Abstracts

Rationale:
Uncontrolled epilepsy creates a constant source of worriment for patients and puts them at high risk of injury. Some of the challenges faced by patients with epilepsy (PWE) can be overcome by developing algorithms capable of predicting epileptic seizures. While a long list of studies have explored this avenue, additional effort is still required before clinical primetime. Identifying recurrent ‘premonitory’ symptoms of seizures and using them to recalibrate seizure prediction algorithms may improve their performances. In this study, we investigated patients’ ability to predict their seizures based on preictal symptoms.
Method:
We collaborated with Canadian organizations of PWE who invited their members to anonymously complete an online survey through February and March 2020. The survey included questions aiming to collect demographics and clinical characteristics (e.g. seizure frequency, epilepsy duration, and postictal symptoms duration). PWE were also asked to answer questions regarding seizure prediction:  1) 'What symptoms, if any, warn you of an oncoming seizure with impaired awareness or involuntary movements? List all recurrent symptoms.' 2) 'How long before the start of a seizure do these symptoms begin?' 3) 'What percent of the time are these warnings followed by a seizure?' PWE who reported warning symptoms were grouped as having auras if their symptoms are felt within the 5 min preceding their seizures or having prodromes if their symptoms are felt more than 5 min before their seizures. Statistical testing and correlation analyses were performed to assess the significance of reported findings.
Results:
While a total of 221 PWE participated in the survey, 25 were excluded because they did not properly complete the seizure prediction section questions.  Of the 196 respondents, 119 listed at least 1 preictal symptom (60.7%) of which 88 reported auras (44.9%), 24 reported prodromes (12.2%) and 7 did not specify the time. Welch’s t-tests showed that the average patient-reported postictal state (PRPS) was significantly longer for the prodrome group than the aura group and the group with no warning symptoms (p< 0.05). A Pearson’s correlation test showed a positive correlation between the duration of the PRPS and the patient-reported preictal time, r(102)= 0.26, p= 0.0084. Epilepsy duration was positively correlated to the specificity at which patients can predict their seizures, r(103)= 0.35, p= 0.0003.
Conclusion:
This work demonstrates that PWE who have had seizures for longer can better predict their seizures and that a longer PRPS is associated with earlier warning symptoms. Our findings suggest that the postictal period, which is usually correlated with seizure severity, can be an indicator of the preictal time.
Funding:
:Authors are grateful to the Canadian Epilepsy Alliance and Épilepsie section de Québec for the dissemination of this survey. This work was supported by Epilepsy Canada, IVADO, FRQNT, and the Canada Research Chair Program (DKN).