Drug resistant epilepsy is best managed by surgery. The goal of presurgical evalution is to correctly identify the epileptogenic zone, defined by the extent of cortical tissue that needs to be removed is order to achieve seizure freedom. When no causative lesion is identified, careful study of interictal activity is mandatory. Complementary analysis methods exist, designed to identify the source of electrical activity recorded with surface electro-encephalogram. While results are interesting in retrospective studies, the real clinical added value needs to be demonstrated with prospective studies. The company Epilog (Epilog, Ghent, Belgium) offers, with EPILOG PreOp®, a long-term EEG analysis to automatically detect epileptiform activity, combined with an estimation of the electrical source localization using a sLORETA inverse solution model. We will propose the EPILOG PreOp analysis to refractory epileptic patients with non-contributive cerebral imaging, under presurgical evaluation. By comparing the therapeutic decision with and without knowledge of the results of EPILOG PreOp®, we will establish the added clinical value of EPILOG PreOp®.

Saint-Luc University Hospital (CUSL) are the sponsor of the study. The study is bi-centric, in collaboration with the William Lennox Neurological Hospital Center (CHNWL).

Will be included prospectively the first thirty patients with refractory focal epilepsy being evaluated at CUSL or CHNWL between for surgery, and for whom the cerebral MRI 3 Tesla (3T) did not show any obvious causal lesion.

The pre-surgical evaluation includes (standard of care - SOC):

• EEG-video scalp monitoring (SOC)
• MRI 3Tesla (SOC)
• PET scan FDG (SOC)
• neuropsychological assessment (SOC) MRI examination will be reviewed by an experienced neuroradiologist at CUSL (SOC).

The automated EEG analysis and electrical source localization with EPILOG PreOp® (non-SOC) will be carried out by Epilog (Epilog, Ghent, Belgium), blind to the rest of the workup. The costs of this analysis are 250 € / patient and will be covered by a clinical research fund. If the patient is referred by the CUSL, the costs are charged to account 830 E. If the patient is referred by the CHNWL, the costs are charged to a local scientific account.

The results of the pre-surgical evaluation will be discussed in a multidisciplinary meeting, blind to the results of EPILOG PreOp®, in order to determine the presumed localization of the epileptogenic zone as well as the management plan. The result of this discussion will be classified as follows:

• A focal surgical resection is possible;

• B the assessment must be completed by an invasive EEG (SOC);

• C the option of resective surgery is rejected. If an invasive EEG is considered, the anatomical location of electrodes is planned. Then, the results of EPILOG PreOp® are presented (by Dr. Susana Ferrao Santos) and their influence on the management is evaluated. The possible change of care in relation to the initial plan is considered to be relevant if the EPILOG PreOp® analysis results in

  1. a modification of the therapeutic decision as defined above,
  2. a modification of the invasive EEG implantation plan or
  3. the use of an additional non-invasive imaging method to delineate the epileptogenic zone or to establish the functional risk.

Patients will then be followed for 6 months to determine the clinical relevance of any management changes derived from the EPILOG PreOp® analysis. We will also evaluate the relationship between the localization of the EPILOG + regions (ie the brain regions involved in the irritative zone according to the results provided by the EPILOG PreOp® analysis), the surgical resection, the pathology analysis results and post-surgical results.

EPILOG PreOp® analysis will be carried out using the full length EEG recording collected during video-EEG monitoring (SOC) at CUSL or CHNWL, ie 7 days of recording. with 19-27 electrodes placed according to the international 10-20 system. The EEG data will be transmitted after acquisition and anonymization. For patients investigated at CHNWL, a high-density (non-SOC) recording with 64-76 electrodes will be offered for 24 to 48 hours if the investigator deems it necessary. The additional costs associated with high density EEG are borne by the CHNWL. The result will also be sent to Epilog for analysis, after anonymisation of the data.

The EPILOG PreOp® analysis will begin with automated detection of interictal epileptiform activity using the Persyst program (Persyst Spike Detector P13, Persyst, San Diego, CA, USA), and interictal activity will be clustered and localized. source will be estimated using the sLORETA inverse solution model, using the methodology described previously.

Pathology For all patients undergoing resective surgery, histological sections will be reviewed (SOC). Focal cortical dysplasias will be classified according to the ILAE (SOC) classification. Non-contributory results include: gliosis, absence of microscopic or histological abnormality.

Inclusion Criteria:

• drug resistant epilepsy under presurgical evaluation
• non lesional 3Tesla MRI scan
• epileptic seizure(s) recorded during video-EEG monitoring

Exclusion Criteria:

• patient does not agree for follow-up during 6 months
• patient has an obvious causal lesion on MRI scan

Investigator has no access to Epilog Preop result during first case description.