1
ANTIMICROBIAL-INDUCED COLITIS IN TWO QUARTER HORSE RACEHORSES Emily Sundman
Dr. Karen Trotter, Basic Science Advisor Dr. Sally Ness, Clinical Advisor Senior Seminar paper
Cornell University College of Veterinary Medicine February 1, 2012
Keywords: C. difficile, antimicrobial-induced colitis, doxycycline, equine

2
Abstract Two Quarter Horse racehorses presented to a referral hospital for acute diarrhea after an
eight day course of doxycycline. Both horses were treated presumptively for antimicrobialinduced colitis and recovered without complications such as severe endotoxemia or laminitis. Presented here is a case report and review of antimicrobial-induced colitis in the adult horse, with emphasis on appropriate testing and mechanism of Clostridium difficile colitis. Keywords: C. difficile, antimicrobial-induced colitis, doxycycline, equine Signalment and History
The patients were two three-year-old Quarter Horse racehorses, a gelding and mare, each approximately 450 kg. They were stabled in adjoining stalls at a racetrack near Dallas, Texas along with approximately thirty other horses also in training. Both were in active training at the track for at least one month under the same trainer. Ten days prior to presentation the two patients, and only the two patients, were treated with doxycycline by the referring DVM for mild upper respiratory signs, including mild mucopurulent nasal discharge, mild mandibular lymphadenopathy, and decreased performance. The exact dose adminstered is not known, but the standard dose is 10 mg/kg twice a day for five days. Two days after the doxycycline was discontinued the mild upper respiratory signs returned in both horses and doxycycline was restarted at the same dose. Blood was drawn from the gelding for a hemogram and serum chemistry panel. The hemogram showed a mild neutrophilia (6.9 thou/uL, ref 2.1-6.7 thou/uL) and mild hyperfibrinogenemia (600 U/L, ref 200-400 U/L). The serum chemistry panel showed a mild hypoproteinemia (6.27 g/dL, ref 6.35-8.10 g/dL). Both horses continued to train while on doxycycline. Both horses went off feed at 5PM and broke with diarrhea at approximately 8PM on the day of presentation. The gelding was down in his stall prior to referral.

3
Clinical Findings-Presentation Six hours after the onset of diarrhea both horses were presented to the referral hospital.
The gelding was quiet, tachycardic (54 beats per minute) and normothermic (100.4F). Capillary refill time was prolonged to three seconds and the mucous membranes were hyperemic. There were decreased borborygmi in all quadrants and his urine was dark coffee colored. The diarrhea was pipestream. The mare was bright, tachycardic (56 beats per minute) and normothermic (101.2F). Capillary refill time was prolonged to three seconds and the mucous membranes were pink. There were decreased borborygmi in all quadrants. Her diarrhea was liquid with small amounts of solid material. Both horses had normal digital pulses in all limbs.
Abdominal ultrasound was performed on both horses. The gelding had thickened large colon wall (0.6 cm) on both the left and right side. Increased intraluminal fluid and hypomotility were noted in both the large colon and small intestine. The mare had thickened large colon wall (0.5 cm) only on the right side. Similarly, significant intraluminal fluid and hypomotility were noted, albeit only in the large colon.
Blood was drawn for a hemogram and serum chemistry panel for both horses. Fecal samples were submitted for Salmonella spp. and Clostridium difficile culture. Abnormalities for the gelding on the hemogram were dehydration (packed cell volume 54 %, ref 30-42 %; total solids 7.8 g/dL, ref 4.5-6.9 g/dL), hyperfibrinogenemia (600 mg/dL, ref 200-400 mg/dL), neutrophilia (11.9 thou/uL, ref 2.1-6.7 thou/uL), lymphopenia (1.1 thou/uL, ref 1.5-5.5 thou/uL), and thrombocytosis (1551 thou/uL, ref 100-600 thou/uL). Abnormalities for the mare on the hemogram were elevated total solids (7.5 g/dL, ref 4.5-6.9 g/dL), neutrophilia (10.8 thou/uL, ref 2.1-6.7 thou/uL), and lymphopenia (1.1 thou/uL, ref 1.5-5.5 thou/uL).

