Memory in children with epilepsy: Utility of the WRAML-2 in generalized and focal epilepsy syndromes.

The material-specific model for memory impairment predicts that verbal memory deficits are seen with left temporal seizures, and visual memory deficits are seen with right temporal seizures (Henkin et al., 2005). In pediatric epilepsy, seizure pathology has not always yielded the expected material-specific memory profiles. This study used the Wide Range Assessment of Memory and Learning-Second Edition (WRAML-2) to assess memory functioning among pediatric patients with epilepsy. The WRAML-2 was administered to 180 youth with epilepsy during their neuropsychological evaluations. Memory and recognition scores correlated significantly with epilepsy severity variables. There were no significant differences in verbal and visual memory and recognition index scores among patients with generalized epilepsy or among those with lateralized or localized electroencephalography (EEG) patterns and lesions on imaging. However, clinically meaningful verbal versus visual discrepancy scores were significantly related to lateralized abnormalities on EEG and magnetic resonance imaging (MRI) results. Most patients with right hemisphere pathology showed the expected material-specific visual memory deficits, while fewer than 15% of the left hemisphere cases showed the expected verbal memory deficits. Over one-third of those with identified left-sided pathology showed clinically significant deficits in visual memory. Findings are incongruent with the material-specific memory model and reflect the fact that early developmental neurological insults can lead to functional reorganization/crowding effects in children with left hemisphere epilepsy. On exploratory analyses, there were no significant differences in discrepancy scores among participants with left, right, and bilateral languages on Wada and functional MRI (fMRI). However, those with right and bilateral language dominance were more likely to show discrepancies that were incongruent with the material-specific model.