“There is a growing body of research on feasibility and effectiveness of EEG-guided approaches to […] inform anesthetic titration.”

General anesthesia induces dramatic neurophysiological changes that can be observed on the electroencephalogram (EEG). In this month’s issue of Anesthesiology, Martin et al. used high-density EEG in 23 children with ages ranging from 5 to 10 yr to better understand the effects of sevoflurane induction on the developing brain. Unlike pediatric EEG studies that have focused on the maintenance or emergence phases of anesthesia, this study examined the period of induction shortly after loss of consciousness associated with dynamic clinical and EEG changes.

The investigators obtained EEG data during the eyes-closed, awake baseline state and compared these to the EEG during the minute after loss of consciousness from a high inspired concentration of sevoflurane. They report the main finding of significantly increased signal intensity (i.e., power) in the delta (1 to 4 Hz) frequency band after loss of consciousness without topographical differences between age groups. This finding mirrors earlier studies of EEG patterns in infants and young children undergoing sevoflurane-induced general anesthesia with the appearance of initial diffuse EEG slowing driven by high-voltage delta oscillations. With continued sevoflurane administration, the EEG has been further characterized by the appearance of faster alpha (8 to 12 Hz) and beta (13 to 30 Hz) oscillations, with a gradual reduction in slow wave amplitude activity. Martin et al. additionally found the following: (1) the relative amount of delta power was greater compared to adults; (2) there was no difference in EEG power across different frequency bands between subjects who were 5 to 6 yr, 7 to 8 yr, and 9 to 10 yr old; (3) low-voltage patterns (e.g., a burst suppression pattern) were observed in 5 of 23 (21.7%) subjects, consistent with previous reports; and (4) the posterior dominant rhythm—the dominant frequency band in the occipital region when a person is relaxed with eyes closed—was lower in the youngest age group (5 to 6 yr), suggesting that the fixed band definition to characterize spectral power may not be suitable for children. This may also explain why commercial depth-of-hypnosis monitors do not typically perform well in younger children.

A strength of the study by Martin et al. was recording with 64 channels of EEG to detect topographical differences. Among the study sample of children aged 5 to 10 yr, the finding of no difference in topographical delta power between age groups might have been a result of ongoing neurodevelopmental processes already underway.  Had this study been conducted in a younger infant age group, differences might have been observed, as robust increases in brain volume, cortical thickness, brain surface area, and synaptic connections are occurring during that stage in development.  Furthermore, high-concentration sevoflurane could have also caused large-scale changes such that topographical differences in delta power between age groups were not discernable.

One of the EEG signatures conferred by gamma-aminobutyric acid–mediated agents (e.g., sevoflurane and propofol) is the alpha (8 to 12 Hz) frequency oscillation, which appears in a dose-dependent manner and represents coordinated neuronal activity between cortical areas and the thalamus. During the first year of life, the appearance of alpha oscillations with frontal localization occurs at 4 to 6 months of age.  A recent study in infants assessed alpha power during sevoflurane induction, and an association was found between decreased alpha power in posterior regions of the head and subsequent low-voltage patterns.  Low-voltage patterns have been shown to be common during pediatric anesthesia and may represent unnecessarily deep anesthesia. This suggests that beyond chronological age alone, alpha power might have utility as a neurophysiological biomarker in identifying children at higher risk of developing low-voltage patterns.

Martin et al. should be commended for the challenging technical achievement of collecting high-density EEG in children during the awake and anesthetized states. This type of data collection may be less feasible at younger ages, particularly during the dynamic period of sevoflurane induction. Adhering to the adage of “first, do no harm” while engaging in clinical research recruitment in the complex, high-acuity perioperative setting can be a formidable challenge! Overcoming this challenge in research efforts like the study by Martin et al. can narrow important knowledge gaps and potentially improve our understanding of neurodevelopment.

What can clinicians take away from the study by Martin et al.? For the anesthesiologist who routinely performs sevoflurane induction in children, it is reassuring that no age-related differences in EEG features were detected between the younger (5 to 6 yr old) and older (9 to 10 yr old) age groups, suggesting that a dose adjustment for induction at these ages may not be necessary. Furthermore, the prevalence of burst suppression, a sign of an electrically quiet cortex and potentially excess anesthesia, was consistent across the three study age groups and lower than previous reports in younger children (0 to 3 yr old). This suggests that the commonly used approach of administering high-concentration sevoflurane to induce anesthesia in children 5 to 10 yr old does not cause excessive “overdosing” of the brain. The peak alpha oscillations that are lower in frequency in the youngest age group (5 to 6 yr old) compared to the older groups warrant further investigation, and perhaps should be taken into account during future development of depth-of-hypnosis monitors in children.

The discoveries that emerge from this study could have clinical implications informing future management of anesthesia in children. There is a growing body of research on feasibility and effectiveness of EEG-guided approaches to assess brain activity and detect low-voltage patterns to inform anesthetic titration. While Martin et al. have provided a valuable contribution to this active area of perioperative research, much work remains to be done. Future research is needed to elucidate whether neurophysiologically guided management of anesthesia can positively impact clinically relevant patient outcomes.