4
The gelding’s abnormalities on the serum chemistry were hyponatremia (129.7 mEq/L, ref 131.5-142.5 mEq/L), hypochloremia (94 mEq/L, ref 94.5-110.0 mEq/L), hypocalcemia (11.2 mg/dL, ref 11.6-13.4 mg/dL), elevated alkaline phosphatase (273.7 U/L, ref 105.2-249.9 U/L), elevated creatinine kinase (CK)(572 mg/dL, ref 142-444.3 mg/dL), hypoammonemia (11.3 mg/dL, ref 13-23.5 mg/dL), and hyperalbuminemia (4.41 g/dL, ref 2.95-3.7 g/dL).
Abnormalities for the mare on serum chemistry were hypocalcemia (11.1 mg/dL, ref 11.6-13.4 mg/dL), hypophosphatemia (2.3 mg/dL, ref 2.6-4.8 mg/dL), decreased lactate dehydrogenase (189.5 U/L, ref 217.6-505.7 U/L), decreased CK (114.8 mg/dL, ref 142-444.3 mg/dL), elevated gamma-glutamyl transferase (GGT) (44.1 U/L, ref 8.5-32.9 U/L), elevated total bilirubin (4.11 mg/dL, ref 0.95-3.65 mg/dL), hypoammonemia (10.5 mg/dL, 13-23.5 mg/dL), and hyperalbuminemia (3.87 g/dL, ref 2.95-3.7 g/dL).
It was determined that the gelding was approximately 9% dehydrated, with electrolyte abnormalities consistent with diarrhea. The elevated CK was attributed to rhabdomyolysis. The mare was approximately 7% dehydrated. The hypocalcemia and hypophosphatemia were attributed to the diarrhea. The decreased lactate dehydrogenase and creatinine kinase were incidental. The mare’s elevated GGT and total bilirubin were within normal limits for a fit race horse.1 Both horses’ hypoammonemia was attributed to anorexia. Differential Diagnosis
Due to the combination of physical exam findings, blood work, abdominal ultrasound findings, history of doxycycline administration, and the fact that horses were acutely affected; the primary differential for the diarrhea was antimicrobial-induced colitis. Additional differentials for the diarrhea were, in order of likelihood, non-steroidal anti-inflammatory drug (NSAID) induced colitis, parasitic infestation (strongyles), toxicity (cantharidin, arsenic),

5
carbohydrate overload, inflammatory bowel disease (IBD), intestinal anaphylaxis, and Potomac Horse Fever. Neither horse had received NSAID medications in the previous eight days.2 Many of the differentials, including carbohydrate overload, IBD, and intestinal anaphylaxis, were deemed less likely due to both horses being affected. Other differentials were discounted due to their history and environment; including parasites, toxicity, and Potomac Horse Fever.
Clinical presentation of antimicrobial-induced diarrhea is similar to other causes of acute colitis in the adult horse. Disease can range drastically in severity, duration, and response to treatment and horses will present with varying degrees of dehydration and sepsis. Clinical pathology data may show evidence of diarrhea (dehydration, hyponatremia, hypokalemia, hypocalcemia, metabolic acidosis, prerenal azotemia, and hypoproteinemia). If septic processes are occurring there may be a degenerative left shift, thrombocytopenia, azotemia, elevated liver enzymes, and hyperlactatemia. On physical exam there may be evidence of dehydration (prolonged capillary refill time, prolonged jugular fill, elevated heart rate) and diarrhea. If the horse is septic there may be tachycardia, hyperemic mucus membranes, and elevated digital pulses.2
Antimicrobial-induced diarrhea is a disease found in horses, as well as other veterinary species and humans. This disease is diagnosed by ruling out other causes of diarrhea, history of recent antimicrobial use, and presence of environmental stressors. Three organisms most often associated with antimicrobial-induced diarrhea are Clostridium difficile, Clostridium perfringens, and Salmonella spp.; with C. difficile being the most common.3 C. difficile is a gram-positive anaerobic, spore-forming bacillus found commonly in the environment. Other causes of acute colitis in adult horses that should be ruled out include NSAID associated right dorsal colitis,

