Abstracts

Rationale: The primary aim of this prospective study is to assess the diagnostic yield of 7 tesla (T) MRI in detection of potentially epileptogenic lesions in patients with lesion-negative 3T MRI. The secondary aim is to determine the sensitivity of cortical thickness maps for detection of suspected focal cortical dysplasia (FCD). Epilepsy patients who have brain lesions identified at MRI are more likely to be seizure-free following surgery as compared to patients with lesion-negative MRI1. 7T brain MRI holds promise to improve detection rates of seizure-causing lesions compared to 1.5T and 3T MRI2. However, the diagnostic yield of 7T MRI with cortical thickness mapping in the detection of epileptogenic lesions in patients with negative or equivocal 3T MRI remains a gap in current knowledge. Methods: Inclusion criteria: Diagnosis of focal-onset, drug-resistant epilepsy; age 6-60 years; prior 3T epilepsy protocol brain MRI interpreted as negative or equivocal for potential epileptogenic lesions. Exclusion criteria: Pregnancy, prior craniotomy, and vagal nerve stimulator. 7T Brain MRI: A Siemens (Erlangen, Germany) Magnetom 7T MRI scanner using 24-channel head coil (Nova Medical; Wilmington, MA) acquired: Sag 3D T1 MP2RAGE 0.7 mm3; Cor 2D T2* GRE 1 mm/0.2 mm; Cor 2D T2 TSE 1mm/0.2 mm; Axial SWI 2mm/1mm; Sag 3D T2 FLAIR 0.8 mm3. No IV contrast or sedation given. MRI interpretation: Two neuroradiologists jointly reviewed 7T MR images with access to clinical information and reported a consensus impression of potential seizure-causing lesions. Diagnostic yield was calculated as percentage of subjects with potential seizure-related findings on 7T MRI. Cortical thickness mapping: Cortical thickness maps were derived from 7T MRI T1 MP2RAGE data using AFNI and FreeSurfer software and thresholded to display only areas of cortical thickness greater than 4 mm. Thickness maps were considered positive if the areas of suspected FCD were rendered. Results: Neuroradiologist review identified 8 potentially seizure-related lesions in 5/7 patients (3 children, 4 adults) who had 7T MRI. A one sample test of proportion showed a statistically significant change in the diagnostic yield of 71% using 7T MRI compared to 0% using 3T MRI in these seven patients (p < 0.001). Lesions included: 3 foci consistent with FCD, 1 cluster of nodular heterotopic grey matter, 2 encephaloceles, 1 focus of juxtacortical hemosiderin, and 1 incomplete hippocampal inversion. 3/3 suspected FCD lesions were detected by the cortical thickness maps (100% sensitivity; figures 1 and 2). Conclusions: 7T MRI with cortical thickness mapping may allow for detection of seizure-related brain lesions in focal, drug-resistant epilepsy patients with lesion-negative 3T MRI. Funding: Support from Grant 1R25 EB016671