Laparotomy and burst suppression-inducing sevoflurane induce subtle long-term changes in anxiety and social behavior in late postnatal mice
Article information
Abstract
Background
Despite strong preclinical evidence, clinical studies have reported minimal effects of early anesthesia on neurodevelopment. This discrepancy may be due to limitations of preclinical studies, including the absence of clear criteria for appropriate anesthetic depth, lack of physiological monitoring, and absence of a surgical insult. Therefore, we aimed to evaluate the effects of sevoflurane in a more clinically relevant setting by addressing these known limitations.
Methods
After confirming robust burst suppression (BS) at 2.5% sevoflurane, postnatal day 17 (PND17) mice were assigned to three groups: a Control group, no intervention; BS (-) group, surgery with 1.4% sevoflurane; and a BS (+) group, surgery with 1.4%–2.5% sevoflurane. Total duration of anesthesia was 2 h, and blood pressure was non-invasively measured in a subset of mice. Long-term behavioral changes were evaluated from PND56 to PND62 at a two-day interval.
Results
We found subtle but significant changes in anxiety and sociability only in BS (+) group mice. Change in anxiety level was observed in the light/dark box test, with the number of transitions between chambers significantly lower in the BS (+) group (P = 0.025). Reduced sociability was observed in the three-chamber test, as mice in the BS (+) group did not significantly prefer the chamber containing a stranger mouse (P = 0.065).
Conclusions
Both surgery and excessive anesthesia depth are necessary to induce subtle yet long-term behavioral changes in young mice. Future preclinical studies should reconsider sevoflurane concentration and account for surgical trauma as a significant factor when investigating anesthesia-induced neurotoxicity.
Introduction
For over two decades, preclinical studies have suggested that early anesthesia can induce unanticipated neuronal cell death, changes in synaptic transmission, and long-term behavioral changes [1]. However, despite strong preclinical evidence for anesthesia-induced neurotoxicity in the developing brain, clinical studies have reported inconsistent results [2–6]. While prospective studies strongly suggest that early anesthesia does not affect general intelligence [7–9], recent studies also suggest long-term changes in specific behaviors and brain structure [2–6]. Although this discrepancy may be due to fundamental differences in the central nervous system between species, other factors could also limit the translation of preclinical results [10,11]. For example, many animal studies have examined the effects of prolonged anesthetic exposure (over 4 h), whereas in clinical studies, anesthetic exposures are typically shorter (less than 2 h). In fact, several preclinical studies also show that a clinically relevant duration of anesthesia does not affect long-term behavior in young mice [12,13]. Such studies suggest the need for preclinical studies to accurately replicate clinically relevant conditions when studying neurotoxicity.
The concentration of anesthetics is a well-established factor that influences neurotoxicity in young animals [11]. However, to properly address the dose-dependency of anesthesia-induced neurotoxicity, we must first determine the appropriate anesthetic dose for young animals. Previous studies have accomplished this by measuring the minimum alveolar concentration (MAC), the concentration required to avoid movement after a nociceptive stimulus. However, MAC primarily reflects the action of anesthetics on the spinal cord rather than the brain. In clinical practice, lower concentrations of anesthetics are often administered to achieve amnesia and hypnotic effects, partly due to the concurrent use of other agents, such as neuromuscular blockers and opioids [14]. One objective method for identifying unnecessary depth of anesthesia is to measure the concentration of anesthetics that induces burst suppression (BS), an electroencephalogram (EEG) pattern characterized by high-amplitude explosive activity alternating with equipotential inhibitory activity [15]. Considering that intraoperative BS may be associated with postoperative delirium in elderly patients [16], the effects of BS may also be significant in the developing brain, as 30%–60% of children undergoing surgery are exposed to BS-inducing concentrations of anesthetics [17–19]. However, the concentration of anesthetics capable of inducing BS has not been properly evaluated in developing mice.
