First time demonstration of seizure source localization in Focal Electrically Administered Seizure Therapy (FEAST).
DOI: Pending DOI
Synopsis
FEAST is a modified novel version of ECT, intended to produce focal unilateral (right) prefrontal seizures thereby minimizing cognitive side effects. FEAST deploys new asymmetric electrode geometry, unilateral placement, and unidirectional current to enhance its focality1. Three different FEAST configurations were used to compare and determine what the optimal directionality of applied current and electrode placement for future FEAST treatments should be. Using 64-channel EEG recordings on depression patients undergoing FEAST treatment with different configurations, we demonstrate that FEAST does in fact induce a focal seizure in right PFC before it generalizes. We tracked the spatiotemporal induction and propagation of FEAST induced seizure by RMS (root mean square)/power activity and band specific spectral analysis using pre-defined ROIs. These results lay the groundwork for optimizing FEAST parameters and its configuration, which will help support an FDA application to approve FEAST for widespread clinical use as an effective antidepression therapy in patients affected by severe depressive episodes.
Background
While conventional bilateral ECT (electroconvulsive therapy) is widely used as a neuromodulation therapy for treating depression, it is associated with memory (retrograde amnesia) and long-term cognitive side effects. Seizure initiation in pre-frontal cortex (PFC) is fundamental to ECT efficacy, whereas seizure expression in medial temporal regions (MTL) is widely thought to contribute to amnestic effects. An advanced form of ECT is urgently needed to minimize the cognitive side effects to improve the quality of life of these patients, while keeping the same efficacy of treatment2. FEAST aims to meet these requirements by employing an asymmetric electrode geometry and placement: small (1.25”) anterior round electrode above the right eye (anode) and a larger posterior oblong electrode (2 x 3”) on right supplementary motor cortex (cathode); and applying unidirectional current (anode to cathode) to direct current preferentially through the right prefrontal cortex, thereby limiting the effects to MTL3. Understanding the spatiotemporal dynamics of seizure induction and its consequent spread during FEAST induced seizures is paramount to optimizing its treatment parameters. Thus, we used FEAST in three different configurations: Regular FEAST (current direction from round anode to oblong cathode with placement as described above), Reverse Polarity (RP) (same electrode placement but direction of current reversed: from oblong cathode to round anode) and reverse configuration (RC) (where the electrode placement was reversed- small round electrode over motor area and oblong electrode above the right eye but current direction remained the same).
Methods
22 ECT eligible TRD participants underwent 64-channel EEG recording for the first 3 sessions which included a FEAST titration (determination of seizure threshold), a FEAST treatment session (Regular FEAST) and a FEAST with Reversed Electrodes Configuration (RC) or Reversed Polarity (RP), the last 2 being cross-randomized, and all other conditions being equal. Each subject either received FEAST/RC or FEAST/RP but never all 3 conditions. A 64 channel Synamps 2 EEG system using Neuroscan software (Compumedics Ltd.) was used to acquire the EEG data, as per the 10-20 system. An EEG cap, with a slit over the right motor cortex to allow insertion of FEAST posterior electrode, was used to record the cortical signals. Electrodes were referenced to an electrode immediately posterior to Cz. EEG signals were recorded using DC and a low pass filter of 100 Hz at 20 kHz sampling rate and an impedance of ≤ 5 kΩ. In total, 58 EEG recordings were collected. The induced seizure epoch was isolated, preprocessed and cleaned and the 1st 30 seconds were extracted for further analysis. RMS (root mean square) of all electrodes was calculated in 2 second sliding windows with 50% overlap and plotted over topographical headplots to observe the spatiotemporal dynamics and propagation of the FEAST seizure in all 3 configurations. Next, specific regions of interest (ROIs) were defined based on electrode placements such as R (right) and L (left) prefrontal area, R and L motor area, vertex etc and band specific spectral analysis (0.5-80Hz) was performed to determine R PF v/s R motor comparison between the 3 FEAST configurations.
Results
RMS (root mean square), a mathematical measure used to calculate magnitude/ power of EEG signals is widely used to quantify brain activity, capturing subtle changes over time4. Observing RMS activity of these signals immediately after seizure induction would help trace seizure spread in a spatiotemporal manner. RMS activity for the first 30 seconds of seizure was plotted over topographical headplots for all 3 FEAST configurations to determine how (and if) different electrode placement and current direction affects seizure induction site and its consequent spread. All three configurations showed distinct patterns of seizure induction and spread from each other; particularly both regular FEAST and reverse polarity FEAST (RP) showed hemispheric asymmetry and activation over right frontal-motor areas, indicating that both those configurations may be more optimal for FEAST than reverse configuration (RC). The figure indicates spatiotemporal activity at 5, 10 and 15 seconds after seizure induction in all 3 configurations. Band specific (delta, theta, alpha, beta, low and high gamma) spectral power analysis was done on predefined ROIs of right prefrontal (RPF) v/s R motor regions to determine what could be driving the right lateralized frontal-motor activation observed in the RMS analysis; since all 3 FEAST configurations have unique electrode placement and current direction over these brain regions. A 2 second window with 0.5 step size was used to calculate FFTs over the initial 5 seconds of seizure epoch. For all six frequency bands, R motor power was higher than R PF for regular FEAST, while the opposite was true for RP. This difference was statistically significant for theta (Wilcoxin test,p=0.044), alpha (p=0.031), beta (p=0.0056) and low gamma (p=0.019) for regular FEAST and for theta (p=0.026), alpha (p=0.003) and beta (p=0.006) for RP. RC on the other hand did not show a clear trend across frequency bands and was not statistically significant. This could be an indication that regular FEAST and reverse polarity FEAST have different seizure induction sites, one prefrontal and one motor, showcasing the influence of current directionality.
Discussion
ECT is the gold standard for acute treatment of severe depression. Unfortunately, it is associated with memory and cognitive impairments which limit its current use and raises the need for more focal targeting of prefrontal regions, while sparing the MTL regions to circumvent this glaring issue. FEAST, a novel form of ECT, aims to address the shortcomings of conventional bilateral ECT by deploying new electrode geometry and placement and unidirectional current. FEAST can be a novel form of ECT with superior cognitive outcomes for treatment resistant depression, paving a new path for neuromodulation for treating neuropsychiatric disorders. FEAST is also easily clinically adaptable and safe while being as efficient, if not more, than regular ECT. FEAST, with multiple configurations also allows for flexible electrode placement and direction of unilateral current and thus can also be adapted for personalized care and treatment of patients based on individual adaptability.
Disclosures
SigmaStim provided the MECTA SpectrumQ machine on loan to Dr Ziad Nahas for administering FEAST. Institutional support was provided by the American University of Beirut Medical Center for data acquisition for this study.
References
1. Karameh and Nahas A Blind Module Identification Approach for Predicting Effective Connectivity Within Brain Dynamical Subnetworks Brain Topography 10.1007/s10548-018-0666-3 (2019).
2. Deng et al How electroconvulsive therapy works in the treatment of depression: is it the seizure, the electricity, or both? Neuropsychopharmacol. 49, 150–162 (2024).
3. Nahas et al A Feasibility Study of a New Method for Electrically Producing Seizures in Man: Focal Electrically Administered Seizure Therapy [FEAST] Brain Stimulation (2013)
4. Filho et al Analysis of Electroencephalographic Signals Using the Rot Mean Square (RMS) Fluctuation Function: Applicable Sample Test Mathematical Modelling and Applications (2024)