|Home | About | Journals | Submit | Contact Us | Français|
Two adult alpacas were presented for recumbency and reluctance to rise. Cantharidin toxicosis was suspected based on clinical and ancillary diagnostic findings. The diagnosis was confirmed by gas chromatography-mass spectrometry of gastric contents and urine. Despite medical treatment, neither alpaca survived. Blister beetle toxicosis has not been previously described in camelids. Challenges in treatment of affected ruminants or pseudoruminants are noted.
Toxicose à la cantharidine chez 2 alpagas. Deux alpagas adultes ont été présentés pour décubitus et une réticence à se lever. La toxicose à la cantharidine a été soupçonnée en se fondant sur des résultats diagnostiques cliniques et auxiliaires. Le diagnostic a été confirmé par chromatographie en phase gazeuse et spectromètre de masse du contenu gastrique et de l’urine. Malgré un traitement médical, les deux alpagas n’ont pas survécu. La toxicose aux cantharides n’avait pas été décrite antérieurement chez les camélidés. Les difficultés du traitement des ruminants ou des pseudoruminants sont signalées.
(Traduit par Isabelle Vallières)
Cantharidin toxicity, caused by ingestion of blister beetles, is well-documented in horses and most commonly causes clinical signs referable to the gastrointestinal and urinary tracts (1–4). The following cases are unusual in that cantharidin toxicosis has not been previously described in camelids, and the inherent challenges in treatment of this condition in ruminants or pseudoruminants have not previously been described. Gastrotomy and gastric lavage may be indicated in these cases to remove the source of cantharidin. Livestock owners should be made aware of the potential for blister beetle toxicosis in their animals especially if they feed alfalfa hay.
Two 2 1/2-year-old nulliparous female huacaya alpacas from the same farm were presented to the Oklahoma State University Boren Veterinary Medical Teaching Hospital (OSU-BVMTH) with an 8-hour history of recumbency and reluctance to rise. The first alpaca (#1) had been found in the pasture that morning in a sternally recumbent position. With strong encouragement the animal would rise but would not walk more than 15 m and appeared uncomfortable during ambulation. The herd veterinarian diagnosed colic of unknown etiology; at that time another alpaca from the same pasture (#2) was noted to be cushed (sternally recumbent) frequently and both animals were referred to OSU-BVMTH.
The alpacas originated from a breeding farm of approximately 200 animals. Feeding history included native grass pasture, grass hay ad libitum, and a 1/2 flake of alfalfa and 0.25 kg of rolled oats per adult animal fed twice daily. Fresh water and a loose mineral supplement formulated for camelids were present at all times. All animals on the farm had been vaccinated against clostridial diseases 9 mo earlier. Both alpacas (#1 and #2) and 12 other breeding-quality females had been moved to a different pasture 2 wk previously. Previous herd medical problems included internal parasitism, enteric salmonellosis, and an episode of multiple cases of grain overload 2 y ago.
Upon arrival at OSU, alpaca #1 (body weight 41.4 kg) was unable to rise. Signs of colic (vocalization, rocking side-to-side while cushed, intermittently rolling into lateral recumbency) were evident. Physical examination revealed a body condition score of 2.5/10, moderate depression, and frequent attempts at caudoventral flexion of her neck. The alpaca was hypothermic [35.7°C, reference range (RR): 37.5°C to 38.9°C], tachycardic (> 100 beats/min, RR: 60 to 90 beats/min), and tachypneic (value not recorded). Mucous membranes were hyperemic and tacky with capillary refill time > 3 s. Petechiae were present on the buccal mucosa and bilateral engorgement of scleral vessels was evident. Dehydration was estimated at 7%. First compartment motility and intestinal borborygmi were absent. These clinical findings were suggestive of shock, either hypovolemic, cardiogenic, or maldistributive in origin. Nasal oxygen was administered (3 L/min) and a jugular catheter was placed to facilitate blood collection for additional diagnostics followed by a shock dose of intravenous fluids (Normosol-R; Abbot Laboratories, North Chicago, Illinois, USA) at 60 mL/kg body weight (BW) per hour, ceftiofur sodium (Naxcel; Pfizer Animal Health, New York, New York, USA) 5 mg/kg BW, IV, potassium penicillin (Pfizerpen; Pfizer Animal Health), 44 000 IU/kg BW, IV, omeprazole (Omeprazole sodium; Premier Pharmacy Labs, Weeki Watchee, Florida, USA), 0.5 mg/kg BW, IV, and Clostridium perfringens type C&D antitoxin, 20 mL IV.
