Bacterial diarrhea and related public health concerns (Proceedings)

Normal intestinal flora

The mammalian GI tract is sterile during fetal development, and the fetus is originally exposed to bacteria during passage through the birth canal. Bacteria are ingested from the local environment and travel through the GI tract, competing with other bacteria and ultimately colonizing in their ideal niche. Initially, as the flora evolves, it is very similar to the mother's GI flora, due to close physical contact and nursing; however, with maturation and independence, the adult GI flora becomes an individualized complex ecosystem based on environment, health, stress, nutrition, and exposure to antibiotics and/or probiotics. This allows for a marked variation of GI flora, even among individuals living in the same environment. Bacteria counts are lower in the stomach and proximal small intestine, due to the low pH and bile, and increase progressively through the intestines, peaking in the large intestine. While the GI flora can remain fairly stable over long periods of time, rapid changes can occur with administration of antimicrobial therapy or infection with virulent pathogens. The normal flora is a symbiotic partner with the host, and the flora's roles include: competing with potential pathogens for binding sites and nutrients, stimulating IgA production, producing antimicrobial compounds to suppress activity of pathogens, contributing to the regulation of GI transit time, carbohydrate fermentation, bile acid metabolism, short-chain fatty acid production, and folic acid synthesis.

Physiology of diarrhea

There are 4 main physiologic mechanisms that can lead to diarrhea: secretory, osmotic, increased permeability, and dysmotility. Most diarrheas are caused by a combination of these mechanisms; however, secretory is the predominant mechanism involved with bacterial diarrhea. The enteric nervous system plays an important role regulating local and distant motility, inflammation, response to toxins, and both baseline secretion and hypersecretion of electrolytes and fluid. Acetylcholine and vasoactive intestinal peptide (VIP) are the major neurotransmitters which stimulate secretion from enterocytes. Acetylcholine activates secretion of chloride using a calcium-dependent signal stimulated by activation of muscarinic receptors on the epithelium. VIP triggers a cAMP-mediated secretion of chloride and water. Inflammation caused by pathogens triggers prostaglandin production which can stimulate epithelial cell secretion directly (PGE2) or activate a neural reflex via acetylcholine and VIP to cause chloride and water secretion (PGI2).

Bacterial pathogenesis

Bacterial pathogens have numerous strategies and virulence properties to increase their chances of becoming established in the GI tract and contributing to disease. These strategies include toxin production, adherence, and mucosal invasion. Toxins can bind to receptors on enterocytes and stimulate cyclic AMP to increase chloride secretion and decrease sodium reabsorption, causing loss of large amounts of water. Adhesion mechanisms are varied. For example enteropathogenic E. coli (EPEC) contain pili and fimbriae, attaching and effacing lesion from type III secretion systems, and intimin, which allow strong adhesion as well as rearrangement of actin, villi destruction, influx of white blood cells, and signal transduction pathways causing electrolyte and water loss. Alternatively, other bacteria such as enteroinvasive E. coli (EIEC) are taken up by M-cells (antigen sampling cells) in the intestine and avoid immune destruction by inducing apoptosis of macrophages. Additional bacteria that can invade either intestinal mucosa or submucosa leading to diarrhea include Salmonella, Shigella, and Campylobacter.