6
Potomac Horse Fever, Salmonellosis, feed overload, IBD, prolonged fasting, neoplasia, or toxins (cantharidin, arsenic, mercury).
The primary differentials for the gelding’s dark urine were rhabdomyolysis and polysaccharide storage myopathy. Prior to the gelding’s referral he had been recumbent in his stall which can lead to muscle ischemia and subsequent rhabdomyolysis due to the horse’s own weight. However, Quarter Horses are well known for carrying the polysaccharide storage myopathy gene. Additional, but less likely, differentials include NSAID toxicity, renal disease, toxin ingestion (red maple), idiopathic renal hematuria, and trauma.4 Initial Treatment
Fluid therapy was initiated in both horses based on their calculated dehydration and an estimate of ongoing losses of 20 liters (L)/day. The gelding was given a bolus of 2 L of hypertonic saline, followed by 30 L of Plasmalyte-A without additives over 5 hours (rate 6 L/hour). The mare was given 30 L of Plasmalyte-A without additives over 5 hours.
Both horses were administered Bio-sponge, Metronidazole, and Syngergi probiotic for treatment of diarrhea. Bio-sponge was inappropriately dosed at 1 60cc tube orally per horse. This is the foal dose. The correct initial dose for an adult horse is 48oz of Bio-sponge powder administered with water via nasogastric tube with subsequent doses every 6 hours of 8 oz powder. Metronidazole was administered orally at 20 mg/kg twice a day. 30 ml of Syngergi probiotic (containing Saccharomyces boulardii and Saccharomyces cerevisiae) was administered orally twice a day. Gastroprotectants, consisting of 20 mg/kg of Sucralfate and 1.14 gm Gastrogard (omeprazole), were administered orally twice a day as a preventative measure. Treatment Discussion

7
Treatment is primarily supportive to mitigate the physiological effects of severe diarrhea. Fluid therapy is the standard, with the most common fluids utilized being Plasmalyte A or Lactated Ringers Solution (LRS). It is important that fluid therapy should address ongoing losses in addition to maintenance. Therefore, acute colitis cases can require up to 50-100L per day in an adult horse. The 9% dehydration in the gelding (450 kg) is a 40.5 L deficit while the 7% dehydration in the mare (450kg) is a 31.5 L deficit. A maintenance fluid rate for an adult horse of 450 kilograms (kg) is 1.1 L/hr. These fluids can follow the administration of up to 2 L of hypertonic saline depending on the degree of dehydration and systemic status. Supplementation with potassium chloride (KCl) or calcium chloride (CaCl) should be considered based on electrolyte abnormalities. If intravenous fluids are not possible, an indwelling nasogastric tube can be used to provide oral fluids. The addition of colloids, such as Hetastarch, can temporarily improve hypoproteinemia.5
Metronidazole is the preferred antimicrobial in the central and eastern United States, however reports of metronidazole resistance in the western United States has started to change this perspective.6 Metronidazole is a nitroimidazole antimicrobial with action against anaerobic bacteria and some protozoa. Anti-endotoxic doses of the NSAID flunixin meglumine are advisable, even without evidence of sepsis. Bio-sponge (di-tri-octahedral smectite) is an orally administered binding agent for toxins A and B of C. difficile and the enterotoxin of C. perfringens. Bismuth subsalicylate can be administered orally and has antiprostaglandin activity in humans; however research in veterinary species is lacking.7 Probiotics in horses have not been demonstrated to have a clinical effect for horses; however there are also no significant contraindications.8