An additional factor that distinguishes preclinical from clinical studies is the presence of inflammation and surgical procedures [6]. Unlike clinical studies, most preclinical studies evaluated the effects of anesthesia without surgery. Although this may be beneficial as it allows direct examination of the neurotoxic effects of anesthesia without confounding factors, it does not reflect clinical conditions. Also, since surgery itself may affect neurodevelopment by inducing neuroinflammation [20–22], it is possible that the neurotoxic effects of deep anesthesia may be more significant in the presence of surgery. Thus, while previous preclinical studies have reported that clinically relevant doses of sevoflurane do not induce long-lasting changes in behavior in late postnatal mice [12,13], the same anesthetic procedure may have a synergistic effect when combined with a surgery insult [20].
Therefore, we hypothesized that exposure to BS-inducing doses of sevoflurane in the presence of surgical trauma may affect neurodevelopment in postnatal day 17 (PND17) mice. When first determining the concentration of sevoflurane associated with BS, we found that BS began to appear at 1.6%, a concentration much lower than those used in previous studies. Based on these findings, we investigated whether BS-inducing concentrations of sevoflurane, combined with laparotomy trauma, could result in long-term behavioral changes in PND17 mice.
Materials and Methods
Animals and study design
All experimental procedures were approved by the Committee of Animal Research at Chungnam National University Hospital (Daejeon, South Korea, CNUH-014-A0009). Age-matched (PND 14) male C57BL/6 mouse pups were housed with a fostering dam (Damul Science) in a room maintained at 24°C, with a 12-hour light/dark cycle and provided with food and water ad libitum. The study adheres to the applicable EQUATOR guidelines. Study design and the number of mice used for each experiment are illustrated in Fig. 1. EEG recordings and BS evaluation, n = 6; blood gas analysis after anesthesia to confirm normal respiration, n = 3; blood pressure measurements during the perioperative period, n = 6. Behavioral tests were performed in separate cohorts of mice, n = 82.
Experimental design overview. A total of 97 mice were used. The number of mice used for each experiment: EEG recordings, n = 6; blood gas analysis, n = 3; and blood pressure analysis, n = 6. The remaining 82 mice were used for behavioral experiments. EEG: electroencephalogram, BS: burst suppression.
EEG electrode implantation
PND 16 mice were continuously anesthetized with 2.5% sevoflurane for electrode implantation in a stereotaxic apparatus (Model 940; David KOPF Instruments). Body temperature was maintained using a heating pad set at 38°C (Kent Scientific). After subcutaneous injection of 1% lidocaine (Dai Han Pharm. Co., Ltd.) and skin disinfection, a midline skin incision was made. Three EEG stainless steel screws (0.10” mouse EEG screws with wire leads; Pinnacle Technology) were surgically implanted into the skull: two bilaterally over the frontal lobes (+ 1.8 mm anterior-posterior [AP] and ± 1.0 mm medial-lateral [ML] from bregma) and one over the left cerebellum, the latter serving as a reference or ground screw. The leads connecting the screws were soldered to an eight-pin surface mount (eight-pin connector for mice, 90-degree pin configuration; Pinnacle Technology), and the screws and head mount were fixed by applying dental cement (Vertex Self-Curing; Vertex-Dental B.V.) [23,24]. Mice were placed in a warmed cage for at least 20 min before being returned to their home cage.
EEG recording
EEGs were recorded in a quiet environment one day after implantation (PND17). After a brief exposure to 5% sevoflurane in an anesthesia chamber (induction), mice were connected to a ventilator (MiniVent model 845, Hugo Sachs Elektronik) using a custom-made facemask (respiratory rate, 120 breaths per minute [bpm]; tidal volume, weight [g] x 20 µl; illustrated in Supplementary Fig. 1). After adjusting the initial concentration of sevoflurane to 1.4% (fraction of inspired oxygen [FiO2] 0.4), its concentration was gradually increased to 2.4% in 0.2% steps and then decreased to 1.4%. EEGs were recorded for 5 min at each sevoflurane concentration in the presence of a preamplifier connected to the surface mount that amplifies and filters signals to ensure relatively clean, artifact-free data (Pinnacle Technology). The levels of oxygen (Teledyne MX300-I; Teledyne Analytical Instruments) and sevoflurane (Vamos, Dräger) were measured in real time throughout the procedure. Data were acquired using Sirenia Acquisition 2.2.2 software (Pinnacle Technology).
