Abstracts

Rationale: Post-seizure cerebral complexity changes can explain postictal deficits. Prolonged postictal periods, caused by failure of corrective mechanisms, may determine chronic deficits (e.g. cognitive) in drug-resistant epilepsy (DRE) patients. This study quantifies the postictal period using a complexity measure and investigates its relationship with pre- and post-epilepsy surgery cognition scores.


Methods: Studied patients had focal DRE, pre and post-surgical evaluation (including age-appropriate Wechsler Intelligence and Memory Scales) at Timone Hospital (1999-2024), spontaneous stereo-electroencephalography(SEEG)-recorded focal seizures without bilateral tonico-clonic evolution and hemispheric dominance evaluation. We quantified the postictal period, PAT – Postictal Alteration Time, using permutation entropy on the SEEG recordings, as described in (Bratu et al. Ann Clin Transl Neurol. 2024). We used 1-3 seizures per patient (if multiple seizures, an average PAT value was calculated). PAT was computed at the global and hemispheric levels. Cognitive scores were drawn from the last evaluations before SEEG and after epilepsy surgery. Hemispheric language dominance was assessed using personalized methods: fMRI, WADA and SEEG (e.g. event-related gamma activations and Ba-Pa auditory evoked potentials). We considered dominant hemisphere domains: Verbal Comprehension Index (VCI), verbal immediate/delayed recall and non-dominant domains: Perceptual Reasoning Index (PRI), visual immediate/delayed recall. Full-scale IQ, Working Memory Index (WMI) and Processing Speed Index (PSI) were considered global domains. Cognitive post-surgical evolution was determined with the formula: ((Pre-SEEG score - Post-surgery score))/(Pre-SEEG score )*100. Correlations between PAT values and cognitive scores were performed using Spearman's rank correlation coefficient.


Results: We analysed 80 patients (196 seizures): 37 with temporal epileptogenic zone network (EZN), 16 including other lobes (temporal-plus) and 27 with extra-temporal EZN. Post-surgery, 41 patients were Engel I and 39 Engel non-I. 71 patients had left hemispheric dominance, 1 patient right and 8 bilateral. EZN was unilateral in 75 cases (34 left, 41 right) and bilateral in 8. Cognitive follow-up ranged from 1.6 to 16.5 years (5.6±3.17). In left-dominant-left-EZN patients, left PAT negatively correlated with VCI in pre-SEEG (p< 0.001, rho -0.74 in temporal-plus and -0.86 in temporal) and post-surgical evolution (p< 0.05, rho -0.65 in temporal-plus, -0.64 in temporal and -0.41 in all left EZN). The same was true in left-dominant-left-temporal cases for left PAT and immediate verbal recall in pre-SEEG and post-surgical evolution (p< 0.05, rho -0.85 and -0.67). Pre-SEEG immediate visual recall also correlated with right PAT in left-dominant-right-temporal-plus patients (p< 0.05, rho -0.53). Moreover, post-surgical evolution of global scores showed weak negative correlations with global PAT (p< 0.05, rho -0.2 for full-scale IQ and WMI and -0.2 for PSI).