8
Clinical Findings and Treatment Changes-Following Morning Both horses significantly improved with the initial therapy five hours of therapy. The
gelding was bright, tachycardic (44 beats per minute) and normothermic (100.8F). Capillary refill time was improved to two seconds; however the mucous membranes remained mildly hyperemic. There were increased borborygmi in all quadrants. He was polyuric but the urine was straw colored. The diarrhea was liquid with particles of solid material. The mare was bright, tachycardic (48 beats per minute) and normothermic (100.5F). Capillary refill time was two seconds and the mucous membranes were pink. There were normal borborygmi in all quadrants and her diarrhea was unchanged. However, strong digital pulses were palpated in all of her limbs.
A hemogram was repeated in both horses. The gelding’s dehydration had resolved (packed cell volume 38%, ref 30-42%; total solids 5.9 g/dL, ref 4.5-6.9 g/dL), however a mild neutrophilia (7.1 thou/uL, ref 2.1-6.7 thou/uL) and lymphopenia (1.0 thou/uL, ref 1.5-5.5 thou/uL) was noted. The mare’s dehydration and white cell abnormalities had resolved (packed cell volume 33%, ref 30-42%; total solids 5.9 g/dL, ref 4.5-6.9 g/dL).
Fluid therapy rates were decreased in both horses based on their improved hydration status. The gelding was given 10 L of Plasmalyte-A with potassium chloride, calcium gluconate, DMSO, Vitamin C, and Vitamin B additives over six hours. The mare was given 10 L of Plasmalyte-A with potassium chloride and calcium gluconate over six hours. Fluids were discontinued after this point.
Each horse was correctly dosed with Bio-sponge, and continued every six hours until 8 pm. One dose of flunixin meglumine was administered to each horse at the 1.1mg/kg dose. Bismuth subsalicylate was administered at the 240 ml/dose every six hours until 8 p.m.

9
Treatments of Metronidazole, Synergi probiotic, and gastroprotectants were continued through 8 p.m. as described previously.
Both horses were discharged to the track after a 2.5 day stay at the hospital. Both were discharged without medications, despite the gelding’s continued soft “cow-pie” consistency manure. The gelding had previously qualified for a race which was running in three days and was required at the track for training and pre-race mandatory drug testing. Two days after discharge fecal culture results were available. The gelding was negative for both Salmonella spp. and C. difficile. The mare was positive for Salmonella spp. and negative for C. difficile. The mare was negative for Salmonella spp. on retest 7 days later. There were two hypotheses for the mare’s positive Salmonella spp. culture. The mare may have had a Salmonellosis infection causing her colitis, or the mare may have been transitorily infected with Salmonella spp. from the track and was shedding due to the stress of colitis and hospitalization.9 As the mare was the less clinically affected horse, had a stressful lifestyle, and was negative on repeated culture the latter was determined to be the more likely scenario. Antimicrobial-Induced Colitis Discussion
Diagnosis of antimicrobial-induced diarrhea is often frustrating and unrewarding. Fecal samples should be tested for C. difficile, C. perfringens, and Salmonella spp.. C. difficile and C. perfringens are considered normal inhabitants of the environment and are frequently found in large quantities in the gastrointestinal tract of neonates of numerous species without causing clinical disease.10 Environmental C. difficile differs from pathogenic C. difficile by toxin production. Culture alone of Clostridium does not prove that the pathogenic form is present or that the Clostridium overgrowth is a cause of disease. Therefore, testing of C. difficile is primarily via toxin tests. Historically, the cell cytotoxin assay has been the gold standard test;