EEG spectrogram analysis and quantification of BS
EEG spectrograms were analyzed using the multitaper method provided by the Chronux toolbox (version 2.1.2; http://chronux.org/) and supplemented by custom MATLAB scripts (MathWorks, Inc.). The parameters for the multitaper analysis included a window length (T) of 1 s with no overlap, a time-bandwidth product (TW) of 3, and five tapers (K), achieving a spectral resolution of 1 Hz.
Suppression was defined as periods during which the EEG voltage amplitude was < 5 µV for > 0.5 s [25]. These suppressions were manually identified through visual inspection of the raw EEG signals. Subsequently, the EEG recordings were segmented into bursts and suppressions using a voltage-based threshold. The BS ratio (BSR) was determined by calculating the percentage of the total recording time during which the EEG signal exhibited suppression [26].
General anesthesia and laparotomy
Mice were arbitrarily (randomly assigned without using a specific selection process or criteria) divided into three groups at PND17: a Control group, no intervention; BS (-) group, in which mice underwent surgery with 1.4% sevoflurane; and a BS (+) group, in which mice underwent surgery with 1.4% and 2.5% sevoflurane. Total anesthesia exposure duration was 2 h. After a brief exposure to 5% sevoflurane in an anesthesia chamber (induction), mice in the BS (-) and BS (+) groups were mask ventilated with a custom-made face mask (Suppl. Fig. 1) connected to a ventilator (MiniVent model 845) and sevoflurane was maintained at 1.4% (FiO2, 0.4; respiratory rate, 120 bpm; tidal volume, weight [g] × 20 µl). Preoperatively, the abdominal area of each mouse was disinfected and draped, and lidocaine cream (9.6%, Vivozon) was applied topically for 15–20 min. For laparotomy, a 1 cm midline incision was made, and the liver, spleen, stomach, small intestine, and colon were systematically explored with sterile forceps. Muscle and peritoneum were sutured with absorbable sutures (Ethicon; CAT # W9575), and the skin was subsequently closed with silk thread (Ailee; CAT#SK617). The duration of the surgical procedure was about 10 min. Postoperatively, mice in the BS (-) group were anesthetized with 1.4% sevoflurane for an additional 80 min (Fig. 2A). During the same postoperative period, sevoflurane concentration was increased to 2.5% for 60 min in the BS (+) group (Fig. 2B). Mice were placed on a heating pad during the entire perioperative period to maintain a rectal temperature of 37°C. Oxygen (Teledyne-MX300) and sevoflurane (Dräger Vamos) levels were measured in real time during the entire procedure. Mice were placed in a warmed cage for at least 20 min before returning to their home cage after the procedures.
Schematic presentation of anesthesia protocols. PND17 mice were divided into three groups: a Control group, no intervention; (A) BS (-) group, the mice underwent surgery with 1.4% sevoflurane; and (B) a BS (+) group, the mice underwent surgery with 1.4% and 2.5% sevoflurane.
Blood pressure measurements and blood gas analysis
Hemodynamic stability in a subset of mice during sevoflurane exposure was confirmed by continuously measuring blood pressure every 2 min by plethysmography using a CODA tail cuff (Kent Scientific). Blood pressure measurements were facilitated by increasing the temperature of the heating pad to maintain a rectal temperature of 38–40°C. Adequate ventilation in a subset of mice in the BS group was determined by blood gas analysis of trunk blood [12].
Behavioral tests
Behavioral tests were performed in two cohorts of mice from PND56 to PND62 at a two-day interval [12,27]. Mice in the first cohort sequentially underwent the following tests: light-dark box test, open field test, three-chamber test, and fear conditioning test. Mice in the second cohort sequentially underwent the elevated plus maze test, light-dark box test, forced swimming test, and the tail suspension test. All behavioral procedures were performed in sound-attenuating chamber (brightness 200-300 lux unless stated otherwise). All experiments were recorded and subsequently analyzed manually by a blinded observer or automatically using EthoVision XT tracking software (Noldus Information Technology).