Diagnostic testing included a complete blood (cell) count (CBC), serum chemistry panel, first gastric compartment (C1) fluid examination, and abdominal and thoracic ultrasound. Abnormalities in the CBC included marked hemoconcentration and leukocytosis with a mature neutrophilia (Table 1). Pertinent serum chemistry findings included azotemia, hyperphosphatemia, hypermagnesemia, increased muscle enzymes, hyperglycemia, hypertriglyceridemia, and hypocarbia (Table 2). Urinalysis was not available. First compartment fluid examination revealed a pH of 7.0 with no motile protozoa. Clinicopathologic findings were consistent with an inflammatory process resulting in shock, acute oliguric renal failure, and muscle damage. The gastrointestinal tract (GIT) was the likely origin given the absence of motile protozoa and evidence of colic. Results of abdominal and thoracic ultrasonography were within normal limits. Less than 2 h after arrival the alpaca became obtunded, lost menace response, and the pupils became dilated and unresponsive to light. Given the rapid deterioration in clinical appearance despite therapy and the grave prognosis regardless of etiology, the owner was contacted and the animal was humanely euthanized with pentobarbital sodium and submitted for postmortem examination.
The second alpaca (42.7 kg) had a body condition score of 2.5/10, was able to ambulate normally and was bright, alert, and responsive but attempted to cush frequently in the chute. Abnormalities on physical examination were limited to mild hypothermia (36.8°C) and C1 atony and lack of a palpable fiber mat. Differential diagnoses for 2 affected animals from the same herd were likely nutritional, infectious, toxic, or metabolic in origin and for these clinical findings included grain overload, salmonellosis, enterotoxemia, and toxin ingestion resulting in death of gastric protozoa.
Examination of C1 fluid revealed a pH of 6.5 with no motile protozoa but normal color and odor. Abdominal ultrasound was unremarkable. A CBC revealed marked leukocytosis characterized by a left shift and monocytosis (Table 1). Pertinent serum chemistry findings included azotemia, increased muscle enzymes, hyperglycemia, and hypocalcemia (Table 2). Considering all clinical and ancillary diagnostic findings, toxins affecting both the gastrointestinal and renal systems were suspected, including cantharidin, oak, arsenic, mercury, lead, or copper.
A jugular catheter was placed and treatment with ceftiofur sodium (Naxcel; Pfizer Animal Health), 5 mg/kg BW, IV, q12h, potassium penicillin (Pfizerpen; Pfizer Animal Health), 44 000 IU/kg BW, IV, q6h, omeprazole (Omeprazole sodium; Premier Pharmacy Labs), 0.5 mg/kg BW, IV, q24h, balanced isotonic IV fluids (Normosol-R; Abbot Laboratories), 100 mL/kg BW per day with added KCl (20 mEq/L), 6% hydroxyethyl starch (Hespan; Braun Medical, Irvine, California, USA), 20 mL/kg BW per day, Clostridium perfringens type C&D antitoxin, 20 mL SQ, q12h, di-tri-octahedral smectite (Equine Bio-sponge; Platinum Performance, Buellton, California, USA), 10 g, PO, q12h, and transfaunation, 750 mL once, was initiated. Calcium gluconate (23%, 100 mL) was added to the first 1 L of fluids. Urine was collected overnight and was grossly normal; however, microscopic hematuria was confirmed on sediment examination (Table 3). The urine was submitted for gas chromatography/mass spectrometry (GC/MS) analysis for cantharidin. On day 2, the patient developed mild diarrhea. A fecal occult blood test (Hemoccult Fecal Occult Blood Test; Beckman Coulter, Fullerton, California, USA) was negative. Repeat chemistry profile indicated essentially no change in azotemia, thus low-dose dopamine constant rate infusion (CRI), 3.5 μg/kg BW per minute, was initiated to improve renal perfusion. Marked hyperglycemia (Table 2) was still present and regular insulin (Novolin R; Novo Nordisk, Princeton, New Jersey, USA), 0.2 IU/kg BW, IV, q1h, was started along with hourly glucose monitoring until blood glucose decreased to normal range. Throughout the rest of the day the alpaca ate small amounts of grass hay. Overnight the patient became dysuric and frank blood was noted at the end of urination.
