Abstracts

Rationale: Generalized convulsive seizures (GCS) are the main risk factor for SUDEP (Hesdorffer et al., 2012). However, the exact mechanism through which they lead to SUDEP remains uncertain. Postictal generalized EEG suppression (PGES) is inconstantly observed in the immediate aftermath of GCS (Lamberts et al., 2013). It has been suggested that PGES might participate to the postictal neurovegetative breakdown associated with SUDEP (Lhatoo et al., 2010; Ryvlin et al., 2013). Individualizing patients with high risk of PGES might therefore be important to better evaluate the risk of SUDEP. Methods: We reviewed the video-EEG recordings of 488 patients included in the REPOMSE study to extract all GCS. REPOMSE study is an ongoing multicentre prospective study, based on the French National Research Network on SUDEP predictors, which includes patients (≥16 years) with drug-resistant partial epilepsy who undergo long-term video-EEG monitoring for presurgical evaluation. Video and EEG recordings of all individualized GCS were reviewed to assess occurrence and duration of PGES and to analyse clinical characteristics of seizures. We then investigated whether factors such as seizure type, total seizure duration, duration of tonic and clonic phases, and seizure occurrence during sleep influenced PGES using multivariate analyses. Results: A total of 99 GCS in 69 patients were included. Sleep stage in the minutes preceding the seizure onset could be determined in 97 of them. Forty-one seizures occurred during sleep, including four during sleep stage 1, 35 during sleep stage 2 and two during sleep stage 3. According to ictal semiology, three types of GCS could be individualized: typical tonic-clonic GCS including bilateral and symmetric tonic posturing (Type 1, n=50), clonic GCS without tonic posturing (Type 2, n= 28), GCS during which unilateral and/or asymmetric tonic posturing was observed (Type 3, n=21). Overall, PGES was observed in 46 GCS (46.5%). The mean duration of PGES was 39.7±23.9 seconds. Occurrence of PGES was primarily influenced by seizure type. PGES occurred in 66% of GCS type 1, 14% of GCS type 2 and 43% of GCS type 3 (p<0.003 by logistic regression). In contrast, duration of PGES did not differ across seizure types. In GCS type 1, risk of occurrence of PGES was significantly increased when the seizure occurred during sleep (Odds ratio (95%CI)=5.33 (1.18-24.01), p=0.029 by logistic regression), without impacting PGES duration, however (p=0.66). Conclusions: PGES were predominantly triggered by GCS presenting bilateral symmetric tonic posturing and occurring during sleep. Bilateral symmetric tonic posturing might indicate involvement of the brainstem structures, which could result in disruption of subcortical-cortical activating pathways, especially during sleep. Although these results might echo with the presumptive mechanisms of SUDEP, long-term follow-up of our cohort will be required to evaluate the relations between PGES risk factors and risk of SUDEP. Sources of funding: French's Ministry of Health (PHRC). Veriano Alexandre was supported by FAPESP foundation.