Open-field test
Mice were placed in the central area of an open-field chamber (40 × 40 × 40 cm), and their activity was monitored for 1 h. General activity was quantified by assessing the total distance traveled, whereas anxiety levels were based on the time spent within the central zone (20 × 20 cm).
Light-dark box test
The light-dark box consists of two compartments, a light compartment, measuring 20 × 30 × 20 cm and illuminated at 900-1000 lux [28], and a dark compartment, measuring 20 × 13 × 20 cm, connected by an open door. Mice were initially placed in the center of the dark compartment. The total time spent by each mouse in the light compartment and the number of transitions between the light and dark compartments were recorded for 10 min to assess anxiety levels.
Three-chamber test
The experiment consisted of three 10-min sessions. During the first session, the mice were allowed to adapt to the central chamber, measuring 40 × 20 × 22 cm [29]. During the second session, the mice were allowed to adapt to all three chambers of the apparatus. Each of the two-side chambers contained a small, empty, cup-shaped cage. Prior to the third session, a novel object was placed inside the cage of the one-side chamber, measuring 40 × 20 × 22 cm, and a novel stranger mouse, of the same age and sex as the subject mouse, was placed in the other cage located inside the opposite chamber of equal size. During the third session, the subject mouse was allowed to freely explore all three chambers. To control side preferences, the positions of the object and stranger mouse were alternated between tests. The preference index (PI) was calculated using the equation: PI (%) = (Mo − Ob) / (Mo + Ob) × 100, where Mo and Ob represent the times spent in the chambers with the novel mouse and the object, respectively.
Elevated plus maze test
The elevated plus maze consisted of two open arms (20 × 5 cm), two closed arms (20 × 5 × 20 cm), and a center zone elevated 50 cm above the floor. Mice were placed in the center of the maze facing an open arm and allowed to explore the maze for 10 min.
Forced swimming test
Each mouse was individually placed in a transparent glass beaker (DongSung Science), 20 cm in height and 12 cm in diameter. The beaker was filled with fresh water at a temperature of 23°C to 25°C. The mice were allowed to swim freely for 6 min while their activity was recorded. The mice were carefully removed from the water, dried with a towel, and returned to their home cages. To maintain consistency, each cylinder was cleaned and refilled with fresh water between tests.
Tail suspension test
Each mouse was suspended by its tail using adhesive Scotch tape, positioned approximately 1.5 cm from the tip of the tail, within a box measuring 30 × 30 × 30 cm and the head of the mouse 5 cm above the bottom of the box. Immobility time was measured for 6 min.
Fear chamber test
Contextual fear memory was evaluated in a conditioning chamber equipped with a metal grid floor (Coulbourn Instruments) located within a sound attenuating chamber. The conditioned stimulus (CS) consisted of a 3-kHz tone presented at 80 dB for 20 s, followed by an unconditioned stimulus (US), consisting of an electric shock at 1 mA for 1 s. After a five-minute habituation period, the mice were subjected to three rounds of CS/US, with a 60-second interval between rounds. After 24 h, to access contextual fear memory, the mice were reintroduced into the same chamber for 5 min. Freezing behavior, indicative of a fear response, was automatically quantified using Freezeframe software (Coulbourn Instruments).
Statistical analysis
The sample size for behavioral tests in the first cohort of mice was calculated using an online software (https://www.sealedenvelope.com/power/continuous-superiority/) with previous fear chamber test results (mean ± SD = 57.24 ± 17.83) [12]. At least 10 mice per group are required when the experimental difference was set to a 40% change (effect size, Cohen’s d = 1.28), with α = 0.05 (two-tailed) and a power of 80%. Although the first cohort of mice consisted of three groups, the alpha level was left unadjusted due to a primary focus on the comparison between the Control and BS groups. The sample size for behavioral tests in the second cohort of mice was calculated using the results of the light-dark box test in the first cohort of mice (mean ± SD = 40.73 ± 10.86). At least seven mice per group were required when using the same conditions as in the first cohort (effect size, Cohen’s d = 1.50). All continuous variables were assessed for normality using the Shapiro–Wilk test and for homogeneity of variance using Bartlett’s test. For comparisons between two groups, if normality was violated, the Mann–Whitney U test was applied. Otherwise, an independent t-test was used, with Welch’s t-test applied if variance homogeneity was not satisfied. For comparisons among multiple groups, one-way ANOVA was performed when both normality and homogeneity assumptions were met. If normality was violated, the Kruskal–Wallis test was used instead. If variance homogeneity was not satisfied, Welch’s ANOVA was performed. Post hoc analysis was performed using Tukey’s HSD or Dunn’s test for between-group comparisons. Paired t-tests were used to compare two conditions within the same group when the normality assumption was met, or the Wilcoxon signed-rank test was used when normality was not satisfied. All statistical analyses were performed using the statistical software R (R Core Team, version 4.2.0), with P values < 0.05 considered statistically significant. Detailed statistical findings are provided in the Supplementary Material 1.
