Abstracts

Rationale: Early surgical intervention in pediatric refractory epilepsy produces seizure relief in the majority of cases and is thought to deter negative developmental outcomes. There is some evidence for intellectual improvement after epilepsy surgery in children, but little is known about the specific cognitive risks associated with different surgical interventions. To address this insufficiency, we studied seizure relief and cognitive outcomes associated with four types of temporal lobe epilepsy surgery: anterior temporal lobectomy (ATL), temporal lobectomy without resection of hippocampus (TLwo), lesionectomy (LE), and selective amygdalohippocampectomy (A-H). Methods: Ninety children (51 male, 39 female) underwent epilepsy surgery in the temporal lobes between 1992 and 2009 at the Medical College of Georgia. 64 underwent ATL, 14 had temporal lobectomies sparing hippocampus (TLwo), 8 had LE, and 4 had A-H. Mean age at the time of surgery was 12.9 years, age of onset was 6.8 years, and presurgical duration of epilepsy was 6.3 years. Seizure type was complex partial seizures of temporal lobe origin. Cause was cryptogenic in 48 (53%) and symptomatic in 42 (47%) of cases. 75 children underwent comprehensive neuropsychological evaluation prior to surgery and at least one-year after surgery. Pre- and post-surgery tests selected for study included WISC/WAIS verbal-conceptual index (VCI) and perceptual-organization index (POI), Boston Naming (BNT), Story Memory (CMS, WRAML), and Face Memory (CMS). Only a subset of children received all tests both pre- and post-surgery: ATL = 46, TLwo = 8, LE = 4, A-H = 3.Results: Similar to previous research with adults, children who underwent TLwo had significantly worse (p < .02) seizure relief outcomes than children with other surgery types. Percentage of each group with Engel I outcomes at one-year postsurgery were: ATL = 88%, TLwo = 56%, A-H = 100%, LE = 86%. Despite this difference in seizure outcomes, there were no significant postsurgical differences between groups on intelligence, naming, or verbal and nonverbal memory tests. Among left TL cases, all 4 surgery subtypes improved on POI (p = .05) and Face Memory (p < .02) one-year postsurgery. Among right TL cases, there were no significant pre- to post-surgery changes in the ATL group. There were not enough right non-ATL cases to make group comparisons. Conclusions: The hippocampus is often spared in epilepsy surgery due to concerns that resection may impair verbal memory functions. Tentatively (due to the small N of the TLwo group), results of this study suggest there was no cognitive benefit in avoiding hippocampal resection among children undergoing temporal lobectomy, and considerably worse seizure relief outcome. Thus, there appears to be little justification for excluding the hippocampus in children who have preserved presurgical memory function and seizures originating in the mesial temporal lobes.