Abstracts

The presurgical assessment of intractable frontal lobe epilepsies (FLEs) remains a challenging issue. The aim of this study is to evaluate the usefulness of non-invasive techniques such as combined conventional MRI (MRI), diffusion-weighted MRI (DWI) and MRSI in FLEs comparing the MR findings with the electrical abnormalities defined by depth recordings.
We studied 7 patients presenting with an intractable FLE and undergoing a presurgical assessment which included a depth recording by stereotactic-EEG (SEEG). All patients first underwent both DWI and MRSI associated with a MRI in a one shot session. DWI abnormalities were defined on a pixel by pixel basis, comparing each patient[rsquo]s brain mean diffusivity (MD) maps with those of 22 control subjects using an ANCOVA (SPM99). At the statistical threshold used (p[lt]0.0005, extent 5 pixels), no significantly different voxels were observed by comparing each individual control to the whole control dataset. MRSI abnormalities were defined comparing the value of each patient[rsquo]s NAA/(Cho+Cr) ratio with those of 10 control subjects on 18 medial and lateral fronto-parietal volumes of interest (localized on 2 MRSI slices). Abnormal values were defined as a ratio inferior to 1 SD. Metabolic maps were also obtained for each patients. The localization of MRI, DWI, MRSI and SEEG abnormalities were then compared.
SEEG demonstrated that the epileptogenic zone (EZ) localization was dorsolateral prefrontal in 2 patients, anterior cingular in 1, orbitotemporal in 1, premotor dorsolateral in 1, supplementary motor area in 1 and post-rolandic in 1. Malformation of cortical development were found on MRI in 4 patients, but corresponded strictly to the EZ in only 2 patients. Three patients were MRI negative. Significant diffusion abnormalities were found in all MRI visible lesions but were also present outside the lesions. These abnormalities were found in at least a part of the EZ in 2 out of the 3 MRI negative patients. One patient had no DWI abnormality. When present, DWI abnormalities were not limited to the EZ only and were sometimes found in electrically normal appearing cortices. Metabolic abnormalities were found in all patients and were located in the MRI lesions and in the EZ of all patients but also extended beyond, in regions involved in the interictal spiking and/or the propagations areas.
For each patient, at least one of the 3 MR techniques depicted alterations in the EZ. Although these MR alterations were not specific, they overlapped, when present, with regions involved in the EZ. These findings demonstrate that the combination of MRI, DWI and MRSI is useful in the presurgical assessment of FLEs in which the EZ definition is still difficult without invasive recording.
[Supported by: INSERM, CNRS, French League Against Epilepsy, ]