Efficacy and Safety of Stored and Newly Prepared Tribromoethanol in ICR Mice

CHRISTINE C. LIEGGI, DVM,1,* JAMES E. ARTWOHL, MS, DVM, DIPLOMATE, ACLAM,1 JORI K. LESZCZYNSKI, DVM, DIPLOMATE, ACLAM,2 NANCY A. RODRIGUEZ, DVM, DIPLOMATE, ACLAM,3 BARRY L. FICKBOHM, PHD, DVM,4 AND JEFFREY D. FORTMAN, DVM, DIPLOMATE, ACLAM1

This study, performed in conjunction with an in vitro evaluation of tribromoethanol (TBE), consisted of three trials with three objectives. The first objective was to compare anesthetic efficacy and short-term pathologic findings of TBE, ketamine–xylazine (K–X), and sodium pentobarbital (NaP). The second objective was to evaluate how changes of TBE that occur during the perceived most favorable and least favorable storage conditions (8 weeks at 5°C in the dark [5D] and 25°C with exposure to light [25L], respectively) affect anesthetic efficacy and short-term pathology when compared to newly prepared TBE. The third objective was to perform a 6-week clinical assessment of animals that received newly prepared TBE. All animals that received TBE (400 mg/kg) and 14 of 15 that received K–X (K, 120 mg/kg; X, 16 mg/kg) were anesthetized, as defined by loss of pedal reflex. In comparison, only 8 of 15 animals administered NaP (60 mg/kg) were anesthetized. Anesthetic duration for animals that received K–X was 31.7 min, which was significantly (P = 0.0085) longer than animals that received TBE (18.5 min). Recovery times for TBE and K–X were not significantly different (26.5 and 27.5 min, respectively). Pathologic lesions associated with TBE administration were significantly (P = 0.001) greater than those associated with K– X. NaP was not associated with any pathologic lesions. The pH of newly prepared and 5D TBE was 6.5 to 7.0, whereas that for 25L TBE was 3.0. Anesthetic induction, duration, recovery times, and pathologic lesions were not significantly different, regardless of the pH or storage condition of the solution. It was noted, however, that the average anesthetic duration for animals administered newly prepared TBE in the second trial was longer (37.7 min) than the first trial that used newly prepared TBE. For the third trial (long-term clinical assessment), the average anesthetic duration for TBE was 46.5 min, significantly (P < 0.025) longer when compared to the first trial that used newly prepared TBE. During the third trial, 10 animals were found dead or moribund. All animals that were found moribund were necropsied and found to exhibit a marked ileus. Because of the variability in anesthetic effectiveness, pathology, and morbidity and mortality associated with the use of TBE, we do not recommend the use of this anesthetic agent in ICR mice.

Tribromoethanol (TBE), previously available under the trade name (9). The present in vivo part of the study had three main objectives: Avertin, has been used as an anesthetic in humans and animals since 1) compare anesthetic efficacy and short-term pathologic effects of TBE, the 1930s. TBE was associated with hepatic damage in humans and ketamine–xylazine (K–X), and sodium pentobarbital (NaP); 2) evalu- has not been used as a human anesthetic since the late 1940s (1), ate how changes that occur during the perceived best and worst storage however TBE continues to be used in laboratory mice because it conditions affect anesthetic efficacy and short-term pathology when rapidly induces an adequate depth of surgical anesthesia with a rela- compared with those of newly prepared TBE; and 3) perform a 6-week tively quick recovery. Much controversy surrounds the use of TBE clinical assessment of animals that received newly prepared TBE. in the laboratory animal science community because of conflicting reports on its efficacy and safety (2-6). Materials and Methods Gardner et al. (2) reported the inability of TBE to reliably pro- TBE preparation. Commercially available TBE powder (2,2,2- duce an adequate anesthetic plane, whereas others describe consistent tribromoethanol, Acros Organics, Pittsburgh, Pa.) from the same lot surgical anesthesia (3, 4). Several studies have evaluated the tissues of number (A015409401) was used for all studies. The bottle of TBE mice following TBE administration and reported various degrees of powder that was used for experiment 1 was different than the bottle pathologic lesions and clinical outcomes. Adverse effects that have used for experiments 2 and 3. To prepare a 1 g/ml (stock) solution, a been associated with the use of TBE in mice include intestinal ileus, measured amount of TBE powder (in g) was placed into a Class B muscle necrosis, serositis of several organs, and death (4-6). Use of borosilicate glass vial (Fisher Scientific, Pittsburgh, Pa.) that was TBE has also been associated with post-anesthetic illness and death wrapped in aluminum foil. A glass pipette or Hamilton syringe was in the rat (7) and Mongolian gerbil (8). These discrepancies may be used to add tert-amyl alcohol (Sigma Aldrich, St. Louis, Mo.), and attributed to different formulations, various doses, different tissue the solution was vortexed for 2 min at room temperature (25°C). To evaluation time points, and/or the lack of a standard method for prepare a 25 mg/ml (working) solution, sterile water for injection preparation and storage of TBE. (Abbott Laboratories, North Chicago, Ill.) was added to the appro- This is the second segment of a two-part study designed to evalu- priate volume of stock solution and vortexed for 2 min at 25°C. ate TBE anesthesia in mice. The first part of the study evaluated the Working solutions were sterile filtered with a 0.22-µm Durapore preparation, storage, and use of TBE with in vitro methodologies polyvinylidene fluoride membrane Millex-GV syringe filter unit (Fisher Scientific, Pittsburgh, Pa.). All solutions were prepared in a University of Illinois at Chicago Biologic Resources Laboratory, 1840 W. Taylor Street M/ fume hood, the preparer wore gloves, glass vials were autoclaved, and C 533, Chicago, Illinois 606121; Cleveland Clinic Foundation, Biologic Resources Unit/ all glassware was triple-rinsed with tert-amyl alcohol for stock solu- 2 NC50, 9500 Euclid Avenue, Cleveland, Ohio 44195 ; Laboratory Animal Services, Medical tions and sterile water for working solutions. College of Georgia, CB 1102, Augusta, Georgia 309123; Bayer CropScience LP, 17445 S. Metcalf Avenue, Stilwell, Kansas 660854 Animals. Female ICR mice (weight, 24 to 30 g; age, 38 to 45 *Corresponding author days) were obtained from Harlan (Indianapolis, Ind.). The specific

Volume 44, No. 1 / January 2005 CONTEMPORARY TOPICS © 2005 by the American Association for Laboratory Animal Science 17 pathogen-free (SPF) mice were monitored by the supplier for the Table 1. Scale used to score pathologic lesions following pathogenic agents: Mycoplasma pulmonis, Sendai virus, 0 – Normal No recognizable lesion is present, or the extent of any mouse hepatitis virus, pneumonia virus of mice, minute virus of mice, observed spontaneous tissue change is within normal mouse parvovirus, mouse encephalomyelitis virus, reovirus-3, biological limits. rotavirus, ectromelia virus, lymphocytic choriomeningitis virus, mouse adenovirus, polyoma virus, mouse cytomegalovirus, Hantaan virus, 1– Minimal The distribution and/or extent of tissue change is the least amount (0-25% affected) or degree discernible with little mouse thymic virus, K virus, Encephalitozoon cuniculi, and Clostridium tissue destruction or change from normal. piliforme. Mice were housed in groups of five in sanitized polysulfone microisolator cages on corncob bedding. They were given ad libitum 2– Mild There is a graduated increase in the distribution and/or access to rodent chow (Harlan Teklad 22/5 Rodent Diet 8640) and extent of tissue change (26-50% affected) but with a still- municipal water in bottles. The animal room was maintained in ac- limited variation from normal structure and limited presumed effect on function. cordance to temperature and humidity recommendations of the Guide for the Care and Use of Laboratory Animals (10), with 14 h of light 3 – Moderate The distribution, amount, and/or extent of tissue change and 10 h of dark. All experimental procedures were approved by the (51-75% affected) is well-developed with definite tissue University of Illinois at Chicago Animal Care Committee and per- destruction or change from normal. There may be a formed in the Biologic Resources Laboratory, an Association for the relatively large area of involvement (focal, multifocal, or diffuse), multiple cell involvement, and/or the majority Assessment and Accreditation of Laboratory Animal Care, Interna- of the organ or tissue may be affected that may greatly tional-accredited facility. compromise the ability of the tissue or organ to function Experimental procedures. (i) Experiment 1—efficacy and safety normally. of TBE. Mice were randomly assigned into one of six groups: group 1, TBE (400 mg/kg); group 2, K–X (ketamine, 120 mg/kg, Fort 4 – Severe The distribution, amount, and/or extent of tissue change (76-100% affected) is the most severe or extensive level Dodge Animal Health, Fort Dodge, Iowa; xylazine, 16 mg/kg, Phoe- possible. The lesion is extensively developed with nearly nix Scientific, St. Joseph, Mo.); group 3, NaP (Nembutal, 60 mg/ total tissue destruction or involvement. Again, kg, Abbott Laboratories, North Chicago, Ill.); group 4, tert-amyl al- depending on the nature of the lesion, this degree of cohol/sterile water (TAASW); group 5, sterile water; and group 6, involvement may greatly compromise the ability of the no treatment. Groups 1 through 5 consisted of 15 mice each, whereas tissue or organ to function normally. group 6 consisted of five mice. Group 4 received a sterile, filtered