Bacterial pathogens of concern

Clostridium spp. are Gram-positive anaerobic spore-forming rods that are part of the normal flora in dogs but can be triggered to form spores and toxins contributing to illness with antibiotic exposure, dietary change, pH change, immunosuppression, or concurrent viral infection. In humans, clostridial infections are more often obtained via ingestion of food contaminated with preformed toxin. The pathogenesis of diarrhea caused by Clostridium perfringens is not well understood in dogs, as the bacteria can be isolated from more than 80% of normal dogs. Diarrhea associated with C. perfringens most typically presents as acute colitis, with blood, mucus, and tenesmus. However, it has also been associated with small or diffuse bowel disease and more chronic diarrhea. Traditionally, diagnosis of Clostridium-related diarrhea has included consistent clinical signs and cytological evaluation of a stained thin fecal smear, with visualization of an increased number of endospores (3-5 per high power field has been considered normal). Unfortunately this diagnostic tool can be unreliable, as sporulation can occur in non-diarrheic dogs and by nonpathogenic strains of Clostridium. In 2001, Weese et al reported no association between spore number and the presence of C. perfringens endotoxin (CPE). Detection of toxin in fecal samples can be performed using a reverse passive latex agglutination assay or an ELISA. Weese's study, as well as a study by Marks et al, reported an association between toxin formation (CPE) and presence of diarrhea. In Weese's study, identification of CPE using ELISA was positive in the feces of 34% of diarrheic and 14% nondiarrheic dogs, and the association was stronger when comparing the combination of presence of both CPE and the cpe gene (identified by PCR) with presence of diarrhea. A combination of fecal ELISA for presence of toxin with PCR for presence of the toxin gene, cpe, may be prove to be the most useful diagnostic strategy. Culture of feces is not useful for diagnosing Clostridium, since they are part of the normal flora, although culture can be used to rule out this infectious agent. Treatment for C. perfringens should include supportive care as needed (fluid therapy, GI protectants, bland or high fiber diet) and antibiotics (ampicillin, metronidazole, erythromycin, or tylosin) for moderate to severe cases. There is little zoonotic risk of transmission from pets to people, and most people become ill from contaminated food.

Clostridium difficile can cause diarrhea in dogs and cats and is a leading cause of hospital-acquired diarrhea in human hospitals and long-term care facilities. Similar to C. perfringens, C. difficile can be isolated from the feces of up to 40% of healthy and diarrheic dogs, and so its role in disease is not fully understood. Infected dogs may be asymptomatic (colonized), or have large or small bowel diarrhea that can range from mild to fatal acute hemorrhagic syndrome. Signs in cats include watery diarrhea, vomiting, and fever. C. difficile has 2 clinically relevant toxins as virulence factors, Toxin A (enterotoxin) and Toxin B (cytotoxin). Diagnosis is based on identifying presence of these toxins using an ELISA. Not all strains of C. difficile produce toxins, and rarely humans and pets can carry strains of C. difficile that do not produce toxins or disease. An association has been made with detection of toxins and presence of diarrhea in dogs. Hospitalization and previous antimicrobial therapy are risk factors for diarrhea from C. difficile in humans and have been investigated in cats and dogs as well. In humans, any antibiotic could predispose an individual to C. difficile infection, but cephalosporins, penicillins, and clindamycin have frequently been implicated. An outbreak of nosocomial C. difficile occurred in a veterinary teaching hospital in Canada, with Toxin A or B detected in feces of 52% of dogs with diarrhea over a 5-month span. Weese's group also performed a prospective study assessing C. difficile in feline and canine patients in the ICU, by culturing their feces on admission and every 3 days during hospitalization. They found that 11% of tested pets had C. difficile upon admission and 8.3% of tested pets developed detectable C. difficile during their hospitalization; 69% of isolates had detectable toxins. Further, they found that receiving antimicrobials prior to hospitalization and receiving immunosuppressive medications during hospitalization were risk factors for acquiring C. difficile. Other canine studies have not found an association between prior antibiotic therapy and detection of C. difficile toxins. Hospital-acquired C. difficile was associated with development of diarrhea in these animals. In cats, a study by Madewell et al found that 9.3% of tested cats in a veterinary teaching hospital had detectable C. difficile, and all infected cats had ≥1 of the following risk factors: antibiotic therapy, antineoplastic therapy, and immunosuppressive viral infection. Treatment for C. difficile infections in humans is typically vancomycin; other therapies, including administration of probiotics and increasing fiber in the diet have been investigated but not proven helpful. In dogs and cats, most infections are self-limiting, and the role of antibiotics for therapy is unclear. For patients with moderate to severe infections, metronidazole for 5 days is the drug of choice. While the risk of C. difficile transmission between dogs, cats, and owners is not yet fully understood, there is suggestion that this transmission does exist. In a study by Lefebvre, dogs who visit human healthcare facilities were 2.4x more likely to carry C. difficile than dogs who participate in other animal-assisted interventions, and dogs that licked human patients or accepted treats from human patients were at increased risk. A second study by Weese et al found that dogs living with immunosuppressed owners were at higher risk of C. difficile colonization, suspected to be via either direct or indirect transmission from their owners; however owners' feces were not tested in this study. Handwashing with soap and water is important to minimize or prevent spread of C. difficile, as alcohol-based sanitizers may not be fully effective against these organisms.

