Delta oscillations phase limit neural activity during sevoflurane anesthesia

Abstract

Understanding anesthetic mechanisms with the goal of producing anesthetic states with limited systemic side effects is a major objective of neuroscience research in anesthesiology. Coherent frontal alpha oscillations have been postulated as a mechanism of sevoflurane general anesthesia. This postulate remains unproven. Therefore, we performed a single-site, randomized, cross-over, high-density electroencephalogram study of sevoflurane and sevoflurane-plus-ketamine general anesthesia in 12 healthy subjects. Data were analyzed with multitaper spectral, global coherence, cross-frequency coupling, and phase-dependent methods. Our results suggest that coherent alpha oscillations are not fundamental for maintaining sevoflurane general anesthesia. Taken together, our results suggest that subanesthetic and general anesthetic sevoflurane brain states emerge from impaired information processing instantiated by a delta-higher frequency phase-amplitude coupling syntax. These results provide fundamental new insights into the neural circuit mechanisms of sevoflurane anesthesia and suggest that anesthetic states may be produced by extracranial perturbations that cause delta-higher frequency phase-amplitude interactions.

Keywords: Neural circuits; Neurophysiology.

Fig. 1

Schematic of the study protocol and data from an illustrative subject. a We acquired data using a randomized, cross-over study design. Anesthetic-drugs were administered for approximately 45 min. b, c End-tidal sevoflurane concentration, behavioral response probability curve, and corresponding frontal electroencephalogram spectrogram during the sevoflurane study visit. d, e End-tidal sevoflurane concentration, behavioral response probability curve, and corresponding frontal electroencephalogram spectrogram during the sevoflurane-plus-ketamine study visit. Shaded regions represent the 95% confidence bounds. Sevo sevoflurane, Ket Ketamine

Fig. 2

Spectral and global coherence analysis of the sevoflurane study visit. a The median frontal spectrograms demonstrate that electroencephalogram oscillations change systematically as a function of the anesthetic state. b Power in the canonical slow-delta and alpha frequency bands change with respect to the anesthetic states. c Frontal spectra and bootstrapped difference of median spectra confirm that electroencephalogram oscillations change systematically as a function of the anesthetic state. Increases in slow oscillation power covaried with anesthetic depth. Alpha oscillation power did not covary with the anesthetic state. d The median global coherograms demonstrate that global coherence also changes systematically as a function of the anesthetic state. Globally coherent beta oscillations were associated with the subanesthetic state, while globally coherent alpha and theta oscillations were associated with the anesthetic states. Globally coherent theta and alpha oscillations did not covary with the anesthetic state. e Global coherence in the canonical theta and alpha frequency bands change with respect to the anesthetic states. f Global coherence spectra and bootstrapped difference of median global coherence confirm that theta and alpha global coherence did not covary with the anesthetic state. Shaded regions represent the 99% confidence bounds of the bootstrapped median power spectra. Black lines represent frequency bands that met our threshold for statistical significance

Fig. 3

Spectral and global coherence analysis of the sevoflurane-plus-ketamine study visit. a Ketamine reduced the power of alpha oscillations during sevoflurane general anesthesia. b Power in the canonical slow-delta and alpha frequency bands change with respect to the anesthetic states. c Frontal spectra and bootstrapped difference of median spectra confirm that the ketamine-induced alpha oscillation power decrease was significant. Ketamine was also associated with a decrease in delta oscillation power and an increase in beta oscillation power. d The median global coherograms demonstrate ketamine reduced the global coherence of theta and alpha oscillations. e Global coherence in the canonical theta and alpha frequency bands change with respect to the anesthetic states. f Global coherence spectra and bootstrapped difference of median global coherence confirm that ketamine significantly reduced theta and alpha global coherence to suggest this dynamic is not fundamental for general anesthesia. Shaded regions represent the 99% confidence bounds of the bootstrapped median global coherence spectra. Black lines represent frequency bands that met our threshold for statistical significance

Fig. 4

Phase-amplitude coupling dynamics associated with sevoflurane-induced anesthetic states. a Frontal comodulograms demonstrated that delta oscillations modulated higher frequencies during sevoflurane-induced anesthetic states. b Frontal phaseampograms between delta and higher frequencies demonstrated that distinct patterns of phase limited neural activity are associated with subanesthetic and anesthetic states. c Frontal circular phasor plots demonstrated that neural activity systematically shifted from π towards 0 phase of delta oscillations as a function of anesthetic depth. The median amplitude vector (red line) was increased from baseline during the anesthetic states. Mean amplitude distribution was not uniformly distributed during sevoflurane-induced anesthetic states. d Frontal comodulograms demonstrated that delta oscillations modulated higher frequencies during the sevoflurane-plus-ketamine induced anesthetic state. e Frontal phaseampograms between delta and higher frequencies demonstrated that the distinct patterns of phase limited neural activity associated with the sevoflurane general anesthetic state were conserved during the sevoflurane-plus-ketamine anesthetic state. f Frontal circular phasor plots also demonstrated that neural activity systematically shifted from π towards 0 phase of delta oscillations as a function of anesthetic depth. The median amplitude vector (red line) was increased from baseline during the anesthetic states. Mean amplitude distribution was not uniformly distributed during sevoflurane-induced anesthetic states. Colored circular dots on phasor plots represent subject level data. Error bars represent standard deviation

Fig. 5

Schematic depicting different cortical origins of low and high-frequency oscillations and the dominant regions of phase-amplitude coupling. During the sevoflurane subanesthetic state, higher frequency activity is limited to the trough of delta oscillations. On the contrary, during sevoflurane general anesthetic states, higher frequency activity is limited to the peak of delta oscillations. Although this dynamic was generated by an anesthetic drug, the present finding suggests that anesthetic states may be produced by extracranial perturbations such as direct current stimulations that cause delta-higher frequency phase−amplitude interactions. Subcortical sources of low-frequency oscillations such as the thalamus is not shown