Results
Sevoflurane concentrations commonly used in preclinical studies induce robust BS in PND17 mice
The sevoflurane concentrations associated with BS were determined by analyzing EEG recordings while sevoflurane was sequentially increased from 1.4% to 2.4% in 0.2% steps and similarly reduced from 2.4% to 1.4% (Fig. 3). We discovered that BS began to occur at relatively low sevoflurane concentrations (1.6%), and that sevoflurane concentrations commonly used to study neurotoxicity in young animals (> 2%) induced robust BS. Based on these results, two sevoflurane concentrations were selected for the present study: 1.4% for anesthesia without BS and 2.5% for inducing robust BS.
Sevoflurane concentrations often used in preclinical studies induce robust BS in PND17 mice. EEGs were recorded during sevoflurane exposure in PND17 mice. (A) Representative EEG recordings and spectrogram analysis at 1.4% and 2.4% sevoflurane. (B) Determination of suppression ratio while gradually increasing sevoflurane concentrations from 1.4% to 2.4% in 0.2% steps and similarly decreasing sevoflurane concentrations from 2.4% to 1.4% (n = 6). Values are presented as means ± SDs.
Blood pressure and ventilation in PND17 mice were relatively well-maintained during sevoflurane exposure
Previous preclinical studies investigating neurotoxicity in young animals have been performed without physiological monitoring. To overcome this limitation, we non-invasively measured blood pressure by using the tail-cuff technique in mice undergoing laparotomy [30]. The surgical procedure lasted 10 min but mice were exposed to sevoflurane for up to 2 h (Fig. 2). We found blood pressure to be stable in mice exposed to 1.4% sevoflurane (BS [-] group, Fig. 4A). However, we also found that blood pressure could not be measured in mice continuously exposed to 2.5% sevoflurane, a dose frequently used to study neurotoxicity. Therefore, we sequentially changed the sevoflurane concentration between 1.4% and 2.5% to obtain blood pressure measurements and maintain hemodynamic stability (BS [+] group, Fig. 2B). Blood pressure remained relatively normal (Fig. 4B), and respiration remained relatively stable (Table 1).
Maintenance of blood pressure during sevoflurane anesthesia and surgery. (A, B) Non-invasive measurement of blood pressure by tail cuff plethysmography in three mice each in the BS (-) and BS (+) groups. Blood pressure was relatively well maintained in both groups. Values are presented as means ± SDs.
Blood Gas Analysis of Trunk Blood in a Subset of Mice in the BS (+) Group after Sevoflurane Anesthesia
Surgery combined with 2.5% sevoflurane induced long-lasting but subtle changes in anxiety and sociability
Although previous studies have reported that a relatively short exposure (≤ 2 h) to sevoflurane does not affect long-term behavior in late postnatal mice [12,13], its effects have not been investigated in combination with surgery. After confirming normal growth following early surgery and sevoflurane exposure (Fig. 5A), we evaluated the effects of sevoflurane using a series of behavioral experiments conducted at PND56 (Fig. 5B). General activity and anxiety levels measured in the open field test did not differ between groups, as the total moved distance and time spent in the center area were comparable (Fig. 5C). However, while there was no significant difference in the time spent in the light chamber, the number of transitions between the light and dark chambers were significantly lower in the BS (+) group in the light-dark box test compared to the control group (Fig. 5D). The light-box test results suggest a subtle change in anxiety in the BS (+) group [31,32]. Sociability, as measured by the three-chamber test, was also reduced in the BS (+) group, as these mice did not significantly prefer the chamber containing a stranger mouse over the chamber with a novel object (Fig. 5E). However, such changes were also subtle, as the PI was comparable to the mice in the Control group (Fig. 5E). Unexpectedly, the PI was significantly increased in the BS (-) group, suggesting enhanced sociability (Fig. 5E). Additionally, no significant differences were observed in freezing behavior during the contextual fear test, indicating that sevoflurane exposure did not impact learning and memory (Fig. 5F). Importantly, reducing the duration of deep anesthesia from 1 h to 30 min prevented these changes (Supplementary Fig. 2), implying that the duration of anesthesia associated with BS is a significant factor for influencing behavioral outcomes.