tert-amyl alcohol/sterile water solution at the same ratio of tert-amyl ment; group 2, sterile water; group 3, TAASW; and group 4, newly alcohol to sterile water that is found in the TBE working solution. prepared TBE working solution (400 mg/kg). Each group consisted Anesthetic agents were diluted with sterile water for injection to con- of 17 mice. After i.p. administration of the agent, induction time, centrations that allowed all animals to receive injection volumes of anesthetic duration, and recovery time were evaluated using meth- 0.29 to 0.48 ml. After intraperitoneal (i.p.) administration of the ods identical to those used in experiments 1 and 2. Mice were weighed agent according to current body weight, mice were maintained on a 24 h after administration and then twice weekly for 6 weeks and Shor-Line temperature-controlled heating board (Schroer Manufac- monitored for clinical signs of illness such as lethargy, unkempt hair turing Co., Kansas City, Mo.) and monitored for induction time, coat, hunched posture, or dehydration. anesthetic duration, and recovery time. Induction time was defined Pathologic evaluation. After a final body weight and clinical as- as time from the injection to the loss of pedal reflex, anesthetic dura- sessment, mice were humanely euthanized via cervical dislocation tion as time between loss and return of pedal reflex, and recovery under CO2 anesthesia. A necropsy was performed and abnormalities time as time from return of pedal reflex to movement around the noted. The gastrointestinal tract (from the terminal esophagus to cage. Pedal reflex was assessed using a Touch-Test Sensory Evaluator the rectum), spleen, and liver were weighed immediately after col- (North Coast Medical, San Jose, Calif.) with a target force of 300 g, lection. The stomach, small intestine, cecum, colon, liver, spleen, which indicates deep pressure sensation. A pedal reflex was defined and body wall were collected for histopathologic evaluation. A seg- as withdrawal of the limb when touched by the sensory evaluator. ment (minimum of 2.5 cm) from the duodenum, jejunum, ileum, Alternate hind paws at alternate sites were assessed every 2 min. Af- and colon was submitted for assessment in saggital section. In cases ter recovery, mice were monitored daily for body weight and general associated with adverse clinical effects, kidney, heart, brain, lung, well-being. For groups 1 through 5, five mice per test group each femur, spine, and the reproductive tract also were evaluated. Tissues were evaluated for pathology at 24 h, 4 days, and 10 days. Group six were fixed in 10% formalin and stained with hematoxylin and eosin (no treatment) was evaluated at 4 days. for evaluation by a veterinary pathologist blinded to the treatment (ii) Experiment 2—efficacy and safety of TBE after various stor- groups. Pathologic lesions were scored on a scale from 0 to 4 reflect- age conditions. Mice were randomly assigned into one of three groups ing the severity of inflammation and percentage of organ affected as to receive TBE at a dose of 400 mg/kg: group 1, newly prepared described in Table 1. Figure 1 depicts examples of grades 0 and 4. TBE working solution; group 2, TBE working solution that had Statistical analysis. For experiment 1, anesthetic duration and re- been stored for 8 weeks at the perceived most favorable storage condi- covery time of mice that lost pedal reflex were evaluated with a tion of 5°C in the dark (5D); and group 3, TBE working solution that one-way analysis of variance (ANOVA), with Tukey’s adjustment had been stored for 8 weeks under the perceived least favorable storage for multiple comparisons to evaluate differences among the agents condition of 25°C in the light (25L). Each group consisted of 15 mice. administered. For experiment 2, a one-way ANOVA was performed After i.p. administration of the agent, induction time, anesthetic dura- with a Tukey–Kramer adjustment for multiple comparisons to evalu- tion, and recovery time were evaluated using methods identical to those ate differences among the three anesthetic groups for induction time, used in experiment 1. After recovery, animals were monitored daily for anesthetic duration, and recovery. For experiment 3, a one-way body weight and general well-being. Five animals per test group each ANOVA, with a Tukey-Kramer adjustment for multiple compari- were evaluated for pathology at 24 h, 4 days, and 10 days. sons was used to compare the anesthetic duration of lethal TBE to (iii) Experiment 3—long-term efficacy and safety of TBE. Mice anesthetic duration of TBE in groups 1, 2, and 3 in experiment 2. were randomly assigned into one of four groups: group 1, no treat- Organ and body weights (experiments 1, 2, and 3) and pathologic

Figure 1. Examples of grade 0 (A) and grade 4 (B) histopathology scores for the body wall. H&E stain; magnification ×21.25.

