Cardiovascular Pathology Possibly Associated with Ketamine/Xylazine Anesthesia in Dutch Belted Rabbits

Robert P. Marini,1 Xiantang Li,2 Neil K Harpster,3 and Charles Dangler1

Background and Purpose: After myocardial necrosis and fibrosis was observed in five rabbits which had been anesthetized a variable number of times, the potential relationship of these lesions and anesthesia was evaluated in 35 other rabbits. Methods: Anesthesia was induced by intramuscular administration of ketamine and xylazine followed by infusion of lactated Ringer’s solution also containing ketamine and xylazine. Group A rabbits (n = 9) were subjected to multiple anesthesias and were evaluated by echocardiography, thoracic radiography, electrocardiography, determination of serum coronavirus titer, vitamin E concentration, and complete necropsy. Prior to a single acute procedure followed by necropsy, group B rabbits (n = 11) were evaluated by echocardiography only. Group C rabbits (n = 10) had never been anesthetized and were necropsied after euthanasia. Group D rabbits (n = 5) had intermediate anesthesia exposure history and were evaluated by echocardiography only. Myocardial fibrosis was scored semi-quantitatively on a scale of 0 to 4. Results: Canine coronavirus test results were negative; hypovitaminosis E was evident, and fibrosis scores were significantly increased in group A, compared with group B or group C, rabbits. ␣ Conclusion: Etiologic differentials included 2-agonist-mediated coronary vasoconstriction with associated myocardial hypoperfusion, hypovitaminosis E and free radical injury, and other anesthetic-induced physiologic trespass.

Spontaneous myocardial disease in rabbits is infrequently tion—have not been reported in the rabbit. documented. Specific causes for myocardial disease include The cases motivating this study comprised five Dutch hypovitaminosis E, rabbit coronavirus infection, salmonello- belted rabbits that were observed over a 12-month period. sis, pasteurellosis, encephalitozoonosis, and administration Clinical signs of disease included sudden death in three rab- ␣ of anesthetic combinations that contain the 2-agonist seda- bits and tachypnea and anorexia in another. One rabbit tive detomidine (1–6). Heart failure or injury has been in- without clinical signs of disease was euthanized for an unre- duced experimentally through rapid ventricular pacing (7); lated cause. These rabbits were part of a cohort of 21 ani- inoculation of Trypanosoma cruzi (8); surgical induction of mals that had been used in an ophthalmology study and had aortic regurgitation (9); left ventricular catheterization with been anesthetized a variable number of times in electroret- subsequent intravenous inoculation of Streptococcus inography and visual-evoked potential studies. For each an- viridans (10); alloxan-induced diabetes (11); gamma irradia- esthesia episode, anesthesia had been induced by ketamine tion (12, 13); ingestion of coffee senna (Cassia occidentalis) (50 mg/kg, intramuscularly [i.m.]) and xylazine (10 mg/kg, (14, 15); and administration of catecholamines (16–19), 6- i.m.) and subsequently maintained by use of a constant-rate hydroxydopamine (20), emetine (21), cantharidin (22), and intravenous infusion ([6 ml/h] of a 1:1 mixture of ketamine the anthracycline antibiotics daunorubicin, detorubicin, and [5 ml, 100 mg/ml] and xylazine [5 ml, 20 mg/ml]) brought to doxorubicin (23, 24). Congenital heart defects have been re- a total volume of 60 ml with lactated Ringer’s solution (final ported sporadically and include ventricular septal defect concentration, 8.3 mg of ketamine and 1.7 mg of xylazine/ and hemocyst (25). A stress-induced episode of cardiomyopa- ml). Rabbits had been anesthetized three to eight times and thy and sudden death was reported in rabbits in association had received 600 to 1,400 mg of ketamine and 120 to 400 mg, with high housing density (26). Experimentally induced Her- cumulatively, of xylazine. One rabbit was used in a short- pesvirus sylvilagus infection in cottontail rabbits causes term procedure, then was submitted to necropsy; for three lymphocytic myocarditis, necrosis of cardiac myocytes, and other rabbits, 3 to 12 weeks elapsed between the last episode eventual myocardial fibrosis (27). To the authors’ knowledge, of anesthesia and death or euthanasia. Histopathologic find- other spontaneous causes of heart disease—such as taurine ings included moderate to extensive myocardial necrosis and deficiency, toxoplasmosis, and Clostridium piliforme infec- fibrosis (four rabbits), and mild multifocal myofiber hypertrophy with interstitial fibrosis (one rabbit). Four rabbits Division of Comparative Medicine, Massachusetts Institute of Technology, had moderate to severe, acute diffuse interstitial pneumoni- Cambridge, Massachusetts1; Animal Resources Center, University of Chi- tis characterized by high numbers of alveolar macrophages; cago, Chicago, Illinois2; Angell Memorial Hospital, Boston, Massachusetts3