Surgery combined with deep anesthesia induced subtle but long-term changes in anxiety and sociability. (A) Effects of surgery and anesthesia on weight gain (n = 6 per group). (B) Timeline of behavioral experiments (Control, n = 11; BS [-], n = 11; n = 11; BS [+], n = 12). Created with biorender.com. (C) Open field test results, showing no differences among groups in total moved distance and time spent in the center region. (D) Light-dark box test results, showing that the numbers of transitions between chambers were significantly lower in the BS (+) group than in the Control group (P = 0.025, one-way ANOVA with post hoc Tukey test), whereas time spent in the light chamber did not differ significantly. (E) Results of three-chamber tests, show that, unlike other groups, mice in the BS (+) group did not prefer the chamber with the stranger mouse, suggesting reduced sociability (P = 0.065, paired t-test). However, there were no significant differences in the PI between all the groups. (F) Fear chamber test results, showing no differences in freezing behavior among the groups, suggesting comparable learning and memory. Values are presented as means ± SDs (*P < 0.05, †P < 0.001, n.s.: not significant).
Our behavior test results are consistent with recent clinical studies that reported significant behavioral problems despite negative changes in general intelligence [2,3]. To further evaluate and validate behavioral changes after surgery and deep anesthesia, additional behavioral experiments were performed in a second cohort of mice (Fig. 6A). The elevated plus maze test that evaluates anxiety levels showed no significant differences in the total duration spent or the number of visits to the open arms across groups (Fig. 6B). However, the number of transitions in the light-dark box test continued to be significantly lower in the BS (+) group compared to the control group (Fig. 6C), suggesting a change in anxiety that was not detected in other anxiety-related tests. Additionally, no differences in depressive-like behavior were observed between groups in the tail suspension and forced swim tests (Figs. 6D and E).
Change in anxiety due to surgery and deep anesthesia was evident only in the light-dark box test but not in the elevated plus maze test. Additional behavioral tests were performed in a different cohort of mice. (A) Timeline of behavioral experiments (Control, n = 9; BS (-), n = 8; BS (+), n = 9). Created with biorender.com. (B) Elevated plus maze test results, showing no differences among groups in the time and number of visits to the open arm, suggesting that anxiety levels were comparable. (C) Results of light-dark box tests, showing that the numbers of transitions between chambers were significantly lower in the BS (+) than in the Control group (P = 0.044, one-way ANOVA with post hoc Tukey test), but that there were no differences in time spent in the light chamber. (D, E) Results of (D) tail suspension. (E) Forced swim tests, showing no differences in total freezing time, suggesting that depression behavior did not differ significantly in the three groups. Values are presented as means ± SDs (*P < 0.05, n.s.: not significant).
Discussion
Previous preclinical studies on anesthesia-induced neurotoxicity have several factors that can limit the clinical translation, including the lack of clear criteria for appropriate anesthetic depth, absence of physiological monitoring, and the lack of a surgical insult. To address these limitations, we recorded EEG to determine deep anesthesia as the sevoflurane concentration associated with robust BS. Our study is the first to incorporate EEG monitoring to quantify BS for adjusting anesthetic doses in young mice, a methodological approach that may improve the clinical relevance of preclinical studies. Based on our EEG data, anesthesia protocols were designed to expose mice to two different concentrations of sevoflurane (1.4% and 2.5%) while maintaining relatively stable hemodynamics, and the combined effects of laparotomy and deep anesthesia on long-term behavior were evaluated. Interestingly, unlike previous studies that showed deep anesthesia (sevoflurane 2.5%) alone does not affect long-term behavior [12,13], we found significant changes in anxiety and sociability only when deep anesthesia was accompanied by surgery.