the time of collection. The mild inflammatory reaction noted in control groups 4 and 5 was most likely due to injection technique and was not significantly greater than the inflammatory reaction seen in other treatment groups. Only mice in group 1 had significantly more inflammation at the time of euthanasia than did those in group 2 (P = 0.001), group 4 (P = 0.005), and group 5 (P = 0.084). Mice in group 1 had the highest incidence of lesions (9 of 15 mice), with a severity grade ranging from 1 to 3 indicating minimal to moderate inflammation. The inflammatory response was either a neutrophil cell population or a mixture of cells that consisted of neutrophils, lymphocytes, plasma cells, and an occasional macrophage, indicating acute or subacute inflammation. The inflammation in some instances not only involved the body wall peritoneum but also af- Figure 2. Mean values and standard deviations for the pathology seen in the fected the overlying muscle, resulting in necrotizing myositis. For all body wall after intraperitoneal injection of an anesthetic agent or control at groups that had pathologic lesions, the average degree of inflamma- 24 h, 4 days, 10 days, and the average of the three time points. Mean values tion decreased significantly from day 1 to day 10 (P < 0.0001). are based on a pathology scale of 0 to 4, with grade 0 indicating normal tissue and grade 4 representing severe lesions. Experiment 2—efficacy and safety of TBE after various storage conditions. Mice in all groups lost pedal reflex within 2 to 4 min. Anesthetic duration and recovery times did not differ significantly lesions (experiments 1 and 2) were evaluated using a two-factor between groups and ranged from 31 to 34.4 min for anesthetic dura- ANOVA design, with treatment group and follow-up time as the tion and 21.8 to 26.9 min for recovery time. Mice that received two factors, with Tukey’s adjustment for multiple comparisons. Pa- newly prepared TBE had an average anesthetic duration of 34.4 min, thology scores were averaged across the three time points for analysis which was significantly longer than the average anesthetic duration of each group and group comparisons. recorded for newly prepared TBE in experiment 1 (18.5 min, P < 0.025). Body and organ weights over the 10-day period were not Results significantly different between the groups. Unlike in experiment 1, a Experiment 1—efficacy and safety of TBE. All 15 mice that re- significant difference in body weights across the measurement days ceived TBE (group 1) and 14 of the 15 that received K–X (group 2) was not seen. As in experiment 1, the body wall peritoneum was the lost pedal reflex, in comparison with only 8 of the 15 mice that re- most affected of the organs examined histologically. Pathology scores ceived NaP (group 3). Mice in control groups 4, 5, and 6 did not did not differ significantly when comparing the three groups. The lose righting reflex, although all mice that received TAASW (group degree of inflammation decreased significantly from day 1 to day 10 4) became ataxic. Of the animals that lost pedal reflex, groups 1 and (P < 0.0001). Results are shown in Fig. 3. 2 had similar average induction times (4.42 and 4.13 min, respec- Experiment 3—long-term efficacy and safety of TBE. All 17 mice tively), and both groups demonstrated a much shorter average that received newly prepared TBE (group 4) lost pedal reflex within induction time than group 3 (9 min). Average anesthetic duration of 2 to 4 min, and mice that received TAASW (group 3) became ataxic. anesthesia was 18.5 min for group 1, 31 min for group 2, and 13 Mice that received newly prepared TBE had an average anesthetic min for group 3. The anesthetic duration for group 2 was signifi- duration of 46.5 min, which was significantly longer than the anes- cantly (P = 0.0085) longer than that for group 1. Average recovery thetic duration recorded for newly prepared TBE in experiment 2 time was similar for groups 1 and 2 (26.5 and 26.4 min) but signifi- (34.4 min, P < 0.025). At 24 h after administration, all mice that cantly longer for animals in group 3 (44.5 min, P = 0.0010 for groups received TBE and 2 of the 17 mice that received TAASW appeared 1 and 2). Values for body weights over the 10-day period were not to be somewhat lethargic with ungroomed hair coats. All mice that significantly different between the groups. A significant difference received TAASW appeared healthy 48 h after injection. Four days (P < 0.0285) in body weights across the measurement days was noted after administration, five mice that had received TBE were found for all groups reflecting weight gain. dead, and four were euthanized as they were determined to be mori- The body wall peritoneum was the most affected of the organs bund. Ten days after administration, one mouse that received TBE examined histologically. Mice in all groups, except for groups 3 and was found dead. Dead and moribund mice were found in random 6, had inflammation involving the body wall peritoneum (Fig. 2) at cages, with filled water bottles and food, ruling out any husbandry