153 Vol 49, No 2 Laboratory Animal Science April 1999 exudation of eosinophils and fibrin; hypercellular, thick al- those of group B in that the latter were ketamine/xylazine- veolar septa; and foci of septal necrosis. Interstitial edema naive at the time of echocardiography. occurred in the interlobular stroma and large arteries, with Rabbits received only ketamine/xylazine systemically; multifocal vasculitis and perivasculitis affecting large- and ophthalmic atropine, phenylephrine, and tropicamide were medium-caliber arteries and veins. The medium- and large- delivered topically to one eye. Animals breathed room air caliber arteries were characterized by medial hypertrophy spontaneously and were kept on circulating hot-water blan- and a few multifocal intimal plaques. The remaining rabbit kets; respiration was continuously monitored. For rabbits of had prominent medial hypertrophy of medium- to large-cali- group A, biomaterials were implanted into the eye at the sec- ber arteries and mild, multifocal periarteritis. Three rabbits ond anesthesia episode to generate compatibility informa- had pleural effusion, one of which also had a diaphragmatic tion before developing a retinal prosthesis. hernia. Echocardiography: Rabbits were sedated with diaz- Myocardial fibrosis had been observed previously by us in epam (1 mg/kg, intravenously [i.v.]), then restrained manu- New Zealand White rabbits given anesthetic combinations ally in right or left lateral recumbency. Seven rabbits from ␣ containing the 2-agonist detomidine (5). A report of group A (six males, one female), six rabbits from group B (six ketamine/xylazine-associated mortality in New Zealand females), and five rabbits from group D (five females) were White rabbits implicates acidosis as the etiologic factor (28). used to determine differences in echocardiographic profiles In that study, however, rabbits were given ketamine/ attributable to ketamine/xylazine exposure. Echo-cardiogra- xylazine after recovery from lidocaine-induced epidural an- phy was performed, using a 7.5-MHz transducer and an ATL esthesia; necropsy results were not reported. We sought to Apogee Model CX (Bothell, Wash.) echocardiograph. correlate the lesions in these rabbits with repeated anesthe- Serum IgG coronavirus titers: Pleural effusion dis- sia episodes of ketamine/xylazine, and attempted to identify ease/infectious cardiomyopathy virus status was evaluated cardiac dysfunction through the noninvasive diagnostic by examining serum from seven group A rabbits, using an techniques of echocardiography, electrocardiography, and indirect fluorescent antibody (IFA) test, for antibody to the thoracic radiography. cross-reacting canine coronavirus (Washington Animal Dis- ease Diagnostic Laboratory, College of Veterinary Medicine, Materials and Methods Washington State University, Pullman, Wash.). Animals: Thirty-five specific-pathogen-free Dutch belted Vitamin E concentrations: Vitamin E concentrations in rabbits (Oryctolagus cuniculus), Hra:(DB) Covance (formerly serum from the five rabbits in group A with the most extensive Hazleton Research Products Inc., Denver, Pa.) were studied. cardiac lesions, as determined by histologic examination, were Twenty-five of the rabbits were individually housed in stain- determined by using normal-phase high-performance liquid less steel cages (24 x 30 x 17 in) with slotted floors and re- chromatography (Animal Health Diagnostic Laboratory, Col- ceived water and a commercial rabbit chow (Purina rabbit lege of Veterinary Medicine, Michigan State University, chow HF326; Ralston Purina Co., St. Louis, Mo.) ad libitum, Lansing, Mich.). in a facility that has animal care and use programs accred- Radiography and electrocardiography: Right lateral ited by AAALAC, International. The environment was con- and ventrodorsal radiographs were obtained from seven rab- trolled for temperature (15.5 to 20.5ЊC), relative humidity bits of group A. These rabbits were sedated with diazepam (1 (40 to 65%), light (12/12-h light/dark cycle), and ventilation mg/kg, i.v.) and restrained in right lateral recumbency. A six- (100% fresh air with 12 air changes/h). Ten rabbits were lead electrocardiogram was obtained, using a PhysioControl housed by the vendor until euthanasia. monitor/recorder (Life Pak 5S; PhysioControl, Redmond, Rabbits were separated into four groups on the basis of Wash.). After electrocardiography, rabbits were anesthetized their exposure to ketamine/xylazine. Group A rabbits (n = 9; with pentobarbital (30 mg/ml solution injected slowly i.v. to one female, eight males; mean age, 21 months) had been effect) for thoracic radiography. Right lateral and anesthetized 8 to 10 times over a period of 1 year to evaluate ventrodorsal thoracic radiographs were obtained. After ra- serial electroretinograms. Rabbits received approximately diography was completed, rabbits were euthanized immedi- 175 mg of ketamine and 35 mg of xylazine over 90 to 120 ately by overdose of pentobarbital (120 mg/kg, i.v.). min in each anesthesia episode; cumulative doses ap- Pathologic examination: Body weight was recorded, proached 1,300 to 1,800 mg of ketamine and 300 to 470 mg and complete necropsy was performed on all rabbits of of xylazine. Group B rabbits (n = 11; nine females, two groups A and B. Heart weight was recorded after trimming males; mean age, 7 months) had been anesthetized once for the aorta at the level of the brachiocephalic artery, pulmo- a terminal procedure and had received cumulative doses of nary arteries and veins, and cranial and caudal vena cavas 350 to 600 mg of ketamine and 70 to 170 mg of xylazine. at the base of the heart for each animal. Body and heart Group C rabbits (n = 10; five females, five males; mean age, weights were not obtained for animals in group C. The heart 30 months) were retired breeders that had never been anes- and other tissues were fixed in neutral-buffered 10% forma- lin. After overnight fixation, the heart was cut transversely thetized. After euthanasia by CO2 inhalation by the vendor, the heart was excised, placed in neutral-buffered 10% for- immediately subjacent to the atrioventricular valves. At this malin, and shipped to our laboratory. Group D rabbits (five level, the following dimensions were obtained: right ven- females; mean age, 20 months) had been anesthetized one to tricular free-wall thickness, interventricular septum thick- three times with ketamine/xylazine and were used only for ness, left ventricular free-wall thickness, right and left echocardiographic evaluation. Group D rabbits differed from ventricular chamber diameters, and total heart width.

