Evaluation of Hypothermia for Anesthesia in Reptiles and Amphibians

Introduction

When reviewing research proposals involving ectothermic animals, members of institutional animal care and use committees (IACUCs) are frequently required to evaluate husbandry methods and research techniques that are not well known to them. Gathering additional information on these techniques is often unrewarding due to the paucity of published accounts and the tremendous diversity among ectotherms.

One such topic of concern is the use of hypothermia for anesthesia. This anesthetic method has been used for over a century (Blair 1971) and is still often proposed by investigators who use amphibians and reptiles. IACUC members frequently express discomfort with the adequacy of this method; they may be unfamiliar with and lack specific knowledge about hypothermia and may suspect that it is insufficient to render an animal insensible to painful procedures such as surgery. Investigators may base their arguments for its use on past successful experiences where the hypothermic animal failed to respond to any stimuli. Investigators may also have experienced better survival rates with hypothermia than with traditional anesthetics. The scientific literature is replete with references to hypothermia used for anesthesia, and there are some concerns that anesthetics can adversely affect an experiment (Smith and others 1991). Also in the literature are counter-balancing statements that regard cold anesthesia as malpractice.¹

The following review is an effort to present information that is known in the area of hypothermia in order to assist IACUC members in evaluating its use. Currently there is insufficient information to determine authoritatively when and if hypothermia is appropriate; however, there are some basic studies that are useful when reviewing the issue.

Research and Regulatory Guidelines

The regulations and guidelines for animal research are not helpful regarding the appropriateness of the use of hypothermic anesthesia for ectotherms. The guidelines for IACUC members that are currently available are not well referenced and are contradictory. Neither the Guide for the Care and Use of Laboratory Animals (Guide) (NRC 1996) nor the standards of the Animal Welfare Act address this issue. The report of the American Veterinary Medical Association (AVMA) Panel on Euthanasia (AVMA 1993) and the Institute of Laboratory Animal Resources (ILAR) report, Recognition and Alleviation of Pain and Distress in Laboratory Animals (NRC 1992, p 115), indicate that hypothermia is unacceptable for euthanasia. Both of these documents assert that while hypothermia decreases metabolism no evidence exists that it raises the pain threshold. The document cited in both cases states ‘there is no evidence that it [hypothermia] raises the pain threshold, i.e., makes the animal less susceptible to painful stimuli’ (Cooper and others 1989, p 11) but does not reference the statement. The listed references of Cooper and others (1989) consist entirely of review articles, books, and correspondence. Paradoxically, Recognition and Alleviation of Pain and Distress in Laboratory Animals further indicates that hypothermia is an effective analgesic for altricial neonates that have not acquired effective ther-moregulation and as an ‘adjunct to general anesthesia in cold-blooded animals’ (NRC 1992, p 83). The Canadian Council on Animal Care (CCAC) Guide to the Care and Use of Experimental Animals acknowledges that hypothermic amphibians appear to be in an unconscious state, but limits the use of hypothermia to nonpainful procedures because unconsciousness has not been assured (CCAC 1980, p 62-69). No references are offered. A newsletter article on pain relief in ectothermic animals indicates that ‘hypothermia is not an analgesic, since nerve conduction and thus the pain response is not abolished by temperature’ (Arena 1990). The statement is unreferenced; however, there is an article listed in the bibliography that relates to temperature effects on nerve conduction in a reptile. This article is an abstract that has been misinterpreted by Arena; both the abstract and the subsequently published article indicate a clear linear relationship between nerve conduction velocity and temperature (Rosenberg 1977, 1978). Decreasing temperature causes decreased nerve conduction velocities. Further, nerve conduction blockage was recorded at temperatures of 1-3.5°C in tortoises (Rosenberg 1978).

The Scientific Literature

A variety of scientific articles are available that shed some light on the interaction between temperature and nerve function.

The literature characterizes the anesthetic aspects of hypothermia in mammals very well. Cold is anesthetic in mammals; evidence for this is readily available (deJong and others 1966; Rossi and Britt 1984; Antognini 1993; Guerit 1994). Anecdotally, local anesthesia in humans by hypothermia is an exceptionally common procedure, for example, it is standard practice for ear piercing. Further, the human anesthetic literature has a wide variety of references to hypothermia. Specifically, peripheral nerve conduction velocity decreases linearly with decreasing temperature, down to 23.5°C in humans anesthetized with halothane (deJong and others 1966). Linear regression analysis predicted that complete conduction blockade would occur at approximately 9°C (deJong and others 1966). These findings agree with studies in a variety of mammalian species. The general mammalian condition is for hypothermia to induce a neuromuscular and cognitive condition comparable to surgical anesthesia at warmer body temperatures (Blair 1971). The evoked potentials of human and cat central nervous systems cannot be elicited below 20°C (Rossi and Britt 1984; Guerit 1994). A recent study has shown that anesthetized goats cooled to about 20°C do not react to painful peripheral stimuli when the anesthetic is removed (Antognini 1993).