Volume 44, No. 1 / January 2005 CONTEMPORARY TOPICS © 2005 by the American Association for Laboratory Animal Science 19 Figure 3. Mean values and standard deviations for the pathology seen in the body wall after intraperitoneal injection of newly prepared TBE (TBE), TBE stored at 5°C in dark (5D) for 8 weeks, and TBE stored at 25°C in light (25L) for 8 weeks at 24 h, 4 days, 10 days, and the average of the three time Figure 4. Appearance of the abdominal cavity of a mouse found moribund points. Mean values are based on a pathology scale of 0 to 4, with grade 0 after injection of newly prepared TBE in experiment 3. indicating normal tissue and grade 4 representing severe lesions. found to be the lowest dose of TBE that resulted in adequate anes- concerns associated with the death of these animals. All mice in con- thesia, based on the toe-pinch reflex and the Touch-Test Sensory trol groups 1 and 2 and most mice in group 3 appeared healthy Evaluator. A dose of 400 mg/kg was found to be the lowest dose that throughout the study. At necropsy, various degrees of distension were consistently produced adequate anesthesia in all ICR mice tested. In seen in the gastrointestinal tract of all mice that were determined our experience, however, the optimal dose of TBE may vary depend- moribund (Fig. 4). The gastrointestinal-tract-to-body weight ratio ing on the mouse strain that is used. < of mice that were clinically ill was significantly (P 0.034) greater TBE commonly is used for the brief surgical procedures necessary than that in control groups, supporting a gross description of ileus. for the production of transgenic mice, such as embryo transfer pro- Histologically, all mice that received TBE demonstrated some de- cedures (3, 4, 12, 15). K–X and NaP are other injectable anesthetic gree of peritonitis, however the mice that died or were euthanized agents that are available to provide surgical anesthesia in mice. In before the end of the study had additional pathologic findings. In- experiment 1, we evaluated the efficacy and safety of newly prepared terstitial pneumonia was observed in these mice and was thought to TBE, K–X, and NaP. The doses chosen for K–X and NaP were based be a consequence of bacterial translocation to the peritoneum, re- on doses that have been reported in the literature (2, 4, 16), personal sulting in septicemia and pneumonia. Bacteria observed in the experience, and unpublished pilot studies (14). The results of ex- abdominal cavity support the idea of an overwhelming infection be- periment 1 demonstrated that NaP did not reliably induce anesthesia, ing the cause of death. and when anesthesia was induced, a short duration of anesthesia was Subsequent to these findings, a study was done to verify the source followed by a prolonged recovery. NaP has been reported to have an of toxicity and determine whether the morbidity and mortality experi- unpredictable dose range, which can result unexpectedly in less-than- enced was a random event (11). Group 1 (10 mice) received working adequate surgical anesthesia in mice (2, 15). NaP can also cause a solution prepared from the same TBE stock used to prepare the work- dose-dependent cardiovascular and respiratory depression and often ing solution administered in experiment 3. Groups 2 and 3 (10 mice is associated with prolonged recovery times. However, in a study each) received working solutions prepared from the same bottle of that compared the heart rates of mouse embryos exposed to TBE or TBE powder used in experiments 2 and 3. The mice were monitored NaP, those exposed to TBE were vulnerable at some developmental for 10 days. Morbidity rates for the three groups in the verification stages to arrhythmias and had significantly lower heart rates than did study ranged from 30% to 50%; the morbidity rate for experiment 3 those exposed