Risk of seizures induced by intracranial research stimulation: analysis of 770 stimulation sessions

Abstract

Objective: Patients with medically refractory epilepsy often undergo intracranial electroencephalography (iEEG) monitoring to identify a seizure focus and determine their candidacy for surgical intervention. This clinically necessary monitoring period provides an increasingly utilized research opportunity to study human neurophysiology, however ethical concerns demand a thorough appreciation of the associated risks. We measured the incidence of research stimulation-associated seizures in a large multi-institutional dataset in order to determine whether brain stimulation was statistically associated with seizure incidence and identify potential risk factors for stimulation-associated seizures.

Approach: 188 subjects undergoing iEEG monitoring across ten institutions participated in 770 research stimulation sessions over 3.5 yr. Seizures within 30 min of a stimulation session were included in our retrospective analysis. We analyzed stimulation parameters, seizure incidence, and typical seizure patterns, to assess the likelihood that recorded seizures were stimulation-induced, rather than events that occurred by chance in epilepsy patients prone to seizing.

Main results: In total, 14 seizures were included in our analysis. All events were single seizures, and no adverse events occurred. The mean amplitude of seizure-associated stimulation did not differ significantly from the mean amplitude delivered in sessions without seizures. In order to determine the likelihood that seizures were stimulation induced, we used three sets of analyses: visual iEEG analysis, statistical frequency, and power analyses. We determined that three of the 14 seizures were likely stimulation-induced, five were possibly stimulation-induced, and six were unlikely stimulation-induced. Overall, we estimate a rate of stimulation-induced seizures between 0.39% and 1.82% of sessions.

Significance: The rarity of stimulation-associated seizures and the fact that none added morbidity or affected the clinical course of any patient are important findings for understanding the feasibility and safety of intracranial stimulation for research purposes.

Conflict of interest statement

Competing interests

Dr Gross serves as a consultant to Medtronic, which is a subcontractor on the MEMES project, and receives compensation for these services. The terms of this arrangement have been reviewed and approved by Emory University in accordance with its conflict of interest policies. Dr Kahana and Dr Rizzuto have started a company, Nia Therapeutics, LLC (‘Nia’), intended to develop and commercialize brain stimulation therapies for memory restoration and has more than 5% equity interest in Nia.

Figures

Figure 1.

Seizure #7 (a) Localization of electrode channels in a 3D brain model; the open circles correspond to seizure #7, the asterisks to seizure #3, and the triangles to seizure #4. (b) iEEG recording from channel 66, the medial-most posterior hippocampal electrode contact, showing a train of stimulation current being delivered. This is followed by clear seizure activity. (c) Seizure activity can be seen most prominently in channel 57, which corresponds to the medial-most anterior hippocampal electrode contacts. (d) Expanded view of the recording from 1500 to 1520 s from panel C.

Figure 2.

Seizure #3 (a) stimulation delivered to entorhinal cortex, shown in gray bars along the top of the plot, followed by clear seizure activity in a neighboring channel, asterisks in figures 1(a)–(c) are expanded views of the seizure activity, which was both clinically and electrographically similar to the patient’s typical seizures.

Figure 3.

Seizure #4 (a), (b) iEEG tracings from the stimulation channels in the posterior hippocampus (triangles in figure 1(a)) show seizure activity clearly starting prior to the onset of stimulation, which is delivered initially at approximately 542 s into the recording (indicated with gray bars along the top of the plot). (c) Overview, and (d) expanded view, captures of seizure activity, electrographically beginning just after 520 s, clearly before the onset of stimulation.