Although previous studies have shown that anesthesia-induced neurotoxicity is dose-dependent [11], such studies focused on total anesthetic duration or anesthetic concentration rather than the depth of anesthesia. Additionally, the appropriate concentration of volatile anesthetics was determined primarily by measuring MAC. However, MAC is based on motor responses mediated primarily at the spinal level [14], and does not directly assess brain activity or depth of anesthesia. To address this limitation, we employed EEG monitoring as a novel method to evaluate the sevoflurane concentration associated with BS, a key marker of excessive anesthetic depth [16]. Our novel approach, by quantifying BS, may enhance our understanding of the relationship between anesthetic depth and neurotoxicity. We found that preclinical studies have administered unnecessarily high concentrations of sevoflurane that induce robust BS in young mice (> 2.0%). Although BS may act to ensure basic cell functions during profound low metabolic states [33], studies in elderly patients suggest that BS may contribute to postoperative delirium [34,35]. While a recent prospective, observational study in pediatric patients reported that BS was not related to emergence delirium [36], the effects of BS on neurodevelopment has not yet been evaluated in pediatric patients. As behavioral changes were only observed when late postnatal mice were exposed to a relatively long duration of BS-inducing sevoflurane (1 h) in combination with surgery, in future studies, both the invasiveness of surgery and the duration of excessive anesthesia may need to be considered when studying the neurodevelopmental impacts of anesthesia.
The lack of physiological monitoring during anesthesia is also a widely recognized limitation of previous studies. However, unlike most previous studies that were performed in neonatal mice (PND7), the present study was conducted at PND17. Although it is difficult to directly compare neurodevelopment, this age has been suggested to be more comparable to that of a human six-month infant [37,38]. PND17 mice are relatively larger in size, enabling blood pressure to be measured non-invasively during sevoflurane exposure using the tail-cuff technique. Although their hemodynamics remained stable, blood pressure could be measured only when the temperature of the heating pad was increased to maintain a rectal temperature of 38–40°C [39]. Since the normal core temperature has been reported as 36.6 [40], behavioral tests were not performed in mice used to measure blood pressure to exclude the possible effects of hyperthermia. However, despite adequate warming, BP measurements could not be obtained with continuous 2.5% sevoflurane exposure. Thus, to maintain relatively stable hemodynamics, continuous exposure to 2.5% sevoflurane was limited to 10 min. While this approach allowed for blood pressure measurements, it is important to note that the number of successful measurements remained reduced. Considering neurotoxicity is often assessed in animals receiving this concentration of sevoflurane, hemodynamic instability should be considered as a significant confounding factor in previous preclinical studies.
Although we found statistically significant changes in anxiety and sociability in mice that received both surgery and deep anesthesia, such changes were subtle. For instance, although the decreased number of transitions between chambers in the light-dark box test suggest changes in anxiety [31,32], there were no differences in the total time spent in the light chamber, the main finding of the light-dark box test. The results of other anxiety-related behavioral tests, including the open field and elevated plus maze tests, also did not differ. This inconsistency indicates that the observed changes in anxiety may be minimal or context-dependent. Similarly, although the three-chamber test results found that mice in the BS (+) group did not prefer the novel stranger chamber, there was no change in the PI between groups. While the significance of our results may be debatable, our findings are consistent with recent clinical studies suggesting changes in specific behaviors, but not in general intelligence [2,3].