While the data in mammals clearly indicates the neuronal transmission blocking effects of hypothermia, the depressed and blocked function returns with rewarming (Blair 1971; Rossi and Britt 1984; Antognini 1993; Guerit 1994). Therefore, hypothermia would not be expected to supply analgesia once the cooled neurons warmed to functioning level. This would place hypothermia in the class of anesthetics that are not ‘analgesic,’ along with such drugs as pentobarbital and halothane. Hypothermia in mammals also has the potential to cause pain because cold temperature can be a noxious stimulus.

While the mammalian experience with hypothermia is useful for establishing anesthetic concepts, the specific details of the effects of hypothermia on reptiles and amphibians must be sought in literature involving ectothermic animals. Some aspects of hypothermia in ectotherms have been described. As is seen in mammals, peripheral nerve conduction velocities decrease in reptiles (Rosenberg 1978) and amphibians (Hutchinson and others 1970) with decreasing temperature. Hypothermia also reduces electroretinographic responses in amphibians (Schaefer and others 1978). The mechanism may be related to a similar reduction in ionic currents in Xenopus neuronal membranes (Frankenhaeuser and Moore 1963). The response of brain-stem auditory evoked potentials to hypothermia has been studied in alligators (Strain and others 1987). While the amplitude of the potentials decreased with decreasing temperature, potentials were recorded in the alligators at stabilized cloacal temperatures of 0.4°C. The authors further comment that ‘the presence of periodic head and limb movements’ was used to assure subject viability at ‘temperature extremes’ (Strain and others 1987). As is expected for these ectotherms, the temperature range in which neurons function is much cooler than is seen in mammals. Studies in hypothermic ectotherms show neuronal function below the temperature in which conduction blockade occurs in mammals (Hunsaker and Lansing 1962; Parsons and Huggins 1965; Walker and Berger 1973; Rosenberg 1977, 1978; Schaefer and others 1978; Strain and others 1987). While it has been shown that hypo-thermia decreases neuronal function and that ‘numbing’ can be anticipated, the critical point at which anesthesia occurs peripherally is when conduction blockade occurs. In a study of tortoises, peripheral neuron transmission generally blocked at 3.5°C but sometimes blocked as low as 1.2°C (Rosenberg 1977). Prior acclimatization of the tortoises to ambient temperatures that were warmer or cooler than room temperature had no apparent effect on the subsequent blocking temperature (Rosenberg 1977, 1978). A study in bullfrogs reported that in vitro, peripheral nerve conduction was blocked at temperatures of 0-2°C (Roberts and Blackburn 1975). Furthermore, in contrast to the findings in mammalian nerves, it has been reported that in bullfrog nerves, pain-carrying C fibers are blocked at higher temperatures than are the neuromuscular A fibers (Roberts and Blackburn 1975). These findings suggest there is a level of hypothermia that would block transmission of noxious stimuli. However, the temperature at which conduction blockade has been recorded is not substantially different from the ultimate body temperature that would be predicted by standard hypothermia-inducing technique (Smith and Varnold 1991). Ice water induction of hypothermia would leave a small margin for error between conduction blockade and frozen tissue. There is further reason for concern because in both endotherms and ectotherms, detectable muscular movements are blocked at warmer temperatures than the temperatures that block neuronal transmission (Hunsaker and Lansing 1962; deJong and others 1966; Rosenberg 1978). These findings suggest that techniques for monitoring the adequacy of anesthesia, such as testing for withdrawal reflexes, would not be useful for monitoring hypothermia as paralysis will occur before local anesthesia.

