Epileptiform discharges triggered with direct electrical stimulation for treatment-resistant depression: Factors that modulate risk and treatment considerations (2024)

Associated Data

Supplementary Materials

Letter to the Editor:

Direct electrical brain stimulation (DES) has been investigated for diverse neuropsychiatric indications, including mood disorders, obsessive compulsive disorder, and chronic pain [1]. DES modulates neural activity of dysfunctional networks to achieve therapeutic effects [2]. Although chronic DES can reduce seizure frequency in people with epilepsy [2,3], its effects—both intended and unintended—are not well established in people without epilepsy. Potential adverse effects of chronic DES include lowering the threshold for and eliciting epileptiform activity [4–6], e.g. afterdischarges, spontaneous discharges, or seizures. Compared to open-loop DES, newer closed-loop DES systems incorporate intracranial electroencephalography (iEEG) recordings, which enable monitoring for epileptiform activity. We report a case of epileptiform activity triggered by chronic closed-loop neurostimulation in a patient enrolled in a clinical trial for treatment-resistant major depressive disorder (trMDD). Our early experience with closed-loop DES provides unique insights into factors related to the emergence of epileptiform activity and the potential safety advantages of DES with concurrent iEEG for non-epilepsy conditions.

A 29-year-old woman with trMDD and no known history of epilepsy enrolled in an ongoing multi-stage clinical trial for personalized closed-loop DES. She had failed a variety of oral agents, intensive psychotherapy, ketamine infusions, and transcranial magnetic stimulation. Multiple rounds of electroconvulsive therapy (ECT; 60 total treatments) previously provided a modest, transient benefit. She thus trialed personalized closed-loop DES, which continuously monitors neural activity intracranially and stimulates in response to detections of neural patterns that represent biomarkers of mood symptom state.

Stage 1 of the clinical trial involved iEEG recording with subacute electrodes, inpatient stimulation-response mapping, and iEEG biomarker discovery. Ten depth electrodes were implanted with distal contacts in bilateral orbitofrontal cortex (OFC), subgenual cingulate (SGC), hippocampus, amygdala, and ventral capsule. Prior to initiating stimulation, rare epileptiform discharges were observed from the left hippocampus. A stimulation safety survey was first performed using low-frequency stimulation (0.5 Hz) at increasing amplitudes (1, 3, 6 mA, 100 μs, 30 s; charge densities of 2.0, 6.0, 11.9 μC/cm², respectively). Epileptiform afterdischarges were observed when stimulating the left hippocampus, left amygdala, and left ventral capsule at 6mA. Additional safety testing was performed with high-frequency stimulation (100 Hz) using short durations (1 and 3 s) and increasing amplitudes (1, 3, 6 mA, 100 μs). Afterdischarges were observed at 6 mA within the right OFC; mesial temporal structures were excluded from high-frequency testing given frequent afterdischarges during low-frequency stimulation. Stimulation-mapping was subsequently performed at amplitudes and regions determined safe, i.e. below afterdischarge threshold (ADT). Notably, ADT decreased after multiple days of testing [7], requiring further reduction of stimulation amplitudes.

Stage 2 involved implantation of chronic electrodes in two regions where stimulation-mapping revealed mood improvement. Chronic electrodes were implanted targeting the right SGC and left nucleus accumbens and SGC (Fig. 1A) using the Responsive Neurostimulation (RNS^®) System (NeuroPace, Inc., Mountain View, CA). Prior to enabling stimulation, iEEG recordings exhibited occasional spontaneous epileptiform discharges in both leads. Chronic stimulation was enabled two months after implantation. Afterdischarges were first observed when stimulation was sequentially increased from 3 to 6.5 mA over the right SGC (bipolar stimulation pathway, 100Hz, 120 μs, 6s; 4.5–9.9 μC/cm²). Stimulation was disabled for 17 days and subsequently reenabled in the left SGC, followed again by the right SGC, using lower amplitudes and a monopolar stimulation pathway to reduce charge density. A marked increase in spontaneous epileptiform discharges was observed with delivering increasing rates of stimulation, i.e. from 50 to 200+ stimulations per hour during testing days and 50+ stimulations per hour during chronic stimulation (monopolar configuration, 142.9 Hz, 160 μs, 4.3 mA, 6s; 8.7 μC/cm²; Fig. 1B). The increase in spontaneous epileptiform discharges correlated with an increase in the iEEG spectral power within the beta–low-gamma band (15–50 Hz; see Supplementary Methods; Fig. 1C,D-4). Stimulation was subsequently disabled for 2.5 months, leading to a progressive reduction in both beta–low-gamma power and rate of epileptiform discharges (Fig. 1C,D-5). During this time, the patient was also started on anti-seizure medications, lacosamide and clobazam. Stimulation was safely reenabled using reduced stimulation amplitude and rates (Fig. 1C,D-6); re-titration of stimulation revealed a lowered ADT as compared to baseline thresholds.