The present study has several limitations. First, although our results suggest a synergistic effect between surgery and deep anesthesia regarding neurotoxicity, the mechanisms underlying the synergistic effects remain unclear. It is possible that surgery-induced inflammation exacerbates the E/I imbalance induced by early sevoflurane exposure [41,42]. Alternatively, deep anesthesia may worsen systemic inflammation, thus affecting neurodevelopment [22,43]. Second, the present study was limited to PND17 mice. Unlike most preclinical studies that have been performed in neonatal mice (PND7), we used PND17 mice as this age may represent a developmental stage closer to that of human infants compared to PND7 [37,38]. However, future studies in diverse age groups are needed, as it is difficult to directly compare neurodevelopment between humans and rodents. Third, due to the male preference seen in prospective clinical studies [7–9], we performed all experiments in male mice. However, since the neurotoxic effects of anesthesia have been shown to differ between male and female rodents [12,20,44], further studies involving both sexes are essential to gain a more comprehensive understanding of the effects of surgery and deep anesthesia. Fourth, EEG measurements were obtained with the use of single electrode for both reference and ground that may compromise the quality of the EEG signals. However, as we used preamplifiers (Pinnacle Technology) that amplifies and filters signals at the animal’s head level to reduce artifacts, we believe our experimental design was sufficient for measuring BS although it may not be adequate for detailed EEG analysis. Fifth, despite rigorous efforts to maintain normal physiological levels during the perioperative period, several technical issues arose. For example, tail blood pressure could be measured only when the rectal temperature was maintained at 38–40°C. Another limitation was the lack of precise respiratory control. Because endotracheal intubation was difficult in late postnatal mice, mask ventilation with a custom-made facemask was applied, and spontaneous ventilation was maintained throughout the perioperative period with ventilator support (respiratory rate, 120 bpm; tidal volume, weight [g] x 20 µl). Although measuring the tidal volume delivered to mice was difficult due to leaks during mask ventilation, blood gas analysis in the present study suggests that ventilation was sufficient, even in mice exposed to 2.5% sevoflurane for long periods of time.
In conclusion, the combined effects of surgery and excessive depth of anesthesia are required to induce long-term behavioral changes in young mice. Rather than examining the isolated effects of anesthesia, future preclinical studies should consider the combined effects of surgery and anesthesia when investigating neurotoxicity in young mice. Furthermore, the concentrations of anesthetic agents should be reconsidered based on EEG monitoring, as previous studies have used doses that induce robust BS and hypotension, unnecessarily increasing the risk of neurotoxicity. Our findings suggest that avoiding excessive anesthetic depth through EEG monitoring may be beneficial for reducing possible behavioral changes following early surgery.
Notes
Funding
This work was supported by the National Research Foundation of Korea (NRF-2022R1A2C2006269, NRF-2022R1A2C2002756, RS-2024-00406568) and the Korean Ministry of Health and Welfare (HR22C1734). This research was supported by the National Research Foundation of Korea (NRF-2022R1A2C2006269 [RS-2022-NR070193], NRF-2022R1A2C2002756 [RS-2022-NR70073], HR22C1734 [RS-2022-KH130308]) and the Korean Ministry of Health and Welfare (HR22C1734 [RS-2022-KH130308]).
Conflicts of Interest
Woosuk Chung has been an editor for the Korean Journal of Anesthesiology since 2020. However, he was not involved in any process of review for this article, including peer reviewer selection, evaluation, or decision-making. There were no other potential conflicts of interest relevant to this article.
Data Availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Author Contributions
Tao Zhang (Conceptualization; Data curation; Investigation; Methodology; Validation; Writing – original draft)
Yulim Lee (Investigation; Methodology; Visualization)
Xianshu Ju (Conceptualization; Project administration; Supervision)
Jiho Park (Resources; Software; Validation; Visualization)
Boohwi Hong (Data curation; Formal analysis; Methodology; Software)
Jianchen Cui (Investigation; Resources; Software; Visualization)
Yeonsu Kim (Formal analysis; Software)
Seongeun Kim (Formal analysis; Software)
Chul Hee Choi (Funding acquisition; Supervision; Visualization; Writing – original draft)
Jun Young Heo (Conceptualization; Funding acquisition; Project administration; Supervision; Visualization; Writing – original draft; Writing – review & editing)
Woosuk Chung (Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Visualization; Writing – original draft; Writing – review & editing)
Supplementary Materials
Custom-made mask used for applying sevoflurane in late postnatal mice.
Surgery combined with 30 minutes of 2.5 % sevoflurane exposure does not affect long-term behavior.
Statistical results.