Neurophysiological correlates of sevoflurane-induced unconsciousness

Abstract

Background: Recent studies of anesthetic-induced unconsciousness in humans have focused predominantly on the intravenous drug propofol and have identified anterior dominance of alpha rhythms and frontal phase-amplitude coupling patterns as neurophysiological markers. However, it is unclear whether the correlates of propofol-induced unconsciousness are generalizable to inhaled anesthetics, which have distinct molecular targets and which are used more commonly in clinical practice.

Methods: The authors recorded 64-channel electroencephalograms in healthy human participants during consciousness, sevoflurane-induced unconsciousness, and recovery (n = 10; n = 7 suitable for analysis). Spectrograms and scalp distributions of low-frequency (1 Hz) and alpha (10 Hz) power were analyzed, and phase-amplitude modulation between these two frequencies was calculated in frontal and parietal regions. Phase lag index was used to assess phase relationships across the cortex.

Results: At concentrations sufficient for unconsciousness, sevoflurane did not result in a consistent anteriorization of alpha power; the relationship between low-frequency phase and alpha amplitude in the frontal cortex did not undergo characteristic transitions. By contrast, there was significant cross-frequency coupling in the parietal region during consciousness that was not observed after loss of consciousness. Furthermore, a reversible disruption of anterior-posterior phase relationships in the alpha bandwidth was identified as a correlate of sevoflurane-induced unconsciousness.

Conclusion: In humans, sevoflurane-induced unconsciousness is not correlated with anteriorization of alpha and related cross-frequency patterns, but rather by a disruption of phase-amplitude coupling in the parietal region and phase-phase relationships across the cortex.

Figure 1. Pharmacological and behavioral profiles

(A) Sevoflurane concentration profile for the experimental session of 6 of the 7 participants (one subject experienced loss of consciousness at 0.4% concentration so was not included in this figure but is included in all other analyses). (B) Behavioral responsiveness profile for participant 5. Loss and recovery of consciousness occurs many times before the participant remains unconscious for at least 10 minutes. Shaded areas highlight the 5-minute alignment epochs used to compare and pool data across participants. E_ON=effect onset, Trans_UN=transition to unconsciousness, Trans_CON=transition to consciousness, LOC=loss of consciousness, ROC=return of consciousness

Figure 2. Group spectral analysis

Scalp electroencephalogram across all channels does not demonstrate an increase in alpha or low-frequency power during sevoflurane-induced unconsciousness. E_ON=effect onset, Trans_UN=transition to unconsciousness, Trans_CON=transition to consciousness

Figure 3. Individual and average topography of alpha power

During sevoflurane-induced unconsciousness, alpha power remains primarily in occipital cortex for 4 participants and undergoes some degree of anteriorization in 3 participants. On average, anteriorization of alpha power is not correlated with sevoflurane-induced unconsciousness. Subject 6 became unconscious at the first concentration of sevoflurane and thus the effective onset period is not applicable (N/A). E_ON=effect onset, Trans_UN=transition to unconsciousness, Trans_CON=transition to consciousness, LOC=loss of consciousness, ROC=return of consciousness

Figure 4. Cross-frequency coupling between low-frequency (0.1 – 1 Hz) phase and alpha (8 – 14 Hz) amplitude

(A) Frontal channels. No significant coupling is observed during conscious epochs or sevoflurane-induced unconsciousness. (B) Modulation index for frontal (purple) and parietal (blue) channels. Values above the dotted line are significant at p < 0.05. (C) Parietal channels. Significant coupling is observed during conscious epochs but not during sevoflurane-induced unconsciousness. E_ON=effect onset, Trans_UN=transition to unconsciousness, Trans_CON=transition to consciousness, LOC=loss of consciousness, ROC=return of consciousness

Figure 5. Phase lag index (PLI) analysis

A PLI value of 1 reflects perfectly consistent phase lead/lag relationship and a value of 0 reflects no consistent phase relationship. (A) Average PLI in the alpha bandwidth across aligned baseline, unconscious and recovery epochs. PLI values significantly decrease during sevoflurane-induced unconsciousness across three brain region combinations. (B) Individual PLI values across baseline, unconsciousness and recovery alignment epochs between frontal and occipital channels. Each individual participant demonstrates a reversible decrease in PLI upon sevoflurane-induced unconsciousness, despite the lack of a consistent shift of alpha power from occipital to frontal cortices. Note that for all empirical data analyses shown in this figure, counterpart surrogate datasets suggested no spurious phase-phase relationships.