Abstracts

Rationale: A novel infraslow, high amplitude EEG waveform determined to be phasic scalp electrodermal activity (EDA) is described to occur with seizures (s-EDA) and independently of seizures (ns-EDA).

Methods: Epilepsy Monitoring Unit (EMU) recordings were screened for EDA in a previously identified “s-EDA Group”: 8 temporal lobe epilepsy (TLE) patients (5 secondary to autoimmune encephalitis (3 anti-LGI1, 2 seronegative), 2 underwent MEG-EEG and 1 simultaneous scalp-stereotactic depth EEG (sEEG); and a “Control Group” (12 consecutive EMU plus 5 simultaneous scalp-sEEG patients). An additional patient was studied during hypothalamic deep brain stimulation (DBS).

Results: The s-EDA Group had 479 EDA events (mean 3.57/day, range 0-32, 249 s-EDA). 11 Control Group patients had 106 EDA events, mostly during sleep (mean 2.1/day, range 0-14, all ns-EDA). Mean EDA duration was 27.1 (±18.9) and 21.3 (±9.8) seconds, and mean amplitude was 515.0 (±482.2) and 521.8 (±540) uV in s-EDA and Control Groups, respectively. Heart rate increased on average 10.1 bpm with s-EDA but did not change with ns-EDA events.

s-EDA was ipsilateral to seizure onset in 78.9% of seizures, bilateral in 21.1% and never contralateral to seizure onset. s-EDA preceded scalp EEG ictal onset in 72.7% of seizures (mean 18.8 seconds before), and followed scalp EEG ictal onset in 27.3% (mean 20.6 seconds after). EDA always preceded scalp EEG ictal onsets in those with autoimmune associated epilepsy, whereas in non-autoimmune associated epilepsy, EEG ictal onsets preceded EDA in 67.9% of cases.

EDA was not visible in MEG or sEEG recordings, and could be transiently abolished by skin scraping, confirming skin localization. Scalp EDA was not correlated with palmar EDA. Unilateral ns-EDA was observed in one patient upon initiation of high-frequency DBS of the ipsilateral hypothalamus, concurrent with tachycardia and a perceived sensation of warmth.

Both ns-EDA and s-EDA occurred in anatomically discrete, stereotyped regions of the head in all participants, most commonly in the parietal-occipital region, less frequently the temporal-zygomatic and central regions. Detailed drawings depicting the distribution of subjective dermal sensations during pilomotor seizures provided by one s-EDA patient corresponded to these anatomical regions.


Conclusions: Scalp-recorded s-EDA is a rare entity discovered in a subset of TLE patients, and may be more commonly associated with autoimmune epilepsy than other causes of TLE. ns-EDA also exists, primarily during sleep. The morphology of s-EDA and ns-EDA is grossly identical and phasic scalp EDA can be elicited by ipsilateral hypothalamic stimulation, presumably representing activation of hypothalamic autonomic regulatory centers. Scalp EDA events occur in anatomically stereotyped regions of the head, indicating the existence of previously unidentified autonomic dermatomes in the scalp. Scalp EDA can occur without involvement of the extremities, suggesting separate neuroanatomical pathways. Unilateral onset s-EDA is a helpful biomarker as it localizes to the hemisphere of seizure onset with 100% specificity, likely due to ictal activation of the ipsilateral hypothalamus.


Funding: none