(SS#VII) Tailored Stimulation of the Centromedian Nucleus of the Thalamus for Generalized Drug-resistant Epilepsy: Insights from a Probabilistic Tractography Study

Introduction: Neuromodulation of the centromedian nucleus (CM) of the thalamus is a promising treatment for generalized drug-resistant epilepsy. Recent studies conducted on Lennox-Gastaut syndrome (LGS) patients suggested a more clinically beneficial “sweet-spot”, corresponding to CM anterolateral component. However, it is unclear if “sweet-spot” specific structural connections may contribute to the beneficial effect of the stimulation. The objective of this study was to investigate the differences in structural connectivity between the CM with the sweet-spot and less clinically beneficial “cold-spot”.

Methods: Diffusion data of 100 subjects from the Human Brain Connectome project were included to the study. FSL probabilistic tractography was performed to investigate the structural connectivity of CM, sweet-spot and cold-spot to 41 cortical and subcortical areas (sensorimotor, associative, basal ganglia, infratentorial and limbic structures). Using the number of normalized streamlines that reached each target, the probability of each seed’s connectivity to the target was calculated and then the mean probability across all subjects was retrieved. Finally, to evaluate the difference between the connectivity of the seed areas the one-way ANOVA was performed and to identify the significant differences between the targets the Bonferroni was used as a posthoc analysis.

Results: One-way ANOVA and posthoc analysis indicated that CM presented statistically significant stronger connectivity in brainstem (F(2,297)=15.7,p < 0.001), basal ganglia (F(2,297)=9,p < 0.001), motor (F(2,297)=63.2,p < 0.001) and sensory (F(2,297)=16.11,p < 0.001) cortices compared to sweet- and cold-spots. Additionally, compared to CM and the cold-spot, the sweet-spot showed statistically significant stronger connections with prefrontal cortices (F(2,297)=15.32,p < 0.001) and the cerebellum (F(2,297)=21.9,p < 0.001).

Conclusion : The “sweet-spot” connectivity pattern explains the beneficial effect of CM stimulation on LGS, a chiefly cortical-driven pathology. However, stimulation of other prominent CM connections (brainstem and basal ganglia) may be more efficacious for different epilepsies, such as absence seizures. Our results suggest a potential role of probabilistic tractography in tailoring CM stimulation.