to NaP (17). Both TBE and K–X produced adequate (long-term study) was 58%. Results from this verification study sug- anesthesia, although K–X was associated with significantly less in- gested that the clinical findings were repeatable, and the source of flammation when compared with TBE, findings consistent with what the toxicity was the TBE powder. This TBE powder did not differ has been reported in the literature (4). A previous study that com- markedly from TBE powder that did not have lethal effects when pared the use of K–X and TBE in embryo transfer procedures reported evaluated with in vitro methodologies. These methods included gas similar histologic findings (4). The same study evaluated the anes- chromatography–mass spectroscopy, carbon nuclear magnetic reso- thetic efficacy of these two anesthetic agents, as well as embryo transfer nance (NMR) spectroscopy, and proton NMR spectroscopy (9). success rate after the use of these agents (4). Both TBE and K–X were found to provide an adequate anesthetic depth with a similar Discussion percentage of surviving offspring, suggesting that there is no advan- There are a wide range of TBE concentrations and doses that have tage to using TBE over K–X in embryo transfer procedures. In been reported in the literature. TBE stock solution concentrations addition, TBE has been reported to reduce both blood pressure and that have been used range from 1:1 to 1:2 ratios of TBE to TAA (3, respirations (18), whereas the unfavorable cardiovascular effects pro- 5, 12). Reported TBE working solution concentrations vary from duced by xylazine can be reversed by yohimbine. Another advantage 1.2% to 2.5% (2-4, 12), and doses range from 240 mg/kg to 450 of K–X over TBE is that both ketamine and xylazine are known to mg/kg (2, 3, 12, 13). The concentrations chosen for the TBE stock provide analgesia, whereas the analgesic effects and mechanisms of and working solutions in the present study were the most common action for TBE are not known. In light of the anesthetic efficacy, reported in the literature (2, 4, 5, 12). The dose of 400 mg/kg that pathologic findings, and pharmacologic knowledge of the anesthetic we used was based on doses reported in the literature (2, 4, 6, 12) agents evaluated, we recommend the use of K–X over TBE and NaP and prior pilot studies (14). In these pilot studies, 375 mg/kg was for injectable anesthesia of female ICR mice.

20 CONTEMPORARY TOPICS © 2005 by the American Association for Laboratory Animal Science Volume 44, No. 1 / January 2005 It is recommended that TBE be stored in cold, dark conditions to avoid confounding effects from circadian rhythm. In experiments 2 prevent formation of the reported decomposition products, and 3, the average duration of anesthesia exceeded that of K–X in dibromoacetaldehyde (DBA) and hydrobromic acid (HBr) (3, 8, 18). experiment 1. These results mirror the varying anesthetic durations It has been theorized that the two compounds, the presence of which reported in the literature (2, 4) and remind us that bottle-to-bottle has been attributed to cause a drop in pH to 5.0 or below, are the variation can be a significant variable. Pathologic lesions induced by most likely cause for toxic effects associated with TBE (3, 13). In the TBE also were variable, as demonstrated by the large standard devia- first part of this study (9), TBE working solution was stored for 8 tions associated with pathology scores. A dose-dependent irritant effect weeks at 25°C in light and dark conditions and at 5°C in light and has been associated with the use of TBE (4), however all mice in the dark conditions. When TBE was stored at 5°C in the dark, pH re- current study received the same dose of TBE. Finally, TBE led to vari- mained at 6.5 throughout 8 weeks of storage, whereas a solution able results clinically. Mice in experiments 1 and 2 appeared healthy stored at 25°C light conditions had a pH of < 3.0 after 8 weeks. throughout the course of the study, whereas animals in experiment 3 Solutions stored at 5°C in light conditions and 25°C in dark condi- experienced significant morbidity and mortality associated with TBE tions had intermediate pH readings. From these pH findings, the administration, a finding that was verified in subsequent studies (11). most favorable storage condition was considered to be 5°C in the The variability of clinical and microscopic effects of TBE reported dark, and the least favorable storage condition was