154 Rabbit Heart Disease

Heart and other tissue specimens, including liver, lung, Table 1. Echocardiographic variables in Dutch belted rabbits spleen, gonads, and skeletal muscles of multiple sites Groups (tongue, triceps, quadriceps, diaphragm), and abdominal AB D Multiple Single (terminal) Intermediate wall were embedded in paraffin, cut at 5-␮m thickness, Echocardiographic anesthesia anesthesia anesthesia and stained with hematoxylin and eosin for histologic variables n = 9 n = 11 n = 5 evaluation. LVEDD 1.01 Ϯ 0.12 1.17 Ϯ 0.19 1.18 Ϯ 0.13 Ϯ Ϯ Ϯ Heart sections were also stained with trichrome and oil LVESD 0.61 0.09 0.70 0.09 0.74 0.15 %∆ D 39.29 Ϯ 4.35 39.50 Ϯ 5.39 37.60 Ϯ 7.44 red O to document fibrosis and fatty change, respectively. LVETa 0.09 Ϯ 0.01 0.08 Ϯ 0.01 0.11 Ϯ 0.00 The degree of myocardial fibrosis was assessed for the VCFb 4.20 Ϯ 0.59 4.74 Ϯ 0.45 3.42 Ϯ 0.66 Ej Fract 75.86 Ϯ 7.03 76.17 Ϯ 5.81 74.20 Ϯ 8.47 right ventricular wall, left ventricular wall, and ventricu- LVFW Th d 0.34 Ϯ 0.08 0.31 Ϯ 0.08 0.30 Ϯ 0.12 lar septum. The degree of myocardial fibrosis was scored IVS Th d 0.37 Ϯ 0.11 0.25 Ϯ 0.05 0.26 Ϯ 0.05 semiquantitatively as 0 (none) if fibrotic changes were not LAD s 0.29 Ϯ 0.49 0.17 Ϯ 0.41 0.40 Ϯ 0.55 AoD d 0.73 Ϯ 0.10 0.67 Ϯ 0.10 0.66 Ϯ 0.09 seen, 1 (minimal) if <25%, 2 (mild) if 25 to 50%, 3 (moder- LA/Ao 1.20 Ϯ 0.27 1.38 Ϯ 0.32 1.41 Ϯ 0.21 ate) if 50 to 75%, and 4 (marked) if >75% of the ventricu- RAD s 0.62 Ϯ 0.05 0.61 Ϯ 0.08 0.70 Ϯ 0.08 Ϯ Ϯ Ϯ lar walls or septum were affected. Hearts from group C AoD d 0.66 0.04 0.69 0.10 0.71 0.08 RA/Ao 0.93 Ϯ 0.10 0.88 Ϯ 0.17 0.96 Ϯ 0.11 rabbits were treated as described previously and were de- EPSS 0.10 Ϯ 0.12 0.05 Ϯ 0.05 0.06 Ϯ 0.09 livered by the vendor. DE Excurs 0.54 Ϯ 0.08 0.55 Ϯ 0.08 0.50 Ϯ 0.10 EF Slope 66.57 Ϯ 26.68 70.17 Ϯ 31.82 77.40 Ϯ 13.79 Statistical evaluation: Echocardiographic variables RVOT Vel 0.84 Ϯ 0.10 0.83 Ϯ 0.10 0.74 Ϯ 0.11 were evaluated by use of one-way ANOVA, followed by LVOT Vel 0.70 Ϯ 0.11 0.65 Ϯ 0.14 0.77 Ϯ 0.16 pairwise comparisons among means, using the Newman- aGroups A, B, and D are significantly different (P Յ 0.05). Keuls test. Heart weight and dimensions were compared by bGroups A and B differ significantly from group D (P Յ 0.05). LVEDD = left ventricular end diastolic dimension, LVESD = left ventricu- using a two-tailed t test (Excel, Version 5.0; Microsoft, lar end systolic dimension, %∆ D = shortening fraction , LVET = left ven- Redmond, Wash.). Pair-wise comparisons of histopathologic tricular ejection time, VCF = velocity of circumferential fiber shortening, scores were evaluated by using a one-tailed Mann-Whitney Ej Fract = ejection fraction, LVFW Th d = left ventricular free wall thickness diastolic, IVS Th d = interventricular septum thickness diastolic, test. A P value < 0.05 was considered significant. LAD s = left atrial dimension systolic, AoD d = aortic root dimension diastolic, LA/Ao = left atrial-to-aortic root ratio, RAD s = right atrial dimension systolic, AoD d = aortic root dimension diastolic, RA/Ao = right Results atrial-to-aortic ratio, EPSS = E-point septal separation, DE excurs = DE Echocardiography: Results are summarized in Table 1; excursion, EF Slope = E-to-F slope, RVOT Vel = right ventricular outflow significant differences were found for two variables. For left tract velocity, LVOT Vel = left ventricular outflow tract velocity. Data are expressed as mean Ϯ SD. ventricular ejection time, in descending order by time, group D was longer than group A, which was longer than group C of either group A (P < 0.01 and < 0.05, respectively) or group (P < 0.05). For velocity of circumferential fiber shortening, values for groups A and B were greater than those for group B (P < 0.001 and P < 0.01, respectively). The right ventricular chamber diameter of group A rabbits was also greater D (P < 0.05). Values for each parameter may have been than that of group C rabbits (P < 0.02). slightly influenced by diazepam tranquilization. Serum IgG coronavirus titers: Serum samples from all Pathologic findings: Gross abnormalities were not seen in hearts from rabbits of groups B and C. Irregular pale or rabbits of group A were test negative by IFA for cross-react- white streaks were seen in the ventricular septa and/or left ing antibodies to canine coronavirus (sera from animals in groups B, C, and D were not tested). ventricular wall of two rabbits from group A. The most prominent histologic change was myocardial fibrosis in Serum vitamin E concentrations: Hemolysis of serum group A rabbits (Figures 1 and 2). Minimal to mild myocar- from two of the five Dutch rabbits originally submitted pre- cluded interpretation of vitamin E concentrations. Mean Ϯ dial fibrosis was also seen in 2 of 10 group C and 1 of 11 group B rabbits. Mild to moderate myocardial fibrosis was SD serum vitamin E concentrations for first submissions seen in 8 of 9 group A rabbits (Figures 2a–c); one rabbit had from three rabbits in group A was 1.85 Ϯ 0.49 ␮g/ml. Electrocardiography and radiography: There were minimal myocardial fibrosis. In mild cases, collagen bundles formed delicate networks and coursed around myofibers, as no abnormalities in rhythm or rate for any of the rabbits revealed by trichrome staining (Figure 2a). In moderate that were evaluated. One rabbit that had been anesthetized a number of times had evidence of right ventricular hyper- cases, collagen bundles formed dense fibrotic plaques and radiated into the adjacent myofibers (Figure 2b). Many of trophy as manifested by the height of the P and R waves and the fibrotic foci were around small arterioles and/or veins the presence of deep S waves in leads I, II, III, and AVF. Ra- diographic abnormalities were not detected. (Figures 2a and b). There was marked variation in size and shape and cytoplasmic vacuolation of the myofibers within Cardiac dimensions and cardiac weight: Postmortem and/or adjacent to the fibrotic foci (Figures 1c and 2c). Sig- cardiac dimensions and cardiac and whole body weights for all rabbits of groups A and B are recorded in Table 2. Signifi- nificant differences in the left ventricle were observed between groups A and B (A > B, P = 0.01) and between groups cant differences were detected between groups A and B for A and C for the left ventricle (A > C, P = 0.005), and inter- body weight (B > A; P = 0.03). Although mean cardiac weight was greater for group A rabbits, the difference was not sig- ventricular septum (A > C, P = 0.004). Significantly, more fibrosis was seen between the left ventricular free wall and nificant. Left ventricular wall and interventricular septum interventricular septum, compared with the right ventricu- thickness was greater in rabbits of group C than in rabbits

