Abstracts

Rationale: Gamma knife stereotactic radiosurgery (SRS) has been proposed as a minimally invasive alternative to anterior temporal lobectomy (ATL) for mesial temporal lobe epilepsy (MTLE).  In the ROSE Trial, we evaluated the hypothesis that minimally invasive SRS would spare verbal memory as compared to ATL.  Previous prospective pilot data had demonstrated that SRS for MTLE resulted in less verbal memory impairment than published results of ATL. Methods: The ROSE Trial was a multicenter prospective trial for patients with unilateral hippocampal sclerosis and concordant video-EEG results who were randomized to either standardized SRS (24Gy to the amygdala, hippocampal head, and parahippocampal gyrus)  Neuropsychological testing was conducted at baseline prior to randomization and at years 1, 2, and 3 post-treatment.  Patients were predominantly English speaking, but patients with Spanish as a primary language were also enrolled.  Tests included the Wechsler Abbreviated Scale of Intelligence (WASI; English cohort), California Verbal Learning Test (CVLT; English cohort), WHO-UCLA Auditory Verbal Learning Test (AVLT; Spanish cohort), Wechsler Memory Scale-III Logical Memory (WMS-III LM; English cohort), and Beck Depression Inventory (combined cohort).  For each measure of interest, a linear mixed-effects model with repeated measures approach was used to estimate mean changes at the follow up years from baseline for the two treatment groups to compare the group difference in the mean changes.  Change scores across evaluations were calculated in reference to baseline data. Results: Data is presented by English cohort (14 SRS, 6 ATL), Spanish cohort (7 SRS, 5 ATL), and combined cohort (21 SRS, 12 ATL).  At time of surgery, the English cohort was significantly older (M=45.50) than the Spanish cohort (M=26.43).  Data was analyzed using a model that adjusted for baseline performance. The WASI Full Scale IQ score increased slightly in those undergoing dominant SRS, but was not clinically significant.  Long delay free recall of a multi-trial word list learning task in the English cohort declined in both treatment groups, but the SRS group had relative sparing compared to the ATL group.  In the Spanish cohort, there were slight declines in long delay free recall of a word list for both treatment groups, but these changes were not significant relative to baseline.  Delayed recall of verbal prose at 3 years was not significantly different from baseline for either treatment group or between groups.  Of note, both treatment groups within the combined cohort demonstrated improvement in depression symptoms across the study period. Conclusions: In this sample of patients undergoing either SRS or ATL in the dominant hemisphere, minimal differences in verbal memory were observed across treatments, although results varied by language cohort.  Within the English cohort, both treatment groups declined in verbal list learning after treatment but less decline was observed in the SRS treatment group.  No change was observed in verbal prose memory within or across groups.  Within the Spanish cohort, slight declines in verbal list learning were observed in both treatments, but these changes were not clinically or statistically significant.  All groups demonstrated reduced symptoms of depression.  Overall, minimally invasive SRS did not significantly spare verbal memory relative to ATL. Funding: NIH-NINDS R01 NS 058634-01A2