Electroencephalogram monitoring during ketamine antidepressant treatment: a pilot study

TL;DR

This study explores how EEG brain activity changes during ketamine treatment for depression, using a low-cost system in a clinic setting.

Contribution

The study demonstrates the feasibility of using low-cost EEG to monitor ketamine effects in real-world clinical settings.

Findings

Ketamine infusions significantly reduced fronto-temporal alpha and theta EEG power in treatment-resistant depression patients.

The largest effects were observed at the TP9 electrode, particularly in the alpha and theta frequency bands.

These findings suggest EEG could be used to inform personalized ketamine treatment in the future.

Abstract

Depression is a major cause of disability world-wide. Up to a third of patients have a treatment-resistant form (TRD), presenting a major challenge. Ketamine has been introduced as a novel rapid-acting antidepressant effective in this population. However, at present, ketamine treatment is not routinely informed by any objective neural markers. Basic research has shown promising electroencephalogram (EEG) changes including a decrease in alpha power. However, clinical translation is lacking.

Assess the feasibility of identifying EEG correlates of ketamine infusions in a routine outpatient setting with a low-cost, easily usable system.

The study was carried out at the Oxford Health Foundation Trust Ketamine Clinic (ethics reference 22/EM/0226). N=18 EEG recordings from N=12 patients were collected (5 women, mean age 44, range 33-62, IV dose 0.5-1mg/kg over 40min). 4-channel EEG was collected with a Muse-S headband at 256Hz, from 5min before to 55min after infusion start. 5s epochs were rejected if gyroscope data indicated head movement above 10 deg/s or if amplitude was above 200μV. A spectrogram (4s window, 3s overlap) as well as band-limited power (theta: 4-8Hz, alpha: 8-13Hz, beta: 13-25Hz) were computed. Significance of changes was found with a repeated measures analysis of variance (RM-ANOVA) on power in 5min segments together with post-hoc Tukey’s P-values.

Across the ketamine infusion recordings, there was a significant effect of time (F=3.65, P=0.0105) and Channel*Time interaction (F=3.80, P<0.001) on the EEG spectrum. Effects were largest on temporal electrodes, particularly TP9 in the alpha and theta bands (Figure 1, Table 1).Table 1:Effect sizes (Cohen’s d) and FDR-corrected ANOVA P-values for ketamine effects on each EEG channel and frequency band. P<0.05 was considered significant (bold). n.s. = not significant (P>0.2).
Channel / Band
TP9AF7AF8TP10Theta1.16 (P=0.019)
0.11 (n.s.)0.11 (n.s.)0.42 (P=0.113)Alpha1.41 (P<0.001)
0.12 (n.s.)0.17 (n.s.)0.605 (P=0.113)Beta1.19 (P=0.112)0.03 (n.s.)0.08 (n.s.)0.21 (n.s.)

Effect sizes (Cohen’s d) and FDR-corrected ANOVA P-values for ketamine effects on each EEG channel and frequency band. P<0.05 was considered significant (bold). n.s. = not significant (P>0.2).

Image:

In a routine outpatient setting, sub-anaesthetic ketamine infusions in TRD patients were associated with decreased fronto-temporal EEG alpha and theta power. Future work should assess the potential of low-cost routine EEG, and alpha desaturation specifically, to inform individualised ketamine treatment.

None Declared

Figures