155 Vol 49, No 2 Laboratory Animal Science April 1999

Table 2. Cardiac dimensions (mm) and body and heart weights (g) Table 3. Distribution of scores for myocardial fibrosis Groups Left ventricular wall scores ABC Variable n = 9 n = 11 n = 10 Group N 0 1 2 3 4 Aa 902250 Body weight (g) 2,025.7 Ϯ 116.3 2,320.6 Ϯ 365.8 ND B 11100100 Heart weight (g) 8.6 Ϯ 1.1 9.1 Ϯ 1.1 ND C 811000 Right ventricular wall 1.44 Ϯ 0.30 1.63 Ϯ 0.56 1.50 Ϯ 0.53 Interventricular septum 4.33 Ϯ 0.71 3.95 Ϯ 0.47 4.95 Ϯ 0.80a Interventricular septum scores Left ventricular wall 4.72 Ϯ 0.67 4.24 Ϯ 0.64 6.10 Ϯ 1.26b Right ventricular Group N 0 1 2 3 4 chamber diameter 7.83 Ϯ 1.48c 7.00 Ϯ 2.09 5.95 Ϯ 1.50 Ab 910620 Left ventricular B 1191100 chamber diameter 6.17 Ϯ 2.22 5.41 Ϯ 2.30 6.60 Ϯ 2.91 C 10100000 Total cardiac dimensions 23.72 Ϯ 1.42 24.55 Ϯ 2.30 24.00 Ϯ 2.98 Right ventricular wall scores aC > A (P < 0.05) or B (P < 0.01). bC > A (P < 0.01) or B (P < 0.001). Group N 0 1 2 3 4 cA > C (P < 0.02). A 918000 B 11101000 C 10100000 aA > B P = 0.01; A > C P = 0.005. bA > C P = 0.004.

Figure 1. Photomicrographs of sections of rabbit heart. (a) Notice myocardial necrosis (N) with mixed inflammatory cell infiltration between (solid arrow) and within (open arrow) myofibers. H&E stain; bar = 43 ␮m. (b) Notice coagulative necrosis (N) and vacuolar degeneration (D) of myofibers of the papillary muscle. H&E stain; bar = 87 ␮m. (c) Notice marked vacuolar degeneration of myofibers (arrows) within or adjacent to fibrotic foci (F), with marked variation in myofiber diameters. H&E stain, bar = 43 ␮m. (d) Notice degenerative myofibers containing fat droplets (arrows). Oil red O stain; bar = 43 ␮m.

lar free wall (P < 0.003 and P < 0.05, respectively). eration and necrosis (Figure 1a). Necrotic fibers were char- Myocarditis and myofiber degeneration and necrosis were acterized mainly by cytoplasmic hypereosinophilia, hyalin- seen principally in the rabbits with one episode of anesthe- ization, fragmentation, and nuclear pyknosis and sia (Figure 1). Prominent myocarditis characterized by karyorrhexis (Figure 1b). Degenerated myofibers were char- mononuclear cell and heterophil infiltration within and/or acterized mainly by their cytoplasmic vacuolization and between the myofibers was seen in rabbits that had been variations in myofiber size and shape (Figure 1c). Many of anesthetized once, with variable degrees of myofiber degen- the degenerative myofibers contained fat droplets as re-

