Abstracts

Rationale: There are two ongoing clinical trials for responsive neurostimulation targeting the centromedian (CM) nucleus of the thalamus to treat drug resistant generalized epilepsy: a feasibility trial for Lennox-Gastaut Syndrome (LGS, NCT05339126) and a pivotal study for idiopathic generalized epilepsy (IGE, NCT05147571). Using chronic intracranial EEG data from the CM obtained by the RNS® System, we compare thalamic electrophysiology between these two distinct epilepsy syndromes during both ictal and interictal periods.

Methods: Bilateral CM data were acquired by the RNS System in both clinical trials. Interictal recordings were collected twice daily; ictal recordings were collected when the participant or caregiver swiped a magnet over the device to identify a clinical event or when the device recorded prolonged detections of ictal patterns. Thalamic oscillatory frequency characteristics were quantified during interictal periods. These features were compared across participants and groups. Ictal recordings were evaluated by trained epileptologists for preictal and ictal onset patterns.

Results: Data were analyzed for 10 participants with LGS and 16 participants with IGE (10 with electrographic seizures). Interictal peaks had similar frequency, power, and bandwidth distributions across cohorts, with the exception of delta, which was more common in participants with LGS. Across cohorts, beta was reliably more prominent during the daytime and sigma tended to be more common at night.

All LGS participants had 2 or more ictal onset patterns. Generally, onset patterns were similar bilaterally (median congruence 75%), though 2 participants had no seizures exhibiting the same onset pattern on the right and left. Low voltage fast activity was the most common onset pattern (43% of seizures), followed by bursts of polyspikes (22%), and then rhythmic < 13Hz activity (17%). High amplitude low frequency spiking (5%) and spike and wave (3%) were less commonly seen. Four of 10 participants with LGS had unilateral seizures recorded.

In IGE participants, rhythmic < 13Hz activity was the most common ictal onset pattern (53% of seizures; most common seizure type in 60% of participants) and was seen in 80% of participants with seizures. The next most common seizure onset type was spike and wave (37% of seizures and dominant seizure type in 2 participants). Ten participants had two additional types of onsets (one seizure each of polyspike and high amplitude delta followed by low amplitude fast activity). Two participants had multiple seizure onset types, and rhythmic < 13Hz activity was one of the seizure onset patterns in both participants.

Conclusions: In both epilepsy syndromes, electrographic data consistent with interictal epileptiform and ictal activity was readily apparent in the CM. With the exception of slower background activity in patients with LGS, interictal CM biomarkers were similar between participants with IGE and LGS. However, ictal onset patterns differed between the two groups. In the LGS group there was greater variability in ictal onset within individual participants.

Funding: NINDS UH3NS109557