Standardizing Perioperative Heat to Improve Mouse (Mus musculus) Recovery

Abstract

Maintaining the core body temperature of anesthetized rodents is essential because of the depression of physiologic homeostasis caused by anesthetics. The maintenance of core body temperature is influenced by the ability of the heating device to provide sufficient heat, the presence of material that might alter heat transfer, and the administration of heat in the period surrounding anesthesia. In this study, optimal heat transfer from 4 unique heating devices, with or without insulating drapes, was determined initially with an inert model. Optimal animal recovery, as evaluated by recovery to baseline activity and normothermic temperature of postoperative animals implanted with an intrabdominal thermometer and monitored with a digitally ventilated caging, was achieved by a device that provided consistent electrically supplied heat at 40 °C throughout the perianesthesia period including 35 minutes before isoflurane anesthetic induction. Animals without preoperative heating required at least 24 hours to return to normal core temperature and normal circadian activity levels.

Introduction

The critical need for maintenance of body temperature during anesthesia is important when small rodents are involved with their high surface area to body mass ratio and high metabolic rate.¹ Hypothermia during general anesthesia has been shown to impact tissue perfusion, alter immune responses, decrease drug clearance, and increase toxicity with resultant delayed recovery and poor surgical outcomes.^2–4 Death can result when these hypothermic effects are compounded with the effects from inhaled anesthetics of decreased respiratory function and cardiac output.¹

Recent publications have described multiple devices and approaches to maintain a normothermic state during anesthesia, but a standardized approach to compare devices has yet to be developed.^5–9 Circulating water blankets, far infrared warming, electrical coil heating pads, and exothermic gel packs are routinely used to warm animals and recovery cages during surgery. Circulating water blankets are one of the most common external heating sources used with rodents. They deliver consistent heat and are usually adjustable, making them a good choice to maintain heat throughout the anesthesia cycle.¹⁰ Far infrared warming uses electromagnetic waves to generate heat that warms deep within the animal.⁸ Electric coil heating pads are easy to find and use but can be unreliable in maintaining temperature and carry a burn risk for the animals.⁵ Exothermic packs are easy to handle, and many are also reusable, providing positive sustainability impacts.⁵ In addition, insulating animals with draping material can reduce heat loss by minimizing exposure to the environment and minimizing burns if warming devices become too hot. Paper draping is inexpensive and can be autoclaved, which make it a good insulation choice.¹¹ Other options such as reflective foils and cling wraps can also be used.

Traditionally, core body temperature in rodents has been measured rectally. This method can be cumbersome for technicians and stressful for the animals. Implantable transponders provide measurable value over more traditional methods, and allow continual temperature readings taken without handling the animal.¹² This results in reduced data errors and earlier insights into animal health.¹³

This study used a nonanimal model substitute to evaluate multiple heating devices and the impact of the insulating material used between the animal substitute and the heating element. An inanimate model was initially selected in accordance with the AstraZeneca Replacement, Reduction, and Refinement (3Rs) focus. The replacement of live animals for the initial selection reduced the total number of animals required for the study. Findings were validated with live animals through monitoring recovery to baseline activity with digital ventilated cages (DVC) after surgical implantation of a temperature monitoring device.

Ethical Review

All procedures involving animals were conducted in a facilty accredited by AAALAC International, with all procedures described in an IACUC-approved animal use protocol. All housing, care, and procedures were in accordance with the Guide for the Care and Use of Laboratory Animals,¹⁴ AstraZeneca’s bioethics policy, and AstraZeneca’s Animal Care and Welfare Standard.

Materials and Methods

Experimental animals and randomization.