156 Rabbit Heart Disease

Figure 2. Photomicrograph of sections of rabbit heart. (a) Notice collagen bundles forming networks between the myofibers (arrows). Trichrome stain; bar = 87 ␮m. (b) Notice marked fibrosis (F) around the small arterioles and/or veins (arrows) and the adjacent myofibers. H&E stain, bar = 87 ␮m. (c) Notice marked variation in size and shape of the myofibers (arrows) within the fibrotic foci (F). H&E stain; bar = 43 ␮m. (d) Notice intense infiltrate of eosinophilic granulocytes and lymphocytes enveloping a medium-caliber pulmonary artery. The intimal layer is thick, and granulocytes are concentrated within the lumen and on the endothelium. H&E stain; bar = 43 ␮m. vealed by oil red O stain (Figure 1d). thickening was frequently seen in the myocardial vascula- The pulmonary vasculature of rabbits in group A (six rab- ture and was typically located in the left cranial descending bits) and group B (five rabbits) frequently had histopatho- artery of the left ventricle. A pathogenetic relation between logic alterations, principally centered on the arterial the eosinophilic infiltrate and the other arterial lesions is branches. The stroma encircling the large pulmonary arter- not obvious at this time. ies and veins often had mild to marked interstitial edema. The periarterial stroma in the lungs often contained multifo- Discussion cal infiltrates of eosinophils accompanied by variable num- This study does not definitively establish a relationship bers of lymphocytes. Arteries of all sizes, including the vasa between anesthesia and the pathologic changes observed, vasorum of large arteries, were affected. Occasionally, focal nor does it identify the nature of a potential relation. A phar- adherence and infiltration of eosinophils at the arterial in- macologic basis is suggested, however, by similar lesions tima were evident. In one instance, eosinophilic infiltration documented in New Zealand White rabbits given the into the medial layer of a pulmonary artery was observed, xylazine-like agent detomidine. Hurley and coworkers (5) associated with marked edema of this arterial segment. The ␣ evaluated the 2-agonist drug detomidine, alone and in com- smooth muscle cells within the medial layer of some me- bination with ketamine or diazepam, and found myocardial dium and large pulmonary arteries were separated by necrosis and fibrosis in five of six rabbits. They theorized poorly stained extracellular matrix or edema. A few of the that the combination of reduction in coronary flow reserve medium- and large-caliber pulmonary arteries had variable and impairment of coronary blood flow caused myocardial degrees of medial hypertrophy. In general, the more intense ischemia with subsequent necrosis. The former is the likely lesions of eosinophilic periarteritis and endarteritis were consequence of the hypoxemia associated with ketamine/ observed among the group A rabbits. Nonocclusive, multifo- xylazine administration in the rabbit. The latter is predicted cal fibrous intimal plaques were observed in the large- and ␣ by the interaction of xylazine with 2-receptors in coronary medium-caliber arteries of many rabbits. Focal intimal vessels. In an excised coronary artery preparation from dogs,

157 Vol 49, No 2 Laboratory Animal Science April 1999 xylazine and medetomidine were found to cause dose-depen- reduction in heart rate would result in an increased pressure- dent coronary artery contraction that was more pronounced rate product. Centrally mediated decrease in sympathetic out- ␣ in smaller vessels (29). Repeated episodes of hypoxemia, and flow attributed to 2-agonists may also have been inadequate xylazine-induced coronary vasoconstriction and initial hy- to spare the myocardium depending on the kinetics of drug re- pertension, presumably preclude satisfaction of the myocar- ceptor interactions at the level of the coronary vasculature. Al- dial oxygen demand and lead to cell death and necrosis. The though to our knowledge the initial hypertension induced by rabbit is a species with limited collateral circulation in the detomidine when administered to rabbits has never been re- myocardium and is therefore predisposed to ischemia in- ported for xylazine, coronary vasoconstriction might still occur duced by coronary vasoconstriction (30, 31). in the absence of systemic hypertension. An analogous situation may exist with the rabbit model of Repeated bouts of hypotension and hypoxemia in rabbits catecholamine-induced cardiomyopathy in which alpha adr- anesthetized repetitively with ketamine/xylazine may be re- energic-mediated coronary vasoconstriction in the face of in- sponsible for lesions of myocardial necrosis and subsequent ␣ creased myocardial work may deplete myocardial energy fibrosis independent of 2-agonist-mediated coronary vaso- stores, leading to left ventricular dysfunction (17–19, 32). constriction. The hypotension is due to the depressive effect ␣ Results of radioactive microsphere studies have indicated of xylazine on sympathetic tone via presynaptic 2-receptors, that norepinephrine-induced cardiomyopathy is