Campylobacter spp. are Gram-negative rods with a polar flagellum and curved shape. There are at least 37 species or subspecies of Campylobacter, with most being considered nonpathogenic. Campylobacter is reportedly the most commonly bacterial cause of diarrhea in humans. Isolation of Campylobacter is common from dogs and cats with and without diarrhea, with prevalence rates being highest in dogs <6months old, stressed, with poor hygiene, or concurrent parasites, or consuming homemade diets. Infection rates in dogs and cats vary depending on their living environment, with highest prevalence seen in dogs in rescue and boarding kennels. The most commonly identified species is C. upsaliensis but disease is thought to occur more often by C. jejuni and C. coli. A study by Burnens et al found an association between disease and presence of Campylobacter (C. jejuni and C. upsaliensis) in puppies (<1yr), but not in older dogs or cats. The pathophysiology of Campylobacter spp. includes adhesion to the jejunum and ileum, followed by multiplication and invasion of the epithelial lining. Organisms secrete two toxins, cytotoxin and a heat label-like toxin, which contribute to mucosal destruction and trigger inflammation. Secretory diarrhea is caused by an increase in cAMP as well as prostaglandins and leukotrienes, leading to electrolyte and water loss. Incubation of Campylobacter is 2-5 days, and clinical signs can include small or large bowel diarrhea (watery, mucoid, or hemorrhagic), vomiting, decreased appetite, fever, dehydration, and lethargy. Although Campylobacter-like organisms can be detected with cytology, this method is not acceptable for diagnosis as it cannot confirm Campylobacter or differentiate between species (pathogenic or not). Bacterial culture is the preferred method for diagnosis, and special media and biochemical testing at an experienced laboratory is required. Isolating C. jejuni or C. coli from a dog or cat with diarrhea can be used as a presumptive diagnosis of campylobacteriosis. While most cases are self-limiting, treatment is recommended for dogs and cats with moderate to severe clinical signs and to minimize zoonotic exposure. The treatment of choice is erythromycin 10-15mg/kg PO q8h; however fluoroquinolones, tylosin, tetracycline, and cefoxitin have also been recommended, with no studies available comparing efficacy of different antibiotic protocols. Treatment is not recommended for healthy carriers, as antibiotics may not clear infection and reinfection may occur. No studies have shown reduced public health risk by treating healthy carrier dogs or cats. Humans can acquire Campylobacter from handling or consumption of raw or undercooked poultry and less commonly other meats, unpasteurized milk, untreated water, contact with shedding pet dogs and cats (especially young animals and those with diarrhea), and foreign travel. Up to 80% of raw poultry is estimated to be contaminated with C. jejuni. Children are at higher risk of illness, and human illness is characterized by diarrhea, high fever, and cramps. Several studies have documented the same strain of Campylobacter in pets and their owners, and so zoonotic risk should be taken seriously, especially in households with puppies and young children.