Adult VAF/Plus C57BL/6CRL female mice (18-20 g; Charles River Laboratories, Wilmington, MA) were randomly housed at 4 per cage in a sterilized GR500 DVC (Tecniplast USA, West Chester, PA) for 14 days before surgery for acclimation and to create an activity baseline. Cages contained a tunnel and 2 types of nesting materials (Bed-r'Nest, The Andersons, Maumee, OH; and Teklad Diamond Twists, Inotiv, West Lafayette, IN). Mice were housed on irradiated 1/4-inch corncob bedding (Inotiv, West Lafayette, IN), with ad libitum access to chlorinated water and irradiated Teklad 2918 diet (Inotiv, West Lafayette, IN). The animal holding rooms were maintained at a 12-hour light/12-hour dark cycle, a temperature of 72 ± 2 °F (22.2 ± 1.2 °C), and a relative humidity of 30% to 70%. Animal health was assessed via cage-side clinical observation and review of the DVC reports.

Study design.

This study was composed of 2 distinct segments. Segment one involved the study of heat absorption in a nonanimal model. A latex balloon with saline was tested on 6 devices to determine which device maintained the internal temperature of the saline at 38 °C. From segment one, one of the top-performing devices was then tested with live animals with or without external heat provided before anesthesia.

Study design: Heat absorption in a nonanimal model.

A nonanimal model constructed using a 12-inch latex water balloon filled with 30 mL of 0.9% saline was used to validate the ability of a heating device to maintain body temperature over time. The saline-filled balloon was comparable in size, weight, and surface area to a 30-g mouse. The nonanimal balloon model was heated in a water bath set to 38 °C. Once the temperature reached 38 °C, the balloon was quickly dried with a paper towel and moved to the test site.

Heating elements evaluated include an electrical coil heating pad (Rodent Warmer ×2 Large; Stoelting, Wood Dale, IL), Stryker T/Pump and recirculating water heating pad (Stryker, Kalamazoo, MI), an exothermic gel pack (Snap Heat 4-inch × 10-inch Medium Gel Pack; Snappy Heat, Fort Worth, TX), an anesthesia system with electrical integrated heating beds (E-Z Systems, Palmer, PA), and a FAR Infrared Rodent Warmer (Kent Scientific, Torrington, CT). All devices were used according to the manufacturer’s instructions and had been calibrated by the manufacturer (Figure 1).

Figure 1.

Nonanimal Mouse Balloon Substitute and Heating Devices. (A) Nonanimal mouse balloon model substitute, (B) Rodent Warmer ×2 Large electrical coil device, (C) Stryker T/Pump recirculating water blanket, (D) Snap Heat exothermic gel pack, (E) FAR Infrared Rodent Warmer, and (F) E-Z Systems V-shaped red (left) and flat yellow (right) heating beds.

The surface temperature of all 5 heating elements was measured over 70 minutes using an infrared thermometer (Traceable Infrared Thermometer Gun; VWR International, Radnor, PA). Four of the 5 heating elements were powered on and set to 38-40 °C. The exothermic gel pack was initiated, and the temperature was recorded with the same infrared hermometer for the same duration of time.

The core temperature of the nonanimal balloon model was measured with an immersion probe (Thermocouple Thermometer 53/54 II B' Fluke, Inc., Everett, WA) for 10 minutes while the model was placed directly on a benchtop with no heating element, as well as directly on the heating element with or without drape material. The 2 drape materials placed to create a barrier between the nonanimal model and the heating element were an absorbent blue pad (Wypall; Kimberly Clark, Irving, TX) or a paper towel.

Anipill (Animals Monitoring; St. Clair, France) temperature models were chosen for surgical implantation in place of a rectal temperature probe. Therefore, bridging data were collected, which repeated the nonanimal model testing without insulating drapes. The core temperature of the nonanimal balloon model was measured with the Anipill temperature monitor for 10 minutes when placed between the heated nonanimal model and the heating surface.

Study design: Anipill implant surgery experimental procedures and outcome measures.