associated whereas hypoxemia presumably results from a combination with reduced coronary blood flow and increased coronary re- of ketamine/xylazine-induced hypotension and reduction in sistance (18). Systemic absorption of topical phenylephrine respiratory rate (5). Hypotension due to administration of in rabbits of this study might have contributed to or have the long-lasting vasodilator minoxidil to rats for 2 days in- been solely responsible for coronary vasoconstriction. In the duces myocardial hemorrhage, arteritis, and necrosis that is case of putative xylazine-induced vasoconstriction, interac- more severe with age and is believed to result from hypoten- ␣ tion of the xylazine with either postjunctional vascular 2- sion and failure to meet myocardial oxygen demand (41). ␣ ␣ or both 1- and 2-adrenergic receptors may be involved in Additional mechanisms that may be of importance include vasoconstriction. Both of these receptor types mediate vaso- myocardial stunning, reperfusion injury (42), oxidative dam- constriction at this site. Xylazine is approximately 100 times age from reactive intermediates (43–46), and xylazine-in- ␣ ␣ more selective for the 2- than the 1-receptor (33). At high duced endothelial toxicosis (47, 48). Alternatively, coronary ␣ doses, however, one could envision saturation of the 2- with vasoconstriction could act in synergy with other forms of he- ␣ subsequent interaction with 1-adrenergic receptors. This modynamic change induced by ketamine/xylazine. situation might be more likely to arise with constant-rate in- Age and sex associations of these lesions are not entirely fusions of xylazine, leading to effects that would not be seen clear from this study due to limitations in rabbit availability. in the usual clinical setting of one or two intramuscular in- Specifically lacking were age- and sex-controlled animals jections. The existence of coronary vasoconstriction and the that had not been anesthetized and were housed at our in- relative contribution of xylazine and phenylephrine could stitution. Drug disposition, metabolism and excretion, bio- not be evaluated in this study. Using specific adrenergic re- chemical change in myocardial function, and changes in ceptor antagonists as premedicants in ketamine/xylazine- myocardial mass relative to coronary vascularization are anesthetized Dutch rabbits might help determine the mechanisms by which age may influence the effect of drugs involvement of these anesthetics in cardiomyopathy. on the heart (41). Age-matched control hearts from the ven- The dosages used for rabbit anesthesia in the studies de- dor, while controlled for age-related and cardiotoxic effects of scribed here were well within the published range for intra- drugs, were not appropriately controlled for diet or other en- muscular use in rabbits. When used by this route, ketamine/ vironmental influences. Additionally, a heritable component xylazine reportedly causes marked cardiopulmonary depres- for this lesion could not be determined through evaluation of sion, including reductions from baseline of 40 to 77% for res- records maintained by the breeder. piratory rate, 43 to 50% for PaO2, 35% for heart rate, 20 to The possibility that hypovitaminosis E was solely respon- 35% for blood pressure, and an increase in PaCO2 of 25 to sible for these lesions, although unlikely, cannot be defini- 50% (34–40). In one report, an initial intramuscularly ad- tively excluded as contributory. Although the Dutch rabbits ministered dosage of ketamine/xylazine (35 mg/kg; 5 mg/kg) in this study had serum concentrations of vitamin E below was followed by a 4-h, constant-rate intravenous infusion of the reference laboratory normal values, there were no skel- 1 mg of ketamine and 0.1 mg of xylazine/min (40). This regi- etal muscle lesions typically associated with nutritional men caused progressive reduction in blood pressure of 49% muscular dystrophy (49–58). This suggests that hypovitami- from baseline values; arterial O2 tension decreased to 45% of nosis was not the cause of myocardial lesions. Alternatively, baseline, but then progressively increased. Necropsy find- myocardial injury from either a direct toxic effect of anesthe- ings were not reported in that study. In comparison, the dos- sia or anesthetic-induced physiologic trespass may have led ages used for our rabbits were 50 mg of ketamine and 10 mg to generation of free radicals and subsequent lipid of xylazine/kg, followed by a constant-rate infusion of 0.83 peroxidation. Oxygen-derived free radical formation may mg of ketamine and 0.166 mg of xylazine/min. Although the markedly increase after reperfusion of a formerly ischemic dosages are comparable, the margin of safety for peripheral myocardium. Marked increase in oxidative stress may have autonomic effects of xylazine have not been determined for been potentiated by diminished vitamin E concentration in the rabbit and may have been exceeded. Initial hypertension these rabbits. ␣ associated with 2-agonist administration without concomitant Presence of myocardial fibrosis did not uniformly result in