E. coli can be isolated from the feces of nearly every healthy pet dog and human, and most strains are non-pathogenic. Certain E. coli possess virulence traits (toxins, adhesion properties, mechanisms for invasion and destruction) that lead to gastrointestinal disease, while other possess properties allowing extraintestinal infection (such as UTIs). Studies have shown that 3.1% of dogs on cattle farms carried E. coli O157:H7, and 7.8% of canine fecal samples collected from beaches carried E. coli O157:H7; however illness in dogs from O157:H7 is not fully understood, as most dogs with O157:H7 have been clinically normal. While rare, intestinal pathogenic strains of E. coli can cause diarrhea in pets, most commonly neonates. E. coli from raw meat is also suspected to play a role in cutaneous and renal glomerulonephritis in greyhounds and has been identified as a causative agent of histiocytic ulcerative colitis (HUC) in dogs. Diagnosis of E. coli diarrhea is complicated, as culture alone will not differentiate between pathogenic and non-pathogenic strains. Virulence testing and identification of E. coli within GI lesions (HUC) can provide more support to this diagnosis. Public health precautions include minimizing human and pet exposure to E. coli O157:H7 from farm animals or undercooked meat, maintaining clean environments for neonates and ensuring optimal colostrum, and encouraging good handwashing.

Salmonella spp., especially Salmonella enterica serotypes Enteritidis and Typhimurium, can cause diarrhea in humans and in pets. The prevalence of Salmonella spp. in the feces of clinically healthy or hospitalized dogs ranges from 1-63% and from cats 1-18%; however a true prevalence is most likely <5%. Pets and humans can acquire Salmonella spp. from raw or undercooked meat, contaminated food, treats, supplements, or water, ingestion of infected birds, or exposure to other infected animals (reptiles). Pets can be subclinical carriers or present with hemorrhagic diarrhea, fever, leukopenia, or septicemia. Diagnosis is made by consistent clinical signs and fecal culture. Recommended antibiotics for treatment of clinically ill pets include TMS, amoxicillin, or enrofloxacin, as well as isolation and supportive care. Pets that are subclinical or have mild clinical signs may have self-limiting disease and recover without antibiotics, as antibiotic therapy may prolong shedding and the carrier state. Pet owners should take precaution when handling raw or undercooked meat or eggs and avoid feeding these to pets, wash hands thoroughly after handling pet food, treats, supplements, and feces, and wash hands after contact with high risk animals (reptiles, poultry).

Antibiotic responsive diarrhea (ARD) is a relatively new, self-explanatory term for chronic or intermittent diarrhea in dogs without a specific known bacterial etiology. This type of diarrhea is suspected to be caused by changes in the GI flora, disrupted barrier function, alterations of the immune response to normal flora, or to undiagnosed pathogenic organisms. Previously, the term SIBO (small intestinal bacterial overgrowth) was used to label this condition; however this was considered by many to be a misnomer, as it is difficult to accurately measure the number of bacteria in the intestine and debate as to what is normal vs. overgrowth. While the diagnosis of ARD is simply remission of diarrhea following a treatment trial of antibiotics, other intestinal (parasitic, food allergy, etc.) and non-intestinal (EPI, hypoadrenocorticism, hepatic, etc.) causes of diarrhea should be investigated as well so that specific diseases can be identified and treated directly. Metronidazole (10mg/kg PO q8-12 hours) or tylosin (20mg/kg PO q8-12 hours) are used most commonly for ARD. While there is no evidence of zoonotic transmission from these patients, dog owners should be advised to wash their hands well after all contact with diarrhea in the event that a zoonotic agent is present.

Select references

Ruaux, C.G. and J.S. Suchodolski. Embracing Diversity: Microflora in health and disease. in 23rd Annual Forum of the ACVIM. 2005. Baltimore, MD. p. 583-585.

Suchodolski, J.S., et al., Application of molecular fingerprinting for qualitative assessment of small-intestinal bacterial diversity in dogs. J Clin Microbiol, 2004. 42(10): p. 4702-8.

Benyacoub, J., et al., Probiotics as Tools to Improve Health: Perspectives for Pets. Compend Contin Educ Vet, 2007. 29 Suppl 2A: p. 11-19.

Hirsh, D.C., The alimentary canal as a microbial habitat. in Veterinary Microbiology, D. Hirsh and Y. Zee, Editors. 1999, Blackwell Science, Inc.: Malden, MA. p. 61-64.