On day 3, cystoscopy revealed multiple hyperemic erosions and a single vesicle in the bladder mucosa. Esophagoscopy disclosed fibrin tags along the esophageal mucosa that were more prevalent in the distal esophagus. In the distal 1/3 of the esophagus, severe patchy and linear to coalescing mucosal disruptions were present that increased in severity to become circumferential diffuse blackened regions of mucosa that extended to the cardia (Figure 1). Butorphanol (Torbugesic; Fort Dodge Animal Health, Fort Dodge, Iowa, USA), 0.2 mg/kg BW, 1/2 SQ and 1/2 IV, q6h, and sucralfate, 1 g tablet administered as a slurry PO, q6h, were initiated. First compartment contractions increased to 3/2 minutes, and normal pelleted stool was present. Polydipsia (consumption of > 10 L of water in 24 h) was noted. Changes on bloodwork showed a marked neutropenia with a degenerative left shift, worsening hypoproteinemia and azotemia, and hypocalcemia. A furosemide (Furoject; Butler Animal Health, Dublin, Ohio, USA) CRI, 0.12 mg/kg BW per hour following a loading dose of 0.12 mg/kg BW, was initiated for persistent azotemia and 40 mEq/L KCl added to fluids in anticipation of kaliuresis in addition to 75 mL of 23% calcium gluconate. Commercial llama plasma (Llama plasma; Triple J Farms, Kent Laboratories, Bellingham, Washington, USA), 544 mL, was administered IV for hypoproteinemia. Cantharidin toxicosis was confirmed by GC/MS on the previously submitted urine sample (8.19 ppm; reference value: 0 ppm). The owners were contacted and instructed to stop feeding all alfalfa hay on the premises and to bring the suspect hay bales to OSU-BVMTH to be examined for blister beetles. An irregular arrhythmia was ausculted once overnight, and cardiac troponin I (cTNI) at that time was increased (0.9 ng/mL, equine RR: 0.0 to 0.06 ng/mL) indicating myocardial damage. Thereafter the alpaca became persistently tachycardic (100 beats/ min). An electrocardiogram (ECG) was attempted; however, the patient was extremely fractious and a diagnostic reading could not be obtained. The dopamine CRI was discontinued as the β2-receptor agonist activity of the drug could have contributed to tachycardia.
On day 4 of hospitalization the alpaca became anorexic and colic was observed. Clinicopathologic changes included hypomagnesemia, thus oral magnesium paste was given, 25 mL, q24h. Commercial llama plasma, 500 mL, was again given IV and thiamine HCl, 20 mg/kg BW, SC, q8h was added. Given development of hypomagnesemia, which can inhibit the release of parathyroid hormone (PTH) along with the intermittent recurrent hypocalcemia, paired PTH and ionized calcium concentrations were submitted and were low [ionized calcium 0.93 mmol/L, RR: 1.25 to 1.75 mmol/L; PTH 0.00658 ng/mL, RR: 1.90 ± 0.05 ng/mL in healthy nonpregnant female dromedary camels (5)].
The following day the patient became reluctant to rise and remained anorexic. Partial parenteral nutrition was administered starting at 1/4 of total caloric requirements per day along with insulin therapy as previously described. Complete blood cell count indicated a sudden decrease in HCT; internal hemorrhage was a major concern. Fecal occult blood remained negative and urine was grossly normal. Abdominal and thoracic ultrasound did not reveal evidence of hemoabdomen or hemothorax. A whole blood transfusion (1 L) from a donor llama was given. Serum chemistry showed decreasing magnesium despite oral supplementation and parenteral magnesium sulfate (20 mL of a 10% solution, SC) was given. Fluids were changed to 0.45% NaCl with 60 mEq/L KCl and 100 mL calcium gluconate added. The alpaca appeared to briefly improve following therapeutic alterations and was seen eating grass hay, but that evening she became progressively more depressed to the point of obtundation, acutely and severely tachycardic (> 180 beats/min), tachypneic, and febrile (40°C). The owner was contacted and elected humane euthanasia given the grave prognosis. The alpaca was submitted for complete postmortem examination.
