Abstracts

Rationale: The medical literature suggests that heart rate changes may precede or occur concurrently with some seizure types. Monitoring heart rate, therefore, may be a useful diagnostic in seizure prediction or detection in some epileptic patients. We analyzed simultaneously recorded EEG and ECG data from a large patient population undergoing observation at an epilepsy monitoring unit to assess the prevalence of ictal tachycardia and the frequency of ictal tachycardia among those patients with multiple recorded ictal events.Methods: 2940 hours of EEG and ECG data from 73 patients (34 male, 39 female; median age 35 yrs.) primarily diagnosed with complex partial seizures, which contained 314 seizures, were analyzed using a custom Matlab program. The average recording time analyzed per patient was 40.3 hours. Matlab performed beat detection on the ECG signal and calculated instantaneous heart rate from the resultant R-R intervals. All seizure onset times were based upon expert review of the concurrent EEG by at least one reading epileptologist. The seizure was classified as ictal tachycardia if the heart rate reached at least 100 bpm during the ictal period, and in comparison to the pre-ictal period, the heart rate increased by either 35 bpm or 55%. The patient data was acquired through a partnership with Optima Neuroscience, and was collected at Allegheny General Hospital (Pittsburgh, PA) and the Medical University of South Carolina (Charleston, SC) under the approval of Western IRB. Seizures, identified via EEG, which did not have usable concurrent ECG due to noise saturation or motion artifact, were excluded from the analysis.Results: Thirty-eight of the seventy-three evaluable patients (52%) analyzed had at least one seizure that demonstrated ictal tachycardia using our working definition. Within the ictal tachycardia patient population, 44% of the total seizures (92 of 209) exhibited heart rate changes which were at least 35 bpm or 55% greater than the pre-ictal period, and at least 100 bpm during the ictus (Group A). The remaining 117 seizures in this subpopulation did not meet the definition for ictal tachycardia, but on average exhibited a 14.7 1.1 bpm and 18.5 1.6% increase over the pre-ictal periods (mean SE) (Group B). Forty-one patients did not have any seizures which met the definition for ictal tachycardia (n = 105 seizures), and in comparison these patients only had nominal changes between the pre-ictal and ictal periods (9.7 1.1 bpm and 14.2 1.8% change; mean SE) (Group C). The heart rate change (bpm) among groups B and C was statistically significant (p < 0.01) at an alpha level of 0.05.Conclusions: Our data indicate that ictal tachycardia was prevalent in the patient population studied, which was primarily comprised of complex partial seizures originating in the temporal lobe. Patients who had at least one seizure that met the definition of ictal tachycardia on average showed mild elevation of the cardiac rate associated with their other seizures. These results suggest that heart rate may serve as a useful diagnostic indicating seizure onset or logging seizure occurrence.