"Eating" epilepsy revisited- an electro-clinico-radiological study

This study aimed to evaluate the clinical, video electroencephalographic and MRI attributes of patients with eating epilepsy (EE). Consecutive patients who were diagnosed with EE and underwent potential pre-surgical work-up from 2003 to 2012 formed the study cohort. Their electro-clinico-radiological and seizure outcome data were obtained from our prospectively maintained medical records. Out of 7094 patients who underwent evaluation for refractory seizures, 47 patients satisfied the criteria for EE. Twenty-three (48.9%) had exclusive EE; the remainder had a combination of predominantly eating-induced and unprovoked seizures with no differences noted in timing of seizures in relation to meals. Lesional epilepsy was seen in 34% of patients, with posterior cortex (PC; posterior temporo-parieto-occipital) predominance. In MRI negative patients, PC interictal epileptiform discharges were present in 34.4% of patients and multifocal in 20.6% of patients compared to the MRI positive group with 12.5% and 6.5%, respectively (p=0.003). Among 24 patients (51.1%) with co-existent unprovoked seizures, developmental delay and PC ictal onset was more prevalent (p=0.013 and 0.029) as compared to exclusive EE. The seizure frequency and outcome did not significantly differ between patients with or without MRI abnormality. Two patients underwent anterior temporal lobectomy, with persistence of their eating seizures postoperatively. EE is a complex reflex epilepsy of cryptogenic and symptomatic etiology. As opposed to the traditionally implied temporo-limbic mechanisms behind epileptogenesis, a multilobar network originating from the PC receiving sensory and visual inputs linked to the limbic-opercular pathways represents a plausible mechanism. Surgical selection should be diligent and cautious in this group of patients.