Mice were shaved on the right flank and provided with meloxicam tablets (0.125 mg meloxicam/tablet; Bio-Serv, Flemington, NJ) on the day before surgery. On the day of surgery, the mice were anesthetized with isoflurane in a heated induction chamber (AD-5000; EZ Systems, Palmer, PA) until loss of righting reflex, and sterile ophthalmic lube was applied to the eyes. Animals in both experimental groups, as defined below, were placed on a heated surgical platform V-Shaped Heated Surgery Bed (38 °C; 4 inch × 10.5 inch; E-Z Systems, Palmer, PA), and the surgical site was prepared with alternating wipes of iodine and alcohol and covered with a Tegaderm bandage as a surgical drape (3M, Saint Paul, MN). Once appropriate anesthetic depth was confirmed by toe pinch, a 1-cm incision over the right flank into the abdomen was made, and the MetriCide 28 High-Level Disinfectant Solution (Metrex, Orange, CA) was sterilized Anipill (Animals Monitoring, St. Clair, France) and placed into the abdomen. The abdominal wall was closed with 5.0 polyglycolic acid suture (Ethicon; Johnson and Johnson, Raritan, NJ), and the skin was approximated using 7-mm sterile surgical staples (BD Autoclip; Becton, Dickinson, and Company, Franklin Lakes, NJ). Animals were returned to a new sterile cage following surgery that had been prewarmed for 20 minutes to 40 °C. Twenty minutes after all animals in the cage were fully ambulatory, the recovery cage was returned to the DVC rack. This approach resulted in variable recovery heating durations: the fourth mouse in each group received the shortest exposure (approximately 20 minutes), whereas earlier mice had progressively longer recovery heating due to the sequencing of surgeries and the use of a single heated cage per group. Animals were evaluated twice daily for pain and distress and nest quality and provided with a meloxicam tablet daily for 3 days after surgery. Temperature data for each mouse were recorded at 2-minute intervals by way of an Anipill implanted into the intraperitoneal space during the first 48 hours postprocedure and then at 15-minute intervals to conserve battery life.

Study design: Surgery groups and sample size.

Two experimental groups were established. Group 1 had the home cages maintained at 40 °C for approximately 30 minutes before anesthesia and during anesthesia induction using the EZ-178 Induction Chamber (E-Z Systems, Palmer, PA) with the E-Z Systems Heated Chamber Warmer positioned beneath and an EZ-HP170 Chamber Connection Plate between the warmer and the chamber. For group 2, an identical equipment setup was used, with the home cage and induction chamber maintained without supplemental heat at 22 °C for 30 minutes before anesthesia. All 8 animals from both groups were placed in a prewarmed recovery cage following anesthesia. Heating elements were activated at least 20 minutes before use.

Statistical methods.

For data analysis, a pairwise analysis was conducted with mouse 1.1 compared with 2.1, 1.2 with 2.2, 1.3 with 2.3, and 1.4 with 2.4, reflecting their matched positions within the treatment sequences within each experimental group with the assumption that mouse-mouse interaction within a cage does not influence body temperature. The dataset for each matched pair of animals was partitioned into distinct temporal stages to capture relevant physiologic and experimental transitions, allowing for uniform segmentation across groups. This structure ensures that comparisons between groups reflect equivalent procedural timing and exposures. Within each stage, the between-group difference in body temperature was estimated independently for each paired comparison. For each stage, the pairwise estimates were subsequently combined using meta-analytical techniques to assess the overall effect of preheating, independent of individual preheating durations.

The aggregation of these paired differences for the overall heating effect was formulated as a noninferiority test. The alternative hypothesis asserts that group 1 (preheated) maintains a body temperature at least l(t) degrees higher than group 2 (nonpreheated). For statistical inference, a generalized additive model incorporating random effects and autocorrelated errors was fitted to all data to ensure unbiased inference even in the presence of complex temperature trajectories. Hypothesis testing was performed using the hypothesis function with a Benjamini-Yekutieli adjusted P value from the R package (R Foundation for Statistical Computing). An adjusted P value of less than 0.05 was considered significant.

Inclusion and exclusion criteria.

Data were excluded from analysis for the first 30 minutes following anesthetic recovery as the temperature fluctuation was due to the Anipill adaptation to the mouse body temperature and variability due to individual mouse recovery. No animals were excluded from analysis. Statistical analyses were not applied to the nonanimal model and the DVC activity levels; rather, these data were empirically described.