158 Rabbit Heart Disease

clinical signs of disease. None of the rabbits in group A exhib- 8. Rossi, M. A. 1990. Microvascular changes as a cause of chronic ited signs consistent with heart failure. Group B rabbits were cardiomyopathy in Chagas’ disease. Am. Heart J. 120:233–236. heavier; this may have reflected episodes of partial 9. Magid, N. M., G. Opio, D. C. Wallerson, et al. 1994. Heart failure due to chronic experimental aortic regurgitation. Am. J. inappetance in group A animals. Echocardiographic indices of Physiol. 267:556–562. cardiac performance did not differ markedly between the 10. Tomlinson, C. W., and N. S. Dhalla. 1976. Alterations in myo- unmanipulated animals and those that had been subjected to cardial function during bacterial infective cardiomyopathy. Am. multiple episodes of anesthesia. Differences identified, al- J. Cardiol. 37:373–381. 11. Fein, F. S., B. Miller-Green, B. Zola, et al. 1986. Reversibility though statistically significant, may not have been clinically of diabetic cardiomyopathy with insulin in rabbits. Am. Physiol. relevant. Cardiac dimension data suggest right-chamber dila- Soc. :H108–H113. tation in group A rabbits, but this difference was not corrobo- 12. Khan, M. Y. 1973. Radiation-induced cardiomyopathy. Am. J. rated by echocardiography. The absence of substantial Pathol. 73:131–140. differences in cardiac dimensions and echocardiographic vari- 13. Khan, M. Y. 1971. Radiation-induced cardiomyopathy in rabbits: an ultrastructural study. Radiat. Res. 47:268–269. ables in association with myocardial fibrosis suggests that 14. O’Hara, P. J., and K. R. Pierce. 1974. A toxic cardiomyopathy these measurements are not sensitive ones for assessment of caused by Cassia occidentalis I. Morphological studies in poi- this syndrome. Thoracic radiography and electrocardiography soned rabbits. Vet. Pathol. 11:97–109. also failed to reflect myocardial disease in all but one of the 15. O’Hara, P. J., and K. R. Pierce. 1974. A toxic cardiomyopathy caused by Cassia occidentalis II. Biochemical studies in poisoned seven rabbits evaluated. Markers of myocardial injury, such as rabbits. Vet. Pathol. 11:110–124. ubiquinone, creatine kinase, and lactate dehydrogenase, might 16. Lee, J. C., and S. E. Downing. 1986. Ventricular performance provide insight into the kinetics of these effects in rabbits (45), in diabetic rabbits with norepinephrine cardiomyopathy. Soc. Exp. but to date they have not been evaluated in intact animals. Bio. Med. 181:345–350. The findings reported here suggest that cardiovascular le- 17. Downing, S. E., and V. Chen. 1985. Myocardial injury following endogenous catecholamine release in rabbits. J. Mol. Cell. sions are induced when ketamine/xylazine is administered as Cardiol. 17:377–387. described here. Reduction in the number or duration of anes- 18. Simons, M., and S. E. Downing. 1985. Coronary vasoconstric- thetic episodes, modification of the dosages used, or use of ven- tion and catecholamine cardiomyopathy. Am. Heart J. 109:297– tilatory support or enhanced oxygen may have mitigated 304. 19. Lee, J. C., and D. P. Sponenberg. 1985. Role of ␣-1- anesthetic-associated cardiovascular injury. Other causes, in- adrenoceptors in norepinephrine-induced cardiomyopathy. Am. cluding bacteria, viruses, and toxins, have not been exhaus- J. Pathol. 121:316–321. tively investigated but seem unlikely. The importance of the 20. Lurie, K. G., M. R. Bristow, W. A. Minobe, et al. 1988. rabbit for in vivo vascular studies supports further evaluation 6-hydroxydopamine mediated cardiotoxicity in rabbits. Am. J. of the cardiovascular lesions described here. The roles of vita- Cardiovasc. Pathol. 2:181–191. 21. Khan, M. Y., B. Haider, and I. S. Thind. 1983. Emetine-in- min E, anesthetic-induced physiologic alterations, or a direct duced cardiomyopathy in rabbits. J. Submicrosc. Cytol. 15: effect such as xylazine-induced coronary vasoconstriction dur- 495–507. ing periods of increasing heart work, will be addressed in fu- 22. Friesen, J. M., J. A. J. Ferris, S. S. Rabkin, et al. 1979. ture studies. Pathological features of cantharidin-induced toxic cardiomyopathy: lack of correlation between electron-microscopic and histopathologic myocardial damage. Forensic Sci. Int. 13:187–192. Acknowledgements 23. Dodd, D. A., J. B. Atkinson, R. D. Olson, et al. 1993. Doxoru- We thank Norman Lefebvre of Covance Research Products for bicin cardiomyopathy is associated with a decrease in calcium tissue donations and animal records. release channel of the sarcoplasmic reticulum in a chronic rabbit model. J. Clin. Invest. 91:1697–1705. 24. Jaenke, R. S., D. Deprez-DeCampeneere, and A. Troust. 1980. Cardiotoxicity and comparative pharmacokinetics of six References anthracyclines in the rabbit. Cancer Res. 40:3530–3536. 1. Weisbroth, S. H., R. E. Flatt, and A. L. Kraus (ed.). 1974. 25. Li, X., J. C. Murphy, and N. S. Lipman. 1995. Eisenmenger’s The biology of the laboratory rabbit, 1st ed. Academic Press, syndrome in a New Zealand White rabbit. Lab. Anim. Sci. 45: San Diego. 618–620. 2. Edwards, S., J. D. Small, J. D. Geratz, et al. 1992. An ex- 26. Weber, H. W., and J. J. VanderWalt. 1973. Cardiomyopathy in perimental model for myocarditis and congestive heart fail- crowded rabbits. S. Afr. Med. J. 47:1591–1595. ure after rabbit coronavirus infection. J. Infect. Dis. 165: 27. Hesselton, R. M., W. C. Yang, P. Medveczky, et al. 1988. Patho- 134–140. genesis of Herpesvirus sylvilagus infection in cottontail rabbits. 3. Newcomer, C. E., J. I. Ackerman, J. C. Murphy, et al. 1984. Am. J. Pathol. 133:639–647. The pathogenicity of Salmonella mbandaka in specific patho- 28. Palmore, W. P. 1990. A fatal response to xylazine and ketamine gen free rabbits. Lab. Anim. Sci. 34:588–591. in a group of rabbits. Vet. Res. Commun. 14:91–98. 4. DeLong, D., and P. J. Manning. 1994. Bacterial diseases. In P. 29. Teng, B., W. W. Muir, and D. E. Mason. 1995. Segmental re- J. Manning, D. H. Ringler, and C. E. Newcomer (ed.), The biology sponse of canine coronary arteries to alpha-2 agonists, xylazine of the laboratory rabbit, p. 129–170. Academic Press, San Diego. and medetomidine. Vet. Surg. 24:537. 5. Hurley, R. J., R. P. Marini, D. L. Avison, et al. 1994. Evalua- 30. Flores, N. A., R. L. Davies, W. J. Penny, et al. 1984. Coronary tion of detomidine anesthetic combinations in the rabbit. Lab. microangiography in the guinea pig, rabbit, and ferret. Int. J. Anim. Sci. 44:472–477. Cardiol. 6:459–471. 6. Koller, L. D. 1969. Spontaneous Nosema cuniculi infection in 31. Maxwell, M. P., D. J. Hearse, and D. M. Yellon. 1987. Species laboratory rabbits. J. Am. Vet. Med. Assoc. 155:1108–1114. variation in the coronary collateral circulation during regional 7. Masaki, H., T. Imaizumi, S. Ando, et al. 1993. Production of myocardial ischaemia: a critical determinant of the rate of evo- chronic congestive heart failure by rapid ventricular pacing in lution and extent of myocardial infarction. Cardiovasc. Res. the rabbit. Cardiovasc. Res. 27:828–831. 21:737–746.

