Abstracts

Rationale:
30-40% of patients who are not controlled by anti-seizure medications exhibit physical and behavioral manifestations comparable to epileptic seizures (ES), but there are no EEG abnormalities during the event.  These events are called psychogenic non-epileptic seizures (PNES). The risks, treatments, and prognosis in patients with ES vs. PNES differ substantially. Long-term inpatient video/EEG/ECG recordings of the event is the gold-standard diagnostic test for discriminating between ES and PNES. However, it is burdensome, expensive, at times inconclusive, and not available in many affected populations worldwide. This study explores whether non-invasive assessments of autonomic function, via cardiac ECG recordings, distinguish ES vs. PNES patients/events. We hypothesized that measures of autonomic function, particularly preceding and following seizures, differ in ES vs. PNES events.
Method:
Consecutive adult patients admitted to the University of Rochester Long-Term Epilepsy Monitoring Unit between September 2018 and June 2020 were screened for eligibility. Participants were excluded  if they had comorbidities, were on medications, or had devices (e.g., VNS, pacemaker) that could interfere with autonomic function, and those in whom the final diagnosis could not be documented. The 150 minutes each prior to and following a typical non-motor event were selected for analysis. The ECG analyst was blinded to the diagnosis of ES (n=8) or PNES (n=11), and performed high resolution time/frequency domain heart rate variability analyses (5 minute epochs with 1 minute sliding window.)
Results:
The temporal evolution and fluctuations in autonomic measures differed surrounding non-motor ES vs. PNES events. Heart rate and autonomic measures for PNES events remained stable throughout the baseline (-150 to -20minutes before the event), pre-ictal (-20 to 0 minutes before the event), post-ictal (0 to 20 minutes after the seizure), and post-ictal return to baseline periods (20 to 150 minutes after the event.) In contrast, there was a large transient (0-5 minute post-ictal) and prolonged (5-75 minutes post-ictal) increase in heart rate following ES events. Large fluctuations in autonomics were noticed surrounding ES. There was a large positive spike in total autonomic function (SDNN) in the initial 5 minutes following an ES, compared to PNES. Relative to ES baseline, PNES baseline, and PNES pre-ictal (-20 to 0 minutes before the event), vagal measures (RMSSD, pNN50, and HF) were elevated during the ES pre-ictal period. In addition, while the PNES measures remained stable throughout, following ES events there was a sudden (0-3 minutes) and continued (up to 75 minutes) decrease in vagal function post-ictally.
Conclusion:
Non-invasive cardiac ECG recordings and advanced ECG analysis demonstrate that the temporal evolution of autonomic changes differ surrounding ES vs. PNES events. These preliminary results fuel future studies to determine whether they provide novel outpatient diagnostic tools, particularly for populations with limited access to epilepsy monitoring units.
Funding:
:University of Rochester Research Award