Results

Heat retention with a nonanimal model.

The surface temperature of all the devices fell below 37 °C within the first minute of placement on the external heating device, except for the exothermic gel packs, which reached 52 °C and gradually decreased throughout the 10-minute monitoring period. Core temperature of the nonanimal balloon model decreased 10 °C when placed directly on a stainless-steel work surface (black triangle). This decrease was partially ameliorated with an insulating paper towel (red square) or absorbent blue pad (blue circle) with a resultant loss of 7 °C, independent of drape type (Figure 2A). The exothermic gel pack caused an increase of 2 °C in the absence of a drape with a paper towel, permitting maintenance of 37 °C and a blue absorbent pad correlating with the typical loss of 2 to 3 °C seen with the infrared, recirculating water blanket, or the electrical coil devices (Figure 2). This temperature loss was not seen in the E-Z heated system when set at 40 °C (Figure 2F). The paper towel and the blue absorbent pad had minimal impact on temperature. From these data, the E-Z anesthesia system was selected for all future anesthetic heating use (Figure 3).

Figure 2.

Core Temperature of the Inanimate Model Using Various Heating Elements and Drape Materials. All active heating elements improved temperature retention compared with the unheated control (black triangle). The type of drape material over the heating source influenced the temperature (red square for paper towel or blue circle for absorbent blue pad). Heating conditions included: A, Direct contact with the laboratory bench (no heat source); B, Electrical coil heating pad (Rodent Warmer ×2 Large, Stoelting); C, Recirculating water heating pad (T/Pump, Stryker); D, Exothermic gel pack (4 × 10 in Medium Gel Pack, Snap Heat); and E, Far-infrared rodent warmer (Kent Scientific). No statistical analyses were performed for these data.

Figure 3.

Comparison of All Heating Devices without Insulating Drape. All heating devices were compared with the one selected to use for live animal surgery. Devices tested include E-Z Systems warming bed under induction chamber (blue circle), E-Z Systems V-shaped red heating bed (green square), E-Z Systems flat yellow heating bed (red inverted triangle), Stryker T/Pump recirculating water blanket (hollow purple circle), Rodent Warmer ×2 Large electrical coil device (hollow brown square), Snap Heat exothermic gel pack (hollow blue triangle), FAR Infrared warming device (orange diamond), and stainless steel surface with no heating device (black triangle). BSC, biosafety cabinet.

Failure to reach normothermia in partial heat group.

Animals in group 2 that did not receive preanesthestic heat (Figure 4, blue) were significantly colder (P value less than 10⁻⁹) with an average of 0.4 °C lower in core body temperature when compared with group 1 mice (pink) that received 35 minutes of 40 °C heating before anesthesia. This temperature difference was noted after the first 30 minutes of Anipill implantation. The time between device implant and 30 minutes after implantation was highly variable due to individual mouse variation and device warming to core body temperature but demonstrated a transient temperature decline followed by a gradual rise. Both treatment groups demonstrated a similar recovery trajectory. The pivotal stage spans 30 minutes to day 1.5 postanesthesia. This correlates to the removal of all external heat from the cage once the fourth animal in the treatment group had fully recovered. Every hourly interval yielded highly significant differences when adjusted for multiple testing with the difference between the 2 experimental groups lost at 25 hours after external heating was discontinued. The mice in both groups demonstrated the temperature fluctuation associated with circadian rhythm with the dark cycle initiated at 4.7 hours after surgical completion as shown on the dark bars in Figure 4.

Figure 4.

Average Temperature Difference between Paired Individuals across Experimental Groups. Group 1 (pink, full heat) was consistently elevated by 0.4 °C over group 2 (blue, partial heat). The light cycle is indicated by white bars (lights on) and filled bars (lights off). The numbers above represent the lower bounds of the differences, and the “>” symbol is interpreted in a statistical sense under a (jointly) false discovery rate of 0.05. The light cycle is indicated by white bars (lights on) and solid bars (lights off) at the bottom.

Recovery of core body temperature validated by activity level.