Gross necropsy lesions in alpaca #1 were limited to the kidneys, GIT, and heart. The renal cortices of both kidneys contained numerous irregular, pale red to yellow wedge-shaped infarcts radiating from the mid cortex to the capsular surface. Mild ulceration was noted in the distal esophagus. The first and second stomach compartments (C1 and C2) contained dark brown-gray fluid and feed material, and the mucosa of the third gastric compartment (C3) was diffusely congested. The serosa of the small intestine was segmentally hyperemic. Histologically, the renal medulla contained multifocal regions in which the renal tubule and collecting duct epithelial cells were variably undergoing degeneration and necrosis. The third stomach compartment and small intestine contained segmental regions of congestion and hemorrhage in the lamina propria and submucosa. Several foci of mild endocardial hemorrhage extended into adjacent myocardium. Hepatic mineral analysis revealed low arsenic, lead, and copper levels (arsenic, 0.037 ppm, reference value: < 0.2 ppm; lead, 0.026 ppm, reference value: < 2 ppm; copper, 27.6 ppm, RR: 30 to 100 ppm). Gastric content GC/MS was positive for cantharidin (0.48 ppm) confirming the presumptive diagnosis of acute cantharidin toxicosis.
Gross lesions in alpaca #2 were observed in the esophagus, C1, C2, peritoneum, lungs, pericardial sac, heart, kidneys, and urinary bladder. Extensive erosion and ulceration were noted in the distal esophagus (Figure 2). The wall of C1 was expanded by submucosal edema; the mucosa of C1 and C2 was dark red (Figure 3). There was extensive petechiation of the peritoneum. Pulmonary edema and multifocal, subpleural hemorrhages were present in the lungs. The pericardial sac contained approximately 50 mL of red-tinged yellow fluid and a focal area of red discoloration was identified in the left ventricle of the heart (Figure 4). Acute renal infarcts were observed in both kidneys. There was extensive petechiation of the urinary bladder mucosa (Figure 5). Histologically, necroulcerative lesions were present in the distal esophagus and C1. Myocardial degeneration, necrosis, and mineralization were seen in the right ventricle. GC/MS analysis for cantharidin was performed on the C1 contents and aqueous humor. Concentrations of cantharidin in these specimens were 0.122 ppm and 0.0 ppm, respectively.
Examination of the suspect alfalfa hay bales for evidence of blister beetles did not reveal any cantharidin-containing species but did reveal ground beetles (Carabidae, Pterostichus spp.) and darkling beetles (Tenebrionidae, Blapstinus spp). Recommendations were made to dispose of the remainder of the alfalfa hay as it was all from the same cutting.
Blister beetle toxicosis has been reported in a variety of species, most commonly horses, but this toxicosis has not been previously documented in camelids (6–11). The toxic principle in blister beetle poisoning is cantharidin, a bicyclic terpenoid vesicant and acantholytic agent whose mechanism of action at the molecular level is not known (12,13).
In Oklahoma alone, 67 species of blister beetle have been identified with Epicauta occidentalis being involved in most of the confirmed equid intoxications (14). The cantharidin content of the hundreds of species of blister beetles in North America varies widely; the toxin is only synthesized by males but is transferable from males to females. The blister beetle Epicauta immaculata can contain an average of 5.2 mg of cantharidin per beetle (15). The lethal dose of cantharidin in horses is approximately 1 mg/kg BW but may be less (1,3,13,16); thus the ingestion of as few as 90 beetles can result in the death of a 454 kg horse.
There are no published controlled studies documenting the LD50 of cantharidin in ruminants or pseudoruminants. An unpublished study (John Reagor, plant toxicologist, retired from Texas Veterinary Medical Diagnostic Laboratory, 2012) found that as little as 0.5 mg/kg was lethal in cattle, indicating ruminants may be more susceptible than horses to cantharidin. The paucity of detailed descriptions of cantharidin toxicity in ruminants may indicate a lack of diagnostic scrutiny when toxicity is suspected or confirmed, or that sublethal or lethal cantharidin toxicosis is underdiagnosed in these species.
