Abstracts

Rationale: Subcortical nuclei play an important role in regulating cortical excitability and seizure propagation. Structural and resting-state functional MRI studies provide evidence for atrophy and deranged functional connectivity of the thalamus in temporal lobe epilepsy (TLE). We hypothesized that thalamic dysfunction is more pronounced in patients with secondarily generalised tonic-clonic seizures (SGTCS) than in those whose seizures do not secondarily generalise. To test this hypothesis, we compared fMRI activation patterns and functional connectivity maps in patients with TLE with and without SGTCS, using a verbal fluency fMRI task known to elicit robust thalamic activation.  Methods: We analysed 105 patients with drug-resistant TLE (49 left/56 right), who underwent a verbal fluency fMRI task. Thirty patients (28.5% - 17 left/13 right) had suffered from at least one SGTCS during the year preceding the investigation. The covert fMRI paradigm consisted of five 30 second blocks requiring the subjects to generate words starting with a visually-presented letter (A/D/E/S/W), alternated with 30 second blocks of cross-hair fixation. In SPM12, convolution of the block onsets vector with the canonical haemodynamic response function modelled task effects. Time series of activation maxima within the left thalamus were extracted and utilised as regressors for an additional seed-based whole-brain functional connectivity analysis. Effects of interest in the thalamus are reported at p < 0.05, FWE-corrected with a 12-mm diameter sphere; whole-brain activations are reported at an exploratory threshold of p < 0.005, k=10. Results: The task elicited the expected activation of fronto-parietal cortices as well as of putamina, thalami and hippocampi/parahippocampal gyri, with predominance of left-sided activation. TLE patients with SGTCS exhibited reduced activation of bilateral anterior and posterior thalami, with left-sided effects more marked than the right, as well as of bilateral hippocampi, parahippocampal gyri, left precuneus, right putamen, right superior frontal and pre-central gyri compared to patients without SGCTS. The latter effects could be identified when contrasting left TLE patients with SGTCS against patients without SGTCS; right TLE patients did not display this effect. Inclusion of language lateralisation index as a covariate of no interest in the model did not change the results. The functional connectivity analysis identified reduced coupling of the left thalamic seed with left anterior cingulate cortex, superior and middle frontal gyri in left TLE patients with SGTCS compared to those without SGTCS, but not in right TLE patients.  Conclusions: Left TLE patients with SGTCS exhibit impaired thalamic activation and reduced thalamo-cortical functional connectivity during a verbal fluency task, consistent with reduced thalamic input to the cortex. This finding suggests that secondary generalization of focal seizures may relate to impaired thalamic outflow, resulting in network dysfunction and increased propensity for more widespread dissemination of ictal discharges. Funding: Wellcome Trust (Programme Grant no 083148), Brain Research Trust (PhD scholarship, LC), Medical Resarch Council, Wolfson Trust, Epilepsy Society, Department of Health's NIHR Biomedical Research Centres funding scheme.