10
however it typically is a 2-3 day test. Numerous ELISA tests exist for both toxin A and B individually as well as for both together. Toxin A ELISA test is a good confirmation of disease, however this will miss toxin A negative strains. Toxin A/B ELISA is acceptable for clinical use but needs a confirmatory test (for example culture or antigen testing) as the specificity and sensitivity of this test is unclear. Antigen ELISA alone is a good screening test, but positives need confirmation with a toxin test. PCR for toxin genes is not useful as a primary test, however it can be used to confirm a positive toxin ELISA result.4 Current recommendation in human hospitals is to use the common antigen (glutamate dehydrogenase) ELISA test as a screening test, with positives confirmed with toxin ELISA tests.11 There are pathogenic C. difficile strains in horses and humans that are toxin A negative and toxin B positive, which makes interpreting positive toxin tests difficult.6 Recent evaluation of six horses who tested positive for antimicrobial-induced diarrhea caused by C. difficile showed that all six horses were positive for toxin A and toxin B, which is the most commonly reported C. difficile in horses. 6
C. perfringens testing is via fecal culture, PCR, or ELISA for enterotoxins. Salmonella testing is via positive fecal cultures, sometimes combined with PCR.
C. difficile has numerous toxins including A, B and binary. Toxin A and toxin B are closely related and located in a pathogenicity locus of the bacteria’s chromosome, whereas binary toxin is located elsewhere on the chromosome. Toxin A is an enterotoxin which binds to known receptors on the apical border of the intestinal epithelial cells. This binding begins a series of reactions that lead to disruption of the cytosolic actin filaments, resulting in decreased cell-cell adhesion and eventually cell death. 12 Toxin B is an efficacious cytotoxin in vitro, and binds preferentially on the basolateral border of colonocytes. Additionally, in-vivo studies have not shown toxin B to be able to penetrate the epithelial border. It is suggested that the damage

11
inflicted by toxin A allows toxin B and binary toxin to translocate to the lamina propria.12 There the toxins incite a significant inflammatory response with subsequent infiltration of neutrophils to the colonic epithelium and submucosa, mast cell degranulation, and secretion of prostaglandins, histamine, and inflammatory cytokines, which causes a secretory diarrhea.13
C. perfringens has numerous subtypes identified by the primary toxin produced, with the C. perfringens type C being the primary severe enteric disease strain in foals and Type A being the most common in the environment.9,14 Recently, there have been reports of several foals having co-infections with C. perfringens type C and C. difficile without a history of previous antimicrobial administration. This raises concern of possible synergistic activity between these two organisms.10
Salmonella spp. causes diarrhea by cytotoxins and exotoxins that lead to hypersecretion and severe inflammatory malabsorptive response. Separately, Salmonella bacteria infiltrate the small intestinal and colonic mucosa where they can be phagocytozed by macrophages, leading to sepsis and prolonged residence in the lymph tissue. Salmonellosis is characterized by four different clinical syndromes; a carrier state without clinical disease, mild disease with neutropenia but without diarrhea or colic, acute or peracute colitis with diarrhea, and septicemia with or without diarrhea. Therefore, separating some presenting syndromes of Salmonellosis from anti-microbial induced diarrhea can be quite difficult.4 Prognosis
Acute colitis in adult horses is very serious. Development of sepsis and laminitis are severe consequences of endotoxin release which can independently lead to death or euthanasia. Numerous retrospective studies in university hospitals have attempted to determine mortality statistics and prognostic indicators. Mortality rates have improved over several decades and

