Impact of Isoflurane Anesthesia on Gastrointestinal Myoelectric Recordings: A Comparative Analysis of Awake and Anesthetized States in Ferrets

Abstract

Gastrointestinal (GI) slow wave recordings have a long history of investigation but currently limited use in clinical diagnosis and treatment. GI myoelectric activity could potentially play a role in the treatment of functional and motility‐related GI diseases and obesity by providing biomarkers of GI function or a target for clinical interventions. Significant issues, however, hinder our understanding of these signals, including: (1) use of abdominal surface recordings with low spatio‐temporal resolution; (2) typical use of few signaling features, such as deviations from dominant slow wave frequencies; and, (3) lack of defined relationships to GI functions, for example, muscle contraction and movement of luminal contents. Thus, there is a critical need to determine the value of GI myoelectric signals for clinical application and to assess stability of these signals in different testing conditions. As a first step, we tested the hypothesis that gastric myoelectric activity differs between awake and anesthetized states. Acute surgical preparations with open abdomen GI recordings are potentially an ideal setting to investigate slow wave physiology, but it is unknown how anesthesia affects gastric myoelectric signals. Here, we chronically implanted 5 ferrets with four gastric serosal surface electrodes, from fundus to antrum. Following recovery, we conducted electrophysiological recordings under awake and isoflurane anesthesia (from 1 to 3%) conditions. By comparing 95% confidence intervals for the Welch's overlapped segment averaging power spectral densities (PSD) estimated from myoelectric recordings in awake and anesthetized states, we observed statistically significant reductions in power in the anesthetized state (p<0.01 for each of 18 electrodes across animals). The reductions in area under PSD curves were over 90% for 14 electrodes, and ranged between 41% and 88% for 4 electrodes. Suppression did not occur for one electrode, and another was discarded due to corrupted recordings. Moreover, signal power after exposure to isoflurane anesthesia was concentrated in narrow frequency bands, with marked bimodalities in 17 electrodes, in contrast to a more uniform power distribution in the awake state. The time courses of PSDs were also markedly different in awake and anesthetized states, with gastric activity presenting little variation over time in the latter, and substantial variability in the former, which may be associated with behavioral movement. Indeed, in the awake state, one animal had distinct and contiguous long epochs of movement (e.g., grooming, standing) and rest, which correlated with high and low power respectively (p<0.01 for 3 channels; p<0.05 for 1 channel in the animal). These data reveal potential changes in gastric myoelectric activity under anesthesia, which could impact post‐operative recovery. This study also provides insight into the changes in gastric myoelectric activity during anesthesia and reinforces the need to conduct recordings in the awake state.