Authors :
Presenting Author: Puck Reeders, PhD – Nicklaus Children's Health System
Nathan Schultheiss, PhD – Clinical Neurophysiologist, Nicklaus Children's Health System; Matt Lallas, MD – Pediatric Neurologist, Nicklaus Children's Health System; Shelly Wang, MD MPH – Nicklaus Children's Health System; Prasanna Jayakar, MD, PhD – Director Brain Institute, Brain Institute, Nicklaus Children's Health System; Michael Duchowny, MD – Nicklaus Children's Health System; Marytery Fajardo, MD – Nicklaus Children's Health System
Rationale:
Responsive Neurostimulation (RNS) is an alternative treatment for patients with drug-resistant epilepsy who are not candidates for an excisional procedure. RNS detects abnormal electrical activity at the electrode and delivers targeted electrical stimulation to interrupt seizure propagation. Surgical therapies using RNS typically target the centromedian nucleus of the thalamus (CM) due to its widespread connectivity with neocortical regions. However, the exact mechanism by which neuromodulation of CM aborts seizures is not completely understood.
Hypothesis.
Understanding whether the anatomical network from CM includes epileptogenic zones may provide important insight in how RNS reduces seizure burden. One hypothesis is that direct connectivity of CM with the region of highest amplitude of the scalp activity distribution leads to more favorable outcomes with RNS.
Aims.
We investigated 1) the structural connectivity of the CM where the RNS electrode was placed in pediatric patients with generalized epilepsy, 2) whether this network includes epileptogenic brain regions with predominant peaks in the spatial current density distribution, and 3) whether there is a relationship with the connectivity measures and seizure burden reduction.
Methods:
We performed probabilistic tractography on pre-surgical diffusion weighted imaging using CM of the thalamus as our seed and whole brain cortical and subcortical regions of interest as our targets. We used streamlines from each voxel in the seed region and recorded when those streamlines ended in one of our whole brain target regions. We created a seed-target distribution to investigate the connectivity profile. Next, we identified the target regions with direct connectivity to CM and compared them o the brain areas with predominant peaks in the spatial current density distribution. Lastly, we explored whether connectivity between brain regions with early seizure activity to CM, is related to lowering of seizure burden post RNS implantation.
Results:
Preliminary findings (
N=2) in Patient 1 revealed predominant peaks in the spatial current density distribution (CDR hotspots) in bilateral superior frontal gyri and left medial orbitofrontal cortex. Tractography revealed corresponding strong connectivity between CM and the superior frontal gyri, as well as to the medial orbitofrontal area. This patient had a reduction in seizure duration and intensity. Patient 2 had CDR hotspots in the rostral area of the middle frontal gyrus, and strong connectivity between CM and this region. This patient had a reduction in seizure intensity and improvement in reported behaviors. Further analyses and a larger sample are necessary to verify the initial findings.
Conclusions:
These results support a relationship between anatomical connectivity of CM with brain regions involved in early seizure activity, and seizure burden reduction post RNS. This relationship may aid in candidate selection, and surgical targeting, and outcome prediction.
Funding: Brain Institute - Data Bank Grant