12
many studies now show mortality of approximately 20% (18.8%, 22%, 27%).4,15,16 Severity of
hyperlactatemia 4-8 hours and 24 hours after admission has been associated with predicting survival in horses with acute colitis.17 Cohen, et al, found that prior administration of
antimicrobials leads to a 4.5 fold increased likelihood of mortality in horses presented to a referral hospital for treatment of diarrhea.18 Studies in mares with foals have shown that the
macrolide and lincosamide classes of antimicrobials can lead to severe diarrhea in adult horses.19,20 Recent work showed that the most common antimicrobials implicated in antimicrobial-induced colitis were enrofloxacin, followed by doxycycline and ceftiofur.15
Additionally, work investigating the pharmacokinetics of doxycycline in adult horses, albeit at
doses higher than used in practice, resulted in severe colitis requiring euthanasia in one of the horses,.21 However, any antimicrobial combined with stress is a risk factor for the development of antimicrobial-induced colitis in the horse.2
References
1) McGowan C. Clinical pathology in the racing horse: the role of clinical pathology in assessing fitness and performance in the racehorse. Vet Clin North Am Equine Pract. 2008;24(2);405-421. 2) Reed SM, Bayly WM, Sellon DC. Equine Internal Medicine. 3rd ed. Saunders 2009 3) Ernst NS, Hernandez JA, MacKay RJ, et al. Risk factors associated with fecal Salmonella shedding among hospitalized horses with signs of gastrointestinal tract disease. J Am Vet Med Assoc. 2004;225(2):275-281.
4) Robinson NE, Sprayberry KA. Current Therapy in Equine Medicine. 6th ed. Saunders, 2008.
5) Schott HC. Fluid therapy: a primer for students, technicians, and veterinarians in equine practice. Vet Clin North Am Equine Pract. 2006;22(1):1-14.
6) Magdesian KG, Dujowich M, Madigan JE, et al. Molecular characterization of Clostridium difficile isolates from horses in an intensive care unit and association of disease severity with strain type. J Am Vet Med Assoc. 2006;228(5):751-755.

13
7) Papich MG. Saunders Handbook of Veterinary Drugs: Small and Large Animal. 3rd ed. Saunders. 2010.
8) Atherton RP, McKenzie HC, Furr MO. Acute colitis: pathophysiology and noninfectious causes. Comp Eq. 2009;4:366-373.
9) Slovis NM. Infectious Gastrointestinal Disorders. In 11th WEVA Proceedings. 2009.
10) Uzal FA, Diab SS, Blanchard P, et al. Clostridium perfringens type C and Clostridium difficile co-infection in foals. Vet Microbio. epub 1 Dec 2011.
11) Rupnik M, Wilcox MH, Gerding DN. Clostridium difficile infection: new developments in epidemiology and pathogenesis. Nat Rev Microbiol. 2009;7(7):526-536.
12) Keel MK, Songer JG. The comparative pathology of Clostridium difficile-associated disease. Vet Pathol. 2006;43(3):225-240
13) Kelly CP, Kyne L. The host immune response to Clostridium difficile. J Med Microbio. 2011;60:1070-1079.
14) Diab SS, Kinde H, Moore J, et al. Pathology of Clostridium perfringens Type C enterotoxemia in horses. Vet Pathol. epub 18 April 2011.
15) Barr BS, Waldridge BM, Morresey PR, et al. Antimicrobial-associated diarrhea in three equine referral practices. In AAEP Proceedings. AAEP convention 56;2010:267-268.
16) Weese JS, Toxopeus L, Arroyo L. Clostridium difficile associated diarrhoea in horses within the community: predictors, clinical presentation and outcome. Equine Vet J. 2006;38(2):185-188.
17) Hashimoto-Hill S, Magdesian KG, Kass PH. Serial measurement of lactate concentration in horses with acute colitis. J Vet Int Med. 2011;25(6):1414-1419.
18) Cohen ND, Woods AM. Characteristics and risk factors for failure of horses with acute diarrhea to survive: 122 cases (1990-1996). J Am Vet Med Assoc. 1999;214(3):382-390.
19) Raisbeck MF, Holt GR, Osweiler GD. Lincomycin-associated colitis in horses. J Am Vet Med Assoc. 1981;179(4):362-363.
20) Gustafsson A, Baverud V, Gunnarsson A, et al. The association of erythromycin ethylsuccinate with acute colitis in horses in Sweden. Equine Vet J. 1997;29(4):314-318.
21) Davis JL, Salmon JH, Papich MG. Pharmacokinetics and tissue distribution of doxycycline after oral administration of single and multiple doses in horses. Am J Vet Res 2006;67(2):310316.