159 Vol 49, No 2 Laboratory Animal Science April 1999

32. Bosso, F. J., F. D. Allman, and C. F. Pilati. 1994. Myocar- 44. de Jong, J. W., P. van der Meer, A. S. Nieukoop, et al. 1990. dial work load is a major determinant of norepinephrine-in- Xanthine oxidoreductase activity in perfused hearts of various duced left ventricular dysfunction. Am. Physiol. Soc. : species, including humans. Circ. Res. 67:770–773. H531–H539. 45. Jiang, J. P., V. Chen, and S. E. Downing. 1991. Modulation of 33. Maze, M., and W. Tranquilli. 1991. Alpha-2 adrenoceptor ago- catecholamine cardiomyopathy by allopurinol. Am. Heart J. nists: defining the role in clinical anesthesia. Anesthesiology 122:115–121. 74:581–605. 46. Bertazzoli, C., L. Sala, E. Solcia, et al. 1975. Experimental 34. Marini, R. P., D. L. Avison, B. F. Corning, et al. 1992. adriamycin cardiotoxicity prevented by ubiquinone in vivo in Ketamine/xylazine/butorphanol: a new anesthetic combination rabbits. Int. Res. Commun. Syst-Med. Sci. 3:468. for rabbits. Lab. Anim. Sci. 42:57–62. 47. Amouzadeh, H. R., S. Sangiah, C. W. Qualls, et al. 1991. 35. Hobbs, B. A., T. G. Rolhall, T. L. Sprenkel, et al. 1991. Com- Xylazine-induced pulmonary edema in rats. Toxicol. Appl. parisons of several combinations for anesthesia in rabbits. Am. Pharmacol. 108:417–427. J. Vet. Res. 52:669–674. 48. Amouzadeh, H. R., C. W. Qualls, J. H. Wyckoff, et al. 1993. 36. Lipman, N. S., R. P. Marini, and S. E. Erdman. 1990. Com- Biochemical and morphological alterations in xylazine-induced parison of ketamine/xylazine and ketamine/xylazine/ pulmonary edema. Toxicol. Pathol. 21:562–571. acepromazine anesthesia in the rabbit. Lab. Anim. Sci. 40: 49. Borgman, R. F. 1966. The effects of feeding rabbits a vitamin 395–398. E-low diet containing oleic acid. Am. J. Vet. Res. 27:809–813. 37. Peeters, M. E., D. Gil, E. Teske, et al. 1988. Four methods for 50. Zalkin, H., and A. L. Tappel. 1960. Studies of the mechanism general anesthesia in the rabbit: a comparative study. Lab. Ani- of vitamin E action. IV. Lipid peroxidation in the vitamin E-defi- mals 22:355–360. cient rabbit. Arch. Biochem. Biophys. 88:113–117. 38. Popilskis, S. J., M. C. Oz, P. Gorman, et al. 1991. Comparison 51. Yamini, B., and S. Stein. 1989. Abortion, stillbirth, neonatal of xylazine with tiletamine-zolazepam (TelazolR) and xylazine- death, and nutritional myodegeneration in a rabbit breeding ketamine anesthesia in rabbits. Lab. Anim. Sci. 41:51–53. colony. J. Am. Vet. Med. Assoc. 194:561–562. 39. Sanford, T. D., and E. D. Colby. 1980. Effect of xylazine and 52. Bragdon, J. H., and H. D. Levine. 1949. Myocarditis in vita- ketamine on blood pressure, heart rate and respiratory rate in min E-deficient rabbits. Am. J. Pathol. 25:265–271. rabbits. Lab. Anim. Sci. 30:519–523. 53. Innes, J. R. M., and P. P. Yevich. 1954. So-called nutritional 40. Wyatt, J. D., R. A. W. Scott, and M. E. Richardson. 1989. muscular dystrophy as a cause of “paralysis” in rabbits. Am. J. Effects of prolonged ketamine-xylazine intravenous infusion on Pathol. 30:555–565. arterial blood pH, blood gases, mean arterial blood pressure, heart 54. Ringler, D. H., and G. D. Abrams. 1971. Laboratory diagnosis and respiratory rates, rectal temperature and reflexes in the of vitamin E deficiency in rabbits fed a faulty commercial ration. rabbit. Lab. Anim. Sci. 39:411–416. Lab. Anim. Sci. 21:383–388. 41. Herman, E. H., J. Zhang, D. P. Chadwick, et al. 1996. Age 55. Nielsen, J. N., and W. W. Carlton. 1995. Colobomatous mi- dependence of the cardiac lesions induced by minoxidil in the crophthalmos in a New Zealand White rabbit, arising from a rat. Toxicology 110:71–83. colony with suspected vitamin E deficiency. Lab. Anim. Sci. 42. Roberts, R., D. Morris, C. M. Pratt, et al. 1994. Pathophysiol- 45:320–322. ogy, recognition, and treatment of acute myocardial infarction 56. Hunt, C. E., and D. D. Harrington. 1974. Nutrition and nutri- and its complications, p. 1107–1112. In R. C. Schlant and R. W. tional diseases of the rabbit. In S. H. Weisbroth, R. E. Flatt, and Alexander (ed.), The heart, arteries and veins, 8th ed. McGraw- A. L. Kraus (ed.), The biology of the laboratory rabbit. Academic Hill, Inc., New York. Press, Inc., New York. 43. van Vleet, J. F., L. Greenwood, V. J. Ferrans, et al. 1978. 57. Gatz, A. J., and O. B. Houchin. 1947. Histological observa- Effect of selenium-vitamin E on adriamycin-induced cardiomy- tions on the vitamin E-deficient rabbit heart (abstr.). Anat. Rec. opathy in rabbits. Am. J. Vet. Res. 39:997–1010. 97:337. 58. Gatz, A. J., and O. B. Houchin. 1947. The histology of vitamin E-deficient rabbit hearts (abstr.). Anat. Rec. 97:462.

160