Abstracts

Rationale: While memory impairments are frequently reported in association with chronic epilepsy, especially when associated with temporal lobe foci, the vast majority of studies have used global memory measures based on normative comparisons. In contrast, relatively few have evaluated specific memory components as they relate to anatomical regions of epileptogenic activity, particularly in children. The present study compared verbal memory performance between epilepsy cases involving pathology in the neocortex, mesiotemporal region, and dual pathology involving both areas. Specific metrics of word list learning and memory were statistically evaluated to determine the best clinical index of hippocampal dysfunction. Methods: 73 subjects 5-22 years of age (mean age=13.05) were given comprehensive neuropsychological testing prior to epilepsy surgery. Participants were 54% male, and 88% right-handed. Histopathologically, 43 were classified as having focal cortical dysplasia (FCD), a developmental aberration of the neocortex, 15 had hippocampal sclerosis (HS), and 15 had dual pathology consisting of both HS and FCD. For all hippocampal involvement, 18 were left lateralized, with 12 on the right. Assessment included the Word List subtest of the Wide Range Assessment of Memory and Learning (WRAML). Scores were evaluated for their ability to predict dysfunction of mesial temporal memory systems based upon known histopathological findings. Results: There were no differences between those with and without hippocampal involvement for delayed memory (p=.23) or total list learning (p=.56). However, mean retention [delayed recall raw score / last learning trial] was statistically lower for those patients with HS than those without (p=.04). In contrast, difference scores [last learning trial-delayed recall score] reflected no contrast (p= .33) between groups. Data for memory recognition were not available for analysis. Conclusions: The data suggest that for word list memory, retention score may be the best indicator of hippocampal involvement. Owing to the complexity of memory processing, networks are vulnerable to disruption at multiple locations, and diminished memory performance does not in and of itself suggest mesial temporal dysfunction. The retention score is a metric of information storage across time, and is less influenced by executive and other processes than are normative scores. Although difference scores are used in some memory tests to compensate for initial memory recall, they may be biased by the magnitude of constituent scores; as noted in this study, difference scores did not correlate well with hippocampal pathology. Statistical power was limited by a small sample, as mesial temporal sclerosis is not frequently seen in young populations. Further study is warranted to replicate findings and determine how lesion location (temporal vs. frontal), lesion laterality, and age of seizure onset affect various aspects of memory performance with use of a retention score.