Electroconvulsive therapy generates a postictal wave of spreading depolarization in mice and humans

Figures

• 

  Fig. 1.. Optical neuroimaging in a mouse model of ECT demonstrates a current threshold where seizure is followed by cortical spreading depolarization. a, Schematic of mouse cranial window with optical access to the dorsal cortex. Within each window, five brass electrodes were attached to the intact skull overlying frontal (#1), temporo-parietal (#2, 3), and occipital (#4, 5) cortices. b, Green-filtered illumination (525/50 nm) was projected through the microscope objective onto the brain; reflected and emitted light from the brain was separated by a 560 nm dichroic mirror into two spatially co-registered channels that were bandpass filtered for isosbestic point hemoglobin reflectance of green illumination (512/25 nm) or jRGECO1a red fluorescence emission (630/92 nm). Electrical stimulation was delivered via pairs of brass electrode pins using an isolated pulse stimulator. See Extended Data Fig. 1 and Supplemental Methods. c, Example video frames (from Extended Data Video 1) of neuronal activity (jRGECO1a relative fluorescence %dF/F) in an individual mouse being stimulated with 5 Hz pulses to electrodes 1 and 4. Stimulation was titrated in successive rounds with increasing current (c’,…

• 

  Fig. 2.. Electrode configuration shapes the spatial topography of seizure and CSD. a, Group average maps of seizure amplitude at each pixel (%dF/F). Each panel represents one of 10 possible pairs of 5 electrodes, averaged across multiple stimulation trials (electrode pair and number of trials in each average below, from n=42 end-titration seizures in n=10 mice). Fluorescence colormap is normalized to each plot’s min and max; absolute values of fluorescence are shown in c and d. Stimulated electrode pairs are depicted as colored dots. Note seizures tended to localize near stimulation electrodes. b, Example individual map of post-ictal CSD trajectory for each electrode configuration, depicting time-to-CSD-peak at each pixel. Note that some CSDs were recruited bilaterally to the entirety of both hemispheres, while others were recruited unilaterally to one hemisphere. Bilateral CSDs typically had some degree of temporal lag between the two hemispheres. Colormaps are normalized to the earliest and latest pixel, with absolute values of overall event durations shown in h. c, Asymmetry index for all detected seizures across all titrations (see Methods) evoked by right (R, n=25), left (L, n=34)…

• 

  Fig. 3.. Pulse current and frequency modulate evoked seizure amplitude and duration, as well as subsequent cortical spreading depolarization. Each line in a and b represents an individual mouse being serially stimulated at a fixed pulse frequency (5, 10, 25, 50, or 100 Hz, yellow-orange-red-black color coding). Mice were first stimulated at 1 mA current, seizure was monitored until return to baseline, and then mice were restimulated in increasing in steps to 2, 5, 10, and 25 mA (rainbow color coding) until the evoked seizure was sufficient to trigger CSD. Individual lines for each mouse titration in a and b thus end at the current step that elicited CSD. The x-axis in a-c corresponds to the current step in the titration, while the y-axes measure three different brain event outcomes. a, Seizure amplitude (y-axis, fluorescence intensity %dF/F) averaged across the whole cortex as a function of current step (x-axis, mA) for each mouse titration (individual lines) at one of five frequencies (yellow-orange-red-black color coding). Data analyzed using mixed effects modeling to account for repeated trials on individual animals (see Methods). n is the number of individual mouse titration trials…

• 

  Fig. 4.. Cerebral hemodynamics during four ECT sessions from a single patient exhibit expected features of post-ictal CSD. a, Diagram of diffuse optical probes (magenta on right at F4, green on left at F3), frontal EEG (grey circles at FP1 and FP2), and right unilateral electrodes (yellow). b, Simplified schematic of non-invasive diffuse optics sensors, using paired red-light sources and detectors to regionally measure both deep (brain) and superficial (skin, scalp) blood flow (diffuse correlation spectroscopy, DCS) from two source-detector separations of 1 cm and 2.5 cm and oxygen saturation (frequency-domain diffuse optical spectroscopy, FD-DOS) from four source-detector separations of 1.5, 2, 2.5, and 3 cm. For clarity, the co-located DCS and FD-DOS probes are separated in the schematic. See Methods and Supplemental Discussion and Methods on diffuse optics. c, Series of four recordings on four different treatment days from the same subject receiving right unilateral ECT. Note, the blood flow data shown are derived from the source-detector pair with largest separation and thus largest depth penetration into the brain (Data at shorter separations are given in Extended Data Fig. 5). The…