The whole cage activity level from the DVC cages indicated that normal activity level was not seen in group 2 animals that did not receive prewarming (Figure 5). Normal activity is shown by yellow and green bars indicating activity during the dark cycle and dark blue bars indicating lack of activity during the light cycle. The dark cycle following anesthesia showed minimal activity (dark blue and light green) for group 2, whereas the group 1 mice that had received prewarming showed heightened activity (orange and yellow bars) in the dark cycle following anesthesia. Normal circadian activity of heightened activity during the dark cycle and minimal activity during the day was seen during the dark cycle that began 28 hours after anesthesia with both groups displaying a similar core body temperature that peaked midway through the dark cycle. This is correlated to the lack of a structured nest in group 2 for the 2 days following anesthesia.

Figure 5.

Average Activity Level across Experimental Groups. Cage level activity as displayed in a heat map (blue as minimal activity with red as maximal activity) in a digitally ventilated cage. The white gap in the heat map is a lack of activity associated with the cage removed from the digital ventilated cages rack for anesthesia and surgery.

Discussion

This study has shown that recovery to baseline activity and normothermic temperature following isoflurane anesthesia and abdominal surgery requires 35 minutes of preoperative heat from a device that can provide heat at 40 °C. Animals without preoperative heating required at least 24 hours to return to normal core temperature and circadian activity.

While many prior studies have demonstrated the critical need for heat during anesthesia, few have discussed the importance of preoperative heat, and none have shown the long-term impact of lack of preoperative heat supplementation. Prior studies^7,15 in rats have demonstrated the importance of prewarming to prevent hypothermia during recovery, but these studies were limited to the immediate postoperative window. The impact when measured by a highly accurate core body temperature extended to the first day postoperative and the first 3 days by activity assessment and nest building. This failure to maintain normal activity and body temperature postoperatively calls into question the impact of short-term anesthesia on animals and when animals are physiologically normal after anesthesia.

While attempting to identify the optimal heating device, an inert model was used to evaluate the heating devices used by different individuals. It was surprising that the exothermic gel packs quickly exceeded the normal body temperature of animals when used without an insulating material risking thermal injury. Many of the other heating devices failed to maintain the normal mouse body temperature of 37 °C, which was further attenuated by the presence of insulating material that is commonly used to minimize cleaning between animals. A clear indication of this study reinforces the manufacturer’s instructions to place the animal directly on the heating surface to allow optimal heating rather than placing material that would impact heat conduction between the mouse and the heating device. These initial studies with a nonanimal model allowed for a rapid selection of an optimal device. The device chosen for the live animal portion of this study was chosen based on ease of operation and a built-in feedback loop to maintain a constant temperature. Subsequent studies using this optimal device validated the importance of perioperative heating, especially preoperatively.

There was an observable warm-up time needed for the Anipill’s internal components to increase in temperature from room temperature to match the animal’s temperature. Future studies will ensure that the Anipill is stored in 40 °C sterile saline before implant.

As seen in rats, prewarming did not impact time to observable recovery¹⁵ with the primary difference seen hours after the postoperative recovery. In fact, the average peak temperature difference between the 2 groups was seen 8 hours postoperative at the peak of the dark cycle.

The implications of this research extend to optimizing postoperative animal care and improving surgical outcomes. By systematically quantifying the preheating effect, the study reinforces the value of individualized temperature management protocols and demonstrates an integrative analytical framework capable of handling high-density physiologic data.

Limitations of the current work include the modest sample size, imposed by logistical and ethical constraints, and residual confounding from cage-level effects and surgery sequencing. Further research with larger cohorts and more extensive controls may help to generalize these observations. While it would also be interesting to track the body temperature change during anesthetic induction and surgery, this would require multiple survival surgeries to implant the transponder and then conduct a subsequent sham surgery. Nevertheless, the approach outlined here is both technically rigorous and aligned with current best practices, supporting its use in future studies of thermoregulation and perioperative animal welfare.

Conflict of Interest

The authors have no conflicts of interest to declare.

Funding

This work was internally funded by AstraZeneca.