Abstracts

Rationale: The unpredictability of seizures constitutes a major burden in epilepsy. Self-prediction studies have primarily utilized self-reported seizure diaries, which can be unreliable indicators of seizure frequency. In this preliminary study we compared patient self-reports of sleep quality, mood, and anxiety with gold-standard seizure diaries consisting of electrographically confirmed seizures.

Methods: One patient implanted with the Medtronic Summit RC+S ™ sensing and stimulating deep brain stimulator participated in this study. Electrodes were implanted into the long axis of the hippocampus and the anterior nucleus of the thalamus (ANT) bilaterally. Therapeutic stimulation was delivered to the ANT and seizures were detected from bipolar sensing electrodes in the anterior hippocampi bilaterally. Several times per week for 12 months, the participant was randomly prompted to complete an ecological momentary assessment (EMA) wherein they answered questions on a 7-item Likert scale. The questions included the 12-item Immediate Mood Scaler (Nahum, 2017) and two questions about sleep quality and amount in the previous 24 hours. EMA scores were then compared with electrographic seizures and self-reported seizures and auras.

Results: The participant’s seizure semiology included abrupt loss of awareness with speech and comprehension difficulties. The participant was typically amnestic to these events. Auras included deja-vu, confusion, and feelings of fear with depersonalization. On average, EMAs completed 6, 12, and 24 hours prior to electrographic seizures were not significantly more likely to include more severe symptoms of depression or anxiety than reports that did not precede electrographic seizures. The same results were observed for EMAs completed 6, 12, and 24 hours preceding self-reported seizures and auras. The participant was significantly more likely to report increased sleep during the 24 hours preceding a self-reported aura (group1: n = 30, group2: n = 79, Wilcoxon Rank Sum, p = 0.001), but not significantly more likely to report increased sleep preceding an electrographic seizure.

Conclusions: Although the inclusion of a single subject and unbalanced samples currently preclude interpretation with a larger, group-based model, these preliminary data are in-keeping with the literature on seizure self-prediction. Self-reported mood and anxiety symptoms are more associated with the pre-ictal state for some patients than others, especially considering that seizures can be amnestic events. Our main finding that only increased self-reported sleep was significantly associated with self-reported events supports further investigation into the role that sleep plays into the subjective experience of epilepsy and seizure diary composition.

Funding: Please list any funding that was received in support of this abstract.: NIH UH3 NS095495.