considered to be in the present study and in the literature may be attributed to several 25°C in the light (9). It should be noted that analysis of these solu- factors. First, there are several sources of commercial TBE powder, tions by using gas chromatography–mass spectrometry and NMR none of which are pharmaceutical grade products. In the first part of spectroscopy did not demonstrate a difference in DBA concentra- this study, we evaluated TBE powder from three different suppliers tions, regardless of the storage condition. in the United States. Analysis of these products demonstrated that In experiment 2, we evaluated how changes that occur during the TBE powder from different suppliers can vary in purity (9). It is also perceived most favorable and least favorable storage conditions af- plausible to expect bottle-to-bottle variation, even from the same fect anesthetic efficacy and short-term pathology when compared to supplier. This may explain some of the variations noted between newly prepared TBE. Anesthetic induction, duration of anesthesia, experiment 1 and experiments 2 and 3. The results of our verifica- and recovery times between the three groups were not significantly tion study subsequent to experiment 3 suggest that there may have different. Pathologic findings associated with stored TBE were not been a change in stored TBE powder (11). This is not surprising, as significantly different, regardless of storage condition, when com- some material safety data sheets (22) recommend that TBE is stored pared with those of newly prepared TBE. The recommendation to and used under nitrogen. Substances often are manipulated under monitor pH as an indirect measure of safety for TBE was not sup- nitrogen when known or thought to be susceptible to an oxidation ported by this study, nor was a difference in efficacy noted to relate reaction. Therefore, it is plausible that a change may occur in TBE to storage condition. powder if the bottle is opened, exposed to air, and then stored. Experiment 3 was performed to determine whether there are long- In conclusion, anesthetic induction times for K–X and TBE were term affects associated with the administration of newly prepared rapid and consistent, however the anesthetic duration produced by TBE. During this experiment, we had significant morbidity and TBE was variable. This variability may reflect bottle-to-bottle varia- mortality associated with the administration of newly prepared TBE. tion or changes that occur to the TBE powder during storage. TBE Pathologic findings in moribund animals were similar to the adverse administration caused significantly more pathology when compared effects of TBE that have been reported in the literature (4-6), with with pharmaceutical-grade anesthetic agents that are available for gross findings consistent with gastrointestinal ileus. The overall patho- use in mice and resulted in significant morbidity and mortality. In logic findings suggest that the cause of death was a chemical peritonitis addition, the monitoring of pH does not necessarily correlate to the with bacterial translocation to the peritoneum, which in turn led to presence of toxic components in or the potential lethality of a TBE septicemia and pneumonia. working solution. Because of the variability of anesthetic effective- The starting TBE powder used in experiment 3 was from the same ness, pathology, and morbidity and mortality associated with the use bottle as that used for experiment 2 and was stored in the original of TBE, we do not recommend the use of this anesthetic agent in amber bottle at 5°C after each use. The preparation of the stock and female ICR mice. There are pharmaceutical-grade anesthetics avail- working solutions was identical for all experiments and included sterile able that appear to provide more consistent and reliable anesthesia, filtration. The subsequent study, which was done to verify the re- fewer pathologic changes, and less morbidity and mortality. sults seen in experiment 3, used the same TBE powder and stock solution used in experiment 3. Similar morbidity rates, as well as Acknowledgments pathologic findings, were seen. These findings suggest that a solid- We would like to thank the ACLAM Foundation for supporting this phase change occurred during the storage of TBE powder in the study. We also thank Robert Anderson, PhD, for performing statistical analy- time between experiments 2 and 3. The stock and working solutions ses and Tara G. Ooms, DVM, for her technical assistance. that resulted in mortality were analyzed via pH strips, NMR, and gas chromatography–mass spectrometry and compared with prior analysis done for non-lethal stock and working solutions. The pH of References the lethal solution was 6.5, and we did not find a measurable in- 1. Anonymous. 