The goal of therapeutic DES is to effect network changes that reduce disease-related symptoms [2,3]; however, this dynamic process may also lead to epileptiform activity and, potentially, epileptogenesis, particularly with cortical stimulation [4,8]. Our case reveals a need to better understand risk factors for epileptogenesis in people without epilepsy undergoing chronic DES. In addition, our case illustrates beta–low-gamma power as a potential biomarker for increased excitability, correlating with increased epileptiform activity. Although this patient never experienced a spontaneous seizure, she had spontaneous epileptiform activity prior to enabling stimulation, which may relate to effects of electrode implantation, an undetected focus of hyperexcitability, or her history of ECT. While ECT has not been associated with an increased incidence of epilepsy [9], its impact on local neurophysiology remains understudied given the uncommon scenario of implanting intracranial electrodes in those with a history of ECT.

We hypothesize that the accumulation of stimulation over weeks with increasing amplitude and stimulation rates led to an increase in epileptiform discharges and reduction of ADT. Decreasing these parameters allowed stimulation to be re-enabled and maintained below the ADT. The progressive increase in beta–low-gamma power and spike rate with stimulation and subsequent reduction upon discontinuation of stimulation over weeks illustrates the dynamic network effects of stimulation. Other stimulation parameters that may affect network excitability include stimulation frequency and location (e.g. white vs. gray matter; mesial vs. neocortical targets) [10]. In closed-loop paradigms, network effects of stimulation may also depend on the biomarker being detected. For example, a low-frequency biomarker might result in stimulation preferentially during periods of drowsiness or sleep, when the propensity for epileptiform activity is elevated [11].

Currently, most non-epilepsy DES applications entail active stimulation without iEEG monitoring. In our closed-loop paradigm, iEEG recordings revealed increased epileptiform activity with specific stimulation paradigms, which may indicate a heightened risk for seizure and would not have been detected through open-loop stimulation modalities. Our case illustrates the potential utility of monitoring iEEG during DES and the need to understand how chronic stimulation may lower the threshold for epileptiform activity.

Supplementary Material

Acknowledgements

Footnotes

Contributor Information

Joline M. Fan, Department of Neurology, University of California, San Francisco, CA, USA; Weill Institute for Neurosciences, University of California, San Francisco, CA, USA.

Ankit N. Khambhati, Weill Institute for Neurosciences, University of California, San Francisco, CA, USA; Department of Neurosurgery, University of California, San Francisco, CA, USA.

Kristin K. Sellers, Weill Institute for Neurosciences, University of California, San Francisco, CA, USA; Department of Neurosurgery, University of California, San Francisco, CA, USA.

Noah Stapper, Department of Psychiatry, University of California, San Francisco, CA, USA.

Daniela Astudillo Maya, Department of Psychiatry, University of California, San Francisco, CA, USA.

Elysha Kunwar, Department of Psychiatry, University of California, San Francisco, CA, USA.

Catherine Henderson, Department of Psychiatry, University of California, San Francisco, CA, USA.

Leo P. Sugrue, Weill Institute for Neurosciences, University of California, San Francisco, CA, USA; Department of Radiology and Biomedical Imaging, University of California, San Francisco, CA, USA.

Katherine W. Scangos, Weill Institute for Neurosciences, University of California, San Francisco, CA, USA; Department of Psychiatry, University of California, San Francisco, CA, USA.

Edward F. Chang, Weill Institute for Neurosciences, University of California, San Francisco, CA, USA; Department of Neurosurgery, University of California, San Francisco, CA, USA.

Vikram R. Rao, Department of Neurology, University of California, San Francisco, CA, USA; Weill Institute for Neurosciences, University of California, San Francisco, CA, USA.

Andrew D. Krystal, Weill Institute for Neurosciences, University of California, San Francisco, CA, USA; Department of Psychiatry, University of California, San Francisco, CA, USA.

References