Abstracts

Rationale: Juvenile Myoclonic Epilepsy (JME) is associated with fronto-thalamic structural abnormalities. Executive dysfunction and functional magnetic resonance imaging (fMRI) evidence of motor system co-activation during cognitive tasks are documented for patients with JME and their unaffected siblings. Fewer studies assessed temporal lobe functions and their imaging correlates. A meta-analysis reported moderate long-term memory deficits, and there is evidence for grey matter abnormalities affecting mesiotemporal lobes as well as lateral temporal and high-order fronto-temporo-parietal association cortices. Here, we utilise an event-related fMRI design to specifically address patterns of mesiotemporal activation associated with successful verbal and visual encoding. Methods: Twenty-eight patients with JME, 12 siblings and 18 healthy controls were presented with 70 single concrete nouns ('words') and 70 unfamiliar face photographs ('faces') in a 3T-MRI scanner. Each item was shown for 3s in 30s blocks, separated by 15s cross-hair fixation. Participants were instructed to memorize the items for subsequent out-of-scanner recall, during which the previously presented items were randomly intermixed with an additional 50% novel stimuli. Recognition accuracy was calculated as true positives minus false positives. Data was analysed using SPM8. Mesiotemporal activations associated with subsequently remembered stimuli were compared against those for stimuli that were subsequently forgotten. Significant activations are reported at p < 0.05, FWE-corrected with a 12-mm diameter sphere. Results: Recognition accuracy for the encoding task did not differ across the three groups (ANOVA, p=0.28 and 0.33 for words and faces, respectively). In healthy controls, verbal subsequent memory effects were seen within the left hippocampus. Significant left hippocampal activation was also detected for unaffected siblings, while no significant mesiotemporal activations were identified in JME patients. Direct comparison revealed significantly lower left hippocampal activation in the JME group compared with controls. As for subsequent visual memory, activations were identified in bilateral hippocampi and parahippocampal gyri in controls. Patients with JME activated the right parahippocampal gyrus and the left posterior hippocampus, while no significant mesiotemporal activations were identified in siblings. Direct comparison between siblings and controls revealed significantly lower activations within a cluster encompassing right amygdala, hippocampus and parahippocampal gyrus. Conclusions: Patients with JME and their unaffected siblings exhibit abnormal fMRI activation patterns despite successful task performance, pointing to impaired recruitment of mesiotemporal structures during encoding. The finding of altered mesiotemporal function in unaffected siblings is suggestive of a genetically-driven phenotype. Funding: Wellcome Trust (Project Grant No 079474), Henry Smith Charity (grant ref 20133416). LC is funded by a PhD scholarship from the Brain Research Trust, and received support from an Ermengildo Zegna Founder's Scholarship.