1989. Fifty-eight years ago in anesthesia and analgesia. Anesth. Analg. 68:540. crease in the quantity of DBA, which was previously reported to 2. Gardner, D. J., J. A. Davis, P. J. Weina, et al. 1995. Comparison of cause adverse affects (9). tribromoethanol, ketamine/acetylpromazine, Telazol/xylazine, pento- Although TBE reliably induced anesthesia over the course of the barbital, and methoxyflurane anesthesia in HSD:ICR mice. Lab. Anim. three experiments, it is of interest that the duration of anesthesia was Sci. 45:199-204. variable, with durations of 18.5, 34.4, and 46.5 min. Circadian 3. Papaioannou, V. E. and J. G. Fox. 1993. Efficacy of tribromoethanol rhythm has been shown to affect anesthetic duration in rodents (19- anesthesia in mice. Lab. Anim. Sci. 43:189-192. 21), however all three experimental trials were started and finished 4. Zeller, W., G. Meier, K. Bürki, et al. 1998. Adverse effects of tribro- at approximately the same times (8:30 a.m. to 3:30 p.m.). Treat- moethanol as used in the production of transgenic mice. Lab. Anim. 32:407-413. ment groups were randomized by cage and by agent administered to

Volume 44, No. 1 / January 2005 CONTEMPORARY TOPICS © 2005 by the American Association for Laboratory Animal Science 21 5. Green, C. J. 1975. Neuroleptanalgesic drug combinations in the an- 14. Rodriguez, N. A., C. C. Lieggi, J. D. Fortman, et al. 2003. Unpub- aesthetic management of small laboratory animals. Lab. Anim. lished data. 9:161-178. 15. Swindle, M. M., G. A. Vogler, L. K. Fulton, et al. 2002. Preanesthesia, 6. Tarin, D. and A. Sturdee. 1972. Surgical anaesthesia of mice: evalua- anesthesia, analgesia, and euthanasia, p. 957-959. In J.G. Fox, L. C. tion of tribromo-ethanol, ether, halothane and methoxyflurane and Anderson, F. M. Loew, et al. (ed.), Laboratory animal medicine, 2nd development of a reliable technique. Lab. Anim. 6:79-84. ed. Academic Press, San Diego, Calif. 7. Reid, W. C., K. P. Carmichael, S. Srinivas, et al. 1999. Pathologic 16. Thompson, J. S., S. A. Brown, V. Khurdayan, et al. 2002. Early ef- changes associated with use of tribromoethanol (Avertin) in the Sprague fects of tribromothanol, ketamine/xylazine, pentobarbital, and isoflurane Dawley rat. Lab. Anim. Sci. 49:665-667. anesthesia on hepatic and lymphoid tissue in ICR mice. Comp. Med. 8. Norris, M. L. and W. D. Turner. 1983. An evaluation of tribromoet- 52:63-67. hanol (TBE) as an anaesthetic agent in the Mongolian gerbil (Meriones 17. Huang, G. and K. K. Linask. 1998. Doppler echocardiographic analysis unguiculatus). Lab. Anim. 17:324-329. of effects of tribromoethanol anesthesia on cardiac function in the mouse 9. Lieggi, C. C., J. D. Fortman, R. Kleps, et al. 2004. An evaluation of embryo: a comparison with pentobarbital. Lab. Anim. Sci. 48:206-209. preparation methods and storage conditions of tribromoethanol. 18. Lumb, W. V. 1963. Small animal anesthesia, p. 201. Lea & Febiger, Contemp. Top. Lab. Anim. Sci. 44(1):11-16. Philadelphia. 10. National Research Council. 1996. Guide for the care and use of labo- 19. Sato, Y., E. Kobayashi, Y. Hakamata, et al. 2004. Chronopharmacological ratory animals. National Academy Press, Washington, D.C. studies of ketamine in normal and NMDA ˜1 receptor knockout mice. 11. Lieggi, C. C., J. D. Fortman, and J. E. Artwhol. 2003. Unpublished Br. J. Anaesth. 92(6):859-864. data. 20. Reinberg, A. 1986. Circadian rhythms in effects of hypnotics and sleep 12. Hogan, B., R. Beddington, F. Costantini, et al. 1994. Manipulating inducers. Int. J. Clin. Pharmacol. Res. 6(1):33-44. the mouse embryo: a laboratory manual, 2nd ed. Cold Spring Harbor 21. Holcslaw, T. L. 1978. Circadian rhythm in hexobarbital sleep time: a Laboratory Press. dose–response evaluation in the mouse. 22(1):49-60. 13. Nicol, T., B. Vernon-Roberts, and D. C. Quantock. 1965. Protective 22. Aldrich Chemical Co., Inc. 2001. Material safety data sheet for prod- effect of oestrogens against the toxic decomposition products of tribro- uct number T48402: product name, 2,2,2-tribromoethanol (99%), moethanol. Nature 208(15):1098-1099. section 3. Aldrich Chemical Co., Inc., Milwaukee, Wis.