements of hippocampal dimensions (HD) for shape [width (HD1), height (HD2)] and medial hippocampal positioning in relation to the parahippocampal gyrus (PHG) length (HD3). A qualitative evaluation of hippocampal malrotational (HMAL) features, dentate gyrus invaginations (DGI) and subicular/CA1 folds (SCF) was also made. Results GCL thickness was increased in HS more than those without (p<0.001). In non-HS cases increased GCL thickness was noted in EP-C compared to NEC (p<0.05) but not between SUDEP and NEC. There was no significant difference in the frequency of DGI, SCF, measurements of hippocampal shape (HD1, HD2 or ratio) or medial positioning between SUDEP, EP-C and NEC groups, when factoring in HS, coronal level and age at death. Comparison between left and right sides within cases showed significantly greater PHG lengths (HD3) on the right side in the SUDEP group only (p=0.018) Conclusions No hippocampal morphometric features were identified in support of either excessive granule cell dispersion or features of HMAL as biomarkers for SUDEP. Asymmetries in PHG measurements in SUDEP warrant further investigation as they may indicate abnormal central autonomic networks.