The critical question whether hypothermia is able to render ectotherms unconscious prior to inducing peripheral neuromuscular blockade has neither been established nor adequately addressed in the literature. Some experimental evidence exists that suggests that hypothermia does not readily induce unconsciousness in these animals. Walker and Berger (1973) described 2 electroencephalographic states of tortoises: (1) behavioral activity and arousal, and (2) behavioral inactivity. EEG findings for both of these states were distinctly different. Cooling the tortoises, while causing behavioral inactivity, caused the EEG to record a state of arousal. EEG findings in the caiman have shown increased EEG activity in some brain regions at 2-4°C over those seen at room temperature (Parsons and Huggins 1965). Hunsaker and Lansing (1962) reported the EEGs of lizards at various body temperatures. Below 2°C, the EEGs were at isoelectric baseline and at 2-3°C, potentials were inconsistently seen. However lizards routinely had EEG responses to blowing air, flashing lights, and adjacent hand clapping at body temperatures of 2-3°C. One lizard was reported to have EEG responses to air puffs at a body temperature of 0.8°C. In this study, heart potentials were not consistently recorded until 5°C, and detectable body movements were generally not seen until body temperatures reached 9-10°C. It is further noteworthy that at room temperature, the use of curare lead to a marked, generalized reduction in EEG amplitude and an EEG response to air puffs remarkably similar to that seen in hypo-thermic conditions (Hunsaker and Lansing 1962). Although EEG responses to stimuli are not equivalent to consciousness, these findings in hypothermia are clearly different from those seen in mammals where isoelectric and unresponsive EEGs are found at much higher temperatures than are needed to block peripheral nerve function (Blair 1971). Since animals must be conscious to function in their environment and ectothermic animals routinely function at lower temperatures, the findings are consistent with an interpretation that hypothermia does not readily induce unconsciousness in reptiles. There does not seem to be EEG data for hypothermic amphibians, and the reptilian studies may not be indicative of amphibian responses to cold. However, since reptiles are known to use behavioral modification to maintain their body temperature above ambient temperature, there is no reason to suggest that reptiles would display better neural function during extreme hypothermia than amphibians who normally function at those lower ambient temperatures.

Hypothermia has been reported to cause brain necrosis in snakes and turtles (Northcutt and Butler 1974; Wang and others 1977). This may be the cause of clinical wasting that has been anecdotally reported to occur irregularly following hypothermia (Johnson 1992). Clinically apparent adverse effects have not been reported following hypothermia in amphibians. Female Xenopus spp. do very well following hypothermia even after repeated episodes including abdominal surgery (personal observation). A wide variety of alternative anesthetic regimes are available for amphibians and reptiles that do not use hypothermia (Kaplan 1969; Northcutt and Butler 1974; Wang and others 1977; CCAC 1980; Cooper and others 1989; Schaeffer 1994; Stoskopf 1994). Hypothermia may be a useful adjunct to the use of chemical anesthetic agents. The anesthetic requirement in both mammals (Eger and Johnson 1987; Antognini 1993) and fish (Cherkin and Catchpool 1964) has been shown to be dramatically less when hypothermia is induced. Studies have found that amphibians have a potent opioid in their skin (Braga and others 1984), which may have some influence on their recovery from surgical procedures. While it has been suggested to house ectothermic animals below their preferred temperature a day or two before using anesthetics (Arena 1990), this strategy could make amphibians less tolerant to pain and less susceptible to the analgesic effects of administered opioids (Stevens and Pezalla 1989).

MS-222 is often considered the anesthetic of choice in amphibians (CCAC 1980; Cooper and others 1989; Johnson 1992; Schaeffer 1994). An excellent review of its pharmacology and use in amphibians has recently been published (Downes 1995). However, a recent standard research methodology text advocates hypothermia, indicating that MS-222 induces the maturation of Xenopus oocytes (Smith and others 1991). While anesthetic interference with research can be a compelling argument against its use, it is unclear from the literature whether MS-222 actually adversely affects oo-cytes. Justification for Smith's statement is not supported following close examination of the citations. Smith initially cites himself in which he reproduces, in part, a table by Baulieu and others (1978). The Baulieu article reports that lidocaine, dibucaine, and tetracaine induce oocyte meiosis and that benzocaine and procaine do not. MS-222 is not reported. Further, the citations for tetracaine are articles dealing with calcium-membrane interactions. None of the references involve oocyte membranes nor do they model clinical application of tetracaine for frog anesthesia. Table 2 in Baulieu and others (1978) is an interpretation of drugs that could induce meiosis based on their interaction with calcium. The actual effects of the drugs on Xenopus oocyte maturation are not reported. While MS-222 may effect Xenopus oocytes, this has not been reported in the literature and its use should not be avoided based on these citations.

Amphibians and reptiles include thousands of very diverse species. Although the available articles related to the subject are inadequate for such a large and diverse group, they generally do not support hypothermia as a clinically efficacious method of anesthesia.

The author would like to acknowledge the community on COMPMED for their assistance and comments. Dr. Mark Suckow contributed valuable editorial assistance.

References