Abstracts

Rationale: Anterior temporal lobectomy (ATL) is the most common surgical procedure for adults with drug-resistant epilepsy. ATL approaches involve removal of varying portions of essential memory structures- hippocampus and parahippocampus. Resection extent and its effect on seizure and memory outcomes remains a topic of debate. Literature reports no different memory deficits for partial vs. complete resection. The results are most consistent for verbal memory decline in left temporal lobe epilepsy (TLE) patients and less so for right TLE patients. Recent studies have focused on the use of fMRI for predicting post-surgical memory outcomes. In the current study, we aimed to evaluate the effect of partial hippocampal resection on memory encoding and recognition in a young adult male with intractable right TLE. Methods: A 19 yo (EP) with TLE 3 years post-partial right ATL and a healthy 41 yo control (HC) were given the same fMRI paradigm on a GE 3T MRI. Pre-and post cognitive and Wada results were available for EP. We adapted a paradigm described by Szaflarski et al. (2004) that involved scene encoding (indoor and outdoor) alternated with rest, and implemented a delayed recognition task. Behavioral responses during the memory paradigms were recorded and % accuracy was compared between the 2 participants using one-tailed t-test for both encoding and recognition. Active voxels according to threshold (pbonf<0.01) in the hippocampus region were used to calculate laterality index (LI). Results: EP's pre-operative Wada indicated bilateral memory representation. Neuropsychology suggested decrease in visual-spatial learning, verbal memory and stable IQ. MRI revealed volume loss in residual hippocampal body, tail and amygdala. EP's fMRI results showed right sided activation during both encoding and recognition blocks (LI = -0.81 and -0.36 respectively). Behavioral comparisons between EP and HC revealed no difference (p=0.23) during encoding and significant difference (p<0.01) during recognition blocks regarding % accuracy. FMRI image analysis indicated an increase in voxel activation between recognition and encoding blocks- 311% (EP) and 679% (HC) increase. Conclusions: This study is unique in that it is the first to show post-operative memory fMRI activation in the hippocampal remnant. Literature has reported minimal difference in memory outcome following partial vs. complete hippocampal resection suggesting that the hippocampus functions as a single unit for memory. Thus, even partial resection impairs function. Ipsilateral activation during memory tasks following partial resection raises questions about the functional process. Behavioral data showed expected findings; right TLE patients are expected to show decline during recognition tasks yet less so during encoding. Preliminary data suggests that a block design encoding and recognition paradigm can be used to predict post-surgical outcomes. Acknowledgements: Authors express their gratitude to Drs. Szaflarski and Vannest for sharing their paradigm. Second and 3rd authors contributed equally.