Clinical signs of cantharidin toxicity vary in severity and are dose-dependent (12). The clinical presentation of blister beetle ingestion is best described in equids (1,4). Signs frequently encountered include abdominal pain, fever, depression, oral mucosal ulcerations, immersion of the muzzle in water, diarrhea, dysuria, stiff gait, increased heart and respiratory rates, congested mucous membranes, sweating, lethargy, and anorexia (1,3,4,13,17). Polyuria, pollakiuria, gross hematuria, ptyalism, melena, and synchronous diaphragmatic flutter from hypocalcemia may also be seen (1,3,4,17). Lethal doses of cantharidin can result in signs of hypovolemic shock or sudden death (2,17). In cattle the most consistent clinical signs in natural intoxication were mass feed refusal and decreased milk production (7). Additional clinical signs reported were salivation, oral ulceration, diarrhea, bruxism, abdominal pain, polyuria, reluctance to move, ataxia, and recumbency (7). The alpacas in this report had clinical presentations very similar to those observed in horses, with evidence of colic being the predominant sign. Diarrhea, dysuria, hematuria, polydipsia, hyporexia, hypomotility of the GIT, and depression were also apparent in alpaca #2. Interestingly, this animal had no oral lesions and continued to eat intermittently despite severe distal esophageal and forestomach ulceration. The absence of oral lesions and the distribution of the esophageal lesions might suggest that passive or partial regurgitation of the toxin in ruminating species exacerbates the severity of the GIT mucosal lesions. Signs consistent with hypovolemic shock were apparent 7 h following the onset of abdominal pain in alpaca #1.
Clinicopathologic findings in these alpacas varied. Consistent findings on initial presentation included neutrophilic leukocytosis, azotemia, hyperglycemia, and increased creatine kinase; these laboratory findings parallel those often found in equine cantharidin toxicosis. Marked hemoconcentration, a frequent finding in the early stages of cantharidiasis in equids (3,4,13), was also present in the first case. Hypoproteinemia from hypoalbuminemia in the second alpaca may be credited to protein-losing gastropathy or enteropathy. Hypocalcemia is the most consistent biochemical finding in horses with cantharidin toxicity (17) and was the most notable abnormality on the initial biochemistry panel performed in the second alpaca. Anecdotally, approximately 75% of horses with blister beetle toxicosis become hypocalcemic. Hypomagnesemia often accompanies hypocalcemia in equids (1,3,17), but was not apparent in either alpaca initially; alpaca #1 was hypermagnesemic, which might be explained by oliguria. Common urinalysis findings in equids with blister beetle toxicity include hyposthenuria and microscopic hematuria (3,4,17); while microscopic and subsequently gross hematuria were present in alpaca #2, hyposthenuria was never evident.
A unique finding in both alpacas was the absence of motile protozoa on C1 fluid analysis. In conjunction with the normal C1 pH this finding was highly suggestive of a severe disturbance of the forestomach flora unrelated to an acidotic event. The possibility exists that protozoal death is also dependent on the amount of cantharidin ingested.
Myocardial necrosis has been reported in cases of blister beetle toxicosis in equids (2,4,18). At necropsy, 4 horses out of 21 with cantharidin toxicity had ventricular myocardial necrosis (4), a higher proportion compared with 1 out of 24 equids necropsied in another report (2). Recently, antemortem diagnosis of myocardial damage has also been documented via increased cTNI in 46% of horses with cantharidiasis (19). Cardiac arrhythmias were not documented in any of the animals included in that prospective study (19). The second alpaca in this case study had increased cTNI with a concurrent arrhythmia, and myocardial degeneration and necrosis with mineralization were confirmed in the right ventricle at necropsy.
The toxic effects of cantharidin and its analogues in mammalian tissues have been ascribed to an affinity and specificity for a cantharidin-binding protein subsequently identified as protein phosphatase type 2A (PP2A) (20–22). Cantharidin inhibits PP2A and protein phosphatase type 1 (PP1) activity (20,21,23–25), and it has been suggested that this may be the in vivo mechanism whereby by these compounds exert their toxic effects (21). Interestingly, inhibition of PP1 has also been demonstrated to decrease PTH secretion from bovine parathyroid cells in vitro(26). This may help explain the low serum PTH concentration despite documented ionized hypocalcemia in case #2 and warrants further investigation. Recent research has established that cantharidin also has inhibitory effects on osteoclast differentiation and bone resorption activity but did not elucidate what if any effects this might have on calcium homeostasis in vivo(27). Either or both of these mechanisms could be responsible for the hypocalcemia associated with cantharidiasis in clinical cases.
