Master's Student UCL Queen Square Institute of Neurology London, England, United Kingdom
Rationale: Epilepsy surgery is an effective treatment for temporal lobe epilepsy (TLE); however, memory decline occurs in up to 40% of people. Functional MRI (fMRI) can be used to determine the pre-operative functional anatomy of memory representations to guide surgical decision making and reduce the risk of memory decline. Methods: Twenty-two healthy controls, eight patients with right TLE (RTLE) and ten with left TLE (LTLE), performed an overt word memory-encoding, recognition and recall fMRI paradigm with a baseline of odd-even number decisions. Statistical Parametric Mapping (SPM) was used to investigate specific brain regions associated with encoding and retrieval processes contrasted with the baseline task. One sample and two sample t-Tests were performed between controls and patient groups. Group activations are reported at p< 0.001 for extratemporal and p< 0.01 for mesial temporal lobe (MTL) activations unless otherwise stated. Results: Controls: For word encoding and recognition, there was left frontal and bilateral posterior MTL and right anterior MTL activations.At recall there were only left MTL activations and no extra-temporal activations.LTLE: Encoding activations were seen in the left parahippocampal gyrus, right hippocampus and within the frontal lobes bilaterally. Word recognition was associated with left posterior hippocampus and right anterior hippocampal activations. Recall was associated with right anterior hippocampus and parahippocampal activations. Similar to encoding, there were bilateral frontal lobe activations at recall.RTLE: Encoding activations were seen in the right posterior MTL and left frontal lobe. Recognition was associated with left posterior cingulum activations and no MTL activations. Recall was associated with right parahippocampal gyrus and left anterior hippocampus and frontal activations.At encoding, both LTLE and RTLE groups had significantly less posterior MTL activations than controls. LTLE showed reduced left frontal whilst RTLE showed reduced bifrontal activations. LTLE showed increased right frontal activations compared to controls (p< 0.01). RTLE showed increased right posterior hippocampal activations compared to controls.At recognition, RTLE showed significantly less right anterior and bilateral posterior MTL activations than controls. On recall, LTLE and RTLE had reduced posterior MTL activations but increased right frontal activations in LTLE compared to controls. Differences in recognition activations were not significant between LTLE and controls. Conclusions: A more specific paradigm for examining memory functional anatomy showed that memory encoding and retrieval are dynamic processes that engage varied task-dependent activations. In controls, left frontal lobe activations are involved in verbal memory encoding however retrieval is predominantly a function of the MTL. Both LTLE and RTLE showed significant task dependent functional reorganization within the MTL and in extratemporal regions. This is concordant with findings of non-material specific memory inefficiencies in people with TLE. Funding: Please list any funding that was received in support of this abstract.: National Institute for Health Research University College London Hospitals Biomedical Research Center (grant number 229811)
