Abstracts

Rationale: The anterior nucleus (ANT) and centromedian nucleus (CMN) of the thalamus have emerged as promising targets for neuromodulation, in the treatment of drug-resistant focal epilepsy (DRFE). While electrical stimulation of thalamic nuclei has become common for seizure suppression, their electrophysiological ictal characteristics remain largely unknown. We aimed to characterize thalamic ictal patterns to better understand the role of the thalamus in seizure initiation and propagation.

Methods: We analyzed seizures from six patients with DRFE who underwent intracranial monitoring via stereoelectroencephalography (SEEG) for presurgical evaluation. All patients had cortical-thalamic depth electrodes involving the ipsilateral ANT and two patients with frontal lobe epilepsy, had additional ipsilateral CMN electrodes. SEEG tracings were reviewed, and ictal pattern data was summarized using descriptive statistics. Thalamic onset timing was defined as synchronous with seizure onset or delayed.

Results: Intracranial EEG from six patients (83.3% female) was analyzed. Mean age at seizure onset and surgical evaluation was 25.7 years (range, 14-46) and 34.3 years (range, 19-53), respectively.

Fifty-three seizures were recorded, of which 14 (26.4%) were focal aware seizures, 16 (30.2%) focal impaired awareness seizures, 3 (5.7%) focal to bilateral tonic clonic seizures, and 20 (37.7%) electrographic seizures (ESZ). Four patients (66.7%) had mesial temporal onset seizures, and two (33.3%) had frontal onset seizures. At the seizure onset zone (SOZ), the most common ictal pattern was bursts of high voltage spikes (HVSp) (25/53, 47.2%), followed by low voltage spikes (LVSp) (14/53, 26.4%) and repetitive spikes with or without direct current shifts (14/53, 26.4%). All seizures demonstrated electrophysiological ipsilateral thalamic involvement via synchronous LVSp (31/53, 58.5%), synchronous low voltage fast activity (LVFA) (8/53, 15.1%), and delayed low voltage fast activity (dLVFA) (14/53, 26.4%).

One patient with lateral frontal seizures showed synchronous LVSp at CMN without ANT involvement. In comparison, another patient with orbital frontal onset seizures had synchronous LVSp at ANT without CMN involvement. A single patient (21 seizures) had variability in ictal pattern in thalamic propagation, showing bursts of HVSp at SOZ with synchronous LVSp at the thalamus (7/21, 33%) versus LVSp at SOZ with dLVFA at the thalamus (14/21, 67%). All dLVFA tracings were ESZs with a 4-12 second delay from SOZ to thalamic nuclei.


Conclusions: We found that the ipsilateral thalamus is consistently involved in seizures across different types and lobes of onset. Most seizures showed synchronous thalamic patterns, indicating rapid relay through the thalamus, while a proportion of ESZ showed delayed thalamic involvement. The involvement of the CMN or ANT based on the brain region of onset highlights the nuanced role of the thalamus in DRFE. This underscores the need to target specific thalamic nuclei for different patients with DRFE. The utility of thalamic SEEG to stratify and predict ideal patients and targets for network neuromodulation warrants further exploration.


Funding: Neurosurgery research and education foundation