Abstracts

Rationale: Simultaneous EEG and functional MRI (EEG/fMRI) is a non-invasive technique with which it is possible to detect hemodynamic changes related to interictal epileptic discharges (paroxysmal slow waves, sharp waves or spikes: IED) identified on scalp EEG. It has been suggested as a useful tool to characterize various forms of focal and generalized epilepsy and may help in the pre-surgical evaluation of patients with refractory focal epilepsy. To better investigate how much the EEG/fMRI is useful in clinical practice, we studied the information that the BOLD adds to the definition of the epileptogenic zone in patients with focal epilepsy. Methods: Consecutive patients with focal epilepsy who underwent EEG/fMRI from April 2009 to April 2010 were retrospectively reviewed. All underwent a 120 min recording session (anatomic acquisition and BOLD fMRI data collected in runs of 6 min with the patient resting). EEG was recorded inside the 3T scanner. For each patient IEDs similar to those recorded outside the scanner were marked by a single electroencephalographer in the filtered EEG. IEDs with different shape or distribution were analyzed as separate event types. The epileptogenic zone was classified as localized, lateralized, bifocal, or uncertain on the basis of seizure semeiology, video-EEG telemetry, MRI, and in some patients SEEG. BOLD responses were reviewed by three experts. The BOLD responses were classified as not contributory if clearly suggestive of an artifactual pattern or if the BOLD map provided no new information compared to the scalp EEG, and contributory if at least one of the marked event types for each patient provided additional information to the EEG about the focus. We considered patients having SEEG, two EEG/fMRI sessions or a focal lesion on MRI as having independent validation (IV). Results: Thirty-one patients were included. In 8 the EEG was not active during the acquisition. The remaining 23 had at least 1 type of IED. Thirteen of 23 had IV (7 SEEG, 3 focal lesion, 3 repetition of the test) and in all cases the BOLD results were confirmed. In 5/23 the BOLD response was not contributory (2 cases with generalized EEG and concordant generalized BOLD, without added information, and 3 with artefactual patterns). In 18/23 the BOLD response was contributory : 13/18 had focal EEG and focal BOLD response with IV, 4 had focal EEG and focal BOLD response concordant with the epileptogenic zone without IV, 1 had generalized EEG and focal BOLD response without IV. In all focal cases, the BOLD response allowed a more precise definition of the source of IEDs than could be obtained from scalp EEG. Conclusions: In patients the examination was useless because no IED was recorded. Considering the remaining 23 patients, in only 5 cases BOLD responses did not provide additional information. In all patients with IV the BOLD responses were strongly corroborated. We therefore assume that in patients without IV, BOLD also provided valid results. With this assumption, more specific localization was gained from EEG/fMRI in 18/23 cases (78.3%), when compared to scalp EEG. Supported by CIHR grant MOP-38079.