Definitive diagnosis of cantharidin toxicosis can be made via GC/MS or high-pressure liquid chromatography, but GC/MS is more sensitive and specific (10,11,16,28). Urine and gastric contents are the diagnostic specimens of choice (10,11,16,28). Urine samples collected more than 72 h after ingestion of the toxin could result in a false negative test result (10); similarly, urine samples obtained after intensive IV fluid therapy may contain undectectable concentrations of cantharidin due to dilution. Of interest in case #2 is that 5 d following presentation cantharidin was still detectable in C1 contents suggesting this as an alternative diagnostic sample in ruminants or pseudoruminants that survive for more than 3 d. Although serum has been used to make a postmortem confirmation of cantharidiasis in a human (29), a sensitive assay for blood has not been developed (16). Detection of cantharidin even at very low concentrations is likely significant (11) regardless of sample type. Aqueous (ocular) fluid was tested for the presence of cantharidin in case #2 in an attempt to identify a readily accessible, non-urine fluid specimen that might have diagnostic value at necropsy. The inability to detect cantharidin in the ocular fluid was not surprising given that cantharidin is irritating to tissues and clinical signs of ocular disease were not observed in this animal.
There is no specific antidote for cantharidin toxicosis, so therapy is generally supportive (3,17). Treatment goals include elimination of the toxin and its source, decreasing toxin absorption, pain management, maintenance of fluid and electrolyte balances, and gastroprotection (3,17). Oral di-tri-octahedral smectite was used as an adsorbent in alpaca #2. Based on a recent study, activated charcoal (1 g/kg BW, PO via orogastric tube, as a slurry) may have been a better selection (17). Given the severity of the forestomach and esophageal ulcerations, case #2 might have benefitted from a C1 gastrotomy for removal of GI contents and C1 compartment lavage to minimize toxin absorption. However, the degree of mucosal damage already present could have compromised surgical recovery and the subacute nature of the intoxication may have mitigated the benefit of a gastrotomy.
Pain management options in blister beetle poisoning include nonsteroidal anti-inflammatories (NSAIDs), α2 agonists, and opioids (17). Nonsteroidal anti-inflammatories were considered suboptimal as GI mucosal ulceration was already present in addition to apparent renal compromise in case #2. Butorphanol was chosen instead, as cantharidin antagonizes the antinociceptive effects of α2 agonists (30) and the μ-opioid agonist morphine (31) in mice, but had no adverse effect on the analgesic effects of κ-opioid receptor agonists (30).
Prognosis in cases of cantharidin toxicity varies from poor (3) to good (17) and is likely dependent on the amount of toxin ingested in addition to timely recognition and aggressive treatment. Horses that survive more than 2 d following toxin ingestion are reported to have a good prognosis for survival (4). Persistent tachycardia, tachypnea, and increased creatine kinase appear to be poor prognostic indicators in affected horses (13). Prognosis in ruminants is difficult to ascertain as very few reports describe experimental or natural cantharidiasis in these species. Deaths have been reported in cattle (7), sheep (10), a goat (11), and in this current case study in 2 alpacas.
Prevention centers around minimizing exposure to suspect alfalfa hay, cubes, or wafers (32). One management strategy utilized by equine owners is to inspect each flake of alfalfa hay for the presence of blister beetles (32). However, hay may still be contaminated even if beetles are not found as cantharidin is released when the beetles are crushed (12). All affected hay should be discarded rather than fed to ruminants as these species are susceptible to cantharidin intoxication (33). Pelleted alfalfa is unlikely to result in exposure to cantharidin as the toxin is drastically diluted during the course of processing (1).
In summary, this is the first report to describe cantharidiasis in a camelid species. Additionally, attempted treatment of blister beetle poisoning in a ruminant or pseudoruminant species has not previously been reported. Livestock owners, including those with cattle, sheep, goats, alpacas, and llamas, should therefore be made aware of the potential for blister beetle toxicity especially if they feed alfalfa hay.
We acknowledge Dr. Roger Panciera and Dr. Keith Bailey for their assistance in the preparation of this manuscript. CVJ
Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office (gro.vmca-amvc@nothguorbh) for additional copies or permission to use this material elsewhere.
All work was done at Oklahoma State University Boren Veterinary Medical Teaching Hospital. No part of this work was supported by a grant.