Swanson, K.S., Using Molecular Techniques to Study Canine and Feline Gut Microbial Ecology. Compend Contin Educ Vet, 2007. 29 Suppl 2A: p. 34-36.

Jones, S.L, and A.T. Blikslager. Role of the enteric nervous system in the pathophysiology of secretory diarrhea. J Vet Intern Med, 2002. 16(3): p. 222-8.

Marks, S.L. and E.J. Kather, Bacterial-associated diarrhea in the dog: a critical appraisal. Vet Clin North Am Small Anim Pract, 2003. 33(5): p. 1029-60.

Weese J.S., H.R. Staempfli, J.F. Prescott, et al. The roles of C. difficile and enterotoxigenic C. perfringens in diarrhea in dogs. J Vet Intern Med 2001;15:374-378.

Weese J.S. Bacterial enteritis and dogs and cats: diagnosis, therapy, and zoonotic potential. Vet Clin Small Anim 2011;41:287-309.

Weese, J.S., R Finley, RR Reid-Smith, et al. Evaluation of Clostridium difficile in dogs and the household environment. Epidemiol Infect 2010;138:1100-4.

Weese J.S. and J. Armstrong. Outbreak of Clostridium difficile-associated diarrhea in a small animal veterinary hospital. J Vet Intern Med 2003;17:813-6.

Clooten J, S. Kruth, L. Arroyo, et al. Prevalence and risk factors for Clostridium difficile colonization in dogs and cats hospitalized in an intensive care unit. Vet Micro 2008;129:209-214.

Madewell BR, Bea JK, Kraegel SA, et al. Clostritidum difficile: a survey of fecal carriage in cats in a veterinary medical teaching hospital. J Vet Diagn Invest 1999;11:50-54.

Lefebvre SL, Reid-Smith RJ, Waltner-Toews D, et al. Incidence of acquisition of MRSA, Clostridium difficile, and other healthcare-associated pathogens by dogs that participate in animal-assisted interventions. J Amer Vet Med Assoc 2009;234:1404-1417.

Parsons BN, Williams NJ, Pinchbeck GL, et al. Prevalence and shedding patterns of Campylobacter spp. in longitudinal studies of kenneled dogs. Vet Journal. 2010. Doi:10.1016/j.tvjl.2010.10.006.

Burnens AP, Angeloz-Wick B, Nicolet J. Comparison of Campylobacter carriage rates in diarrheic and healthy pet animals. Zentralbl Veterinarymed B. 1992;39:175-80.

Bender JB, Shulman SA, Averbeck GA, et al. Epidemiologic features of Campylobacter infection among cats in the upper Midwestern U.S. JAVMA 2005;226:544-7.

Leonard EK, Pearl DL, Janecko N, et al. Factors related to Campylobacter spp. carriage in client-owned dogs visiting veterinary clinics in a region of Ontario, Canada. Epidemiol Infect 2011; doi:10.1017/S0950268810002906.

Altekruse SF, Tollefson LK. Human campylobacteriosis: a challenge for the veterinary profession. JAVMA 2003;223:445-452.

Tenkate TD, Stafford RJ. Risk factors for campylobacter infection in infants and young children: a matched case-control study. Epidemiol Infect 2001;127:399-404.

Hostutler, R.A., et al., Antibiotic-responsive histiocytic ulcerative colitis in 9 dogs. J Vet Intern Med, 2004. 18(4): p. 499-504.

Hancock, D.D., et al., Multiple sources of Escherichia coli O157 in feedlots and dairy farms in the northwestern USA. Prev Vet Med, 1998. 35(1): p. 11-9.

Sanchez, S., et al., Animal issues associated with Escherichia coli O157:H7. J Am Vet Med Assoc, 2002. 221(8): p. 1122-6.

KuKanich KS. Update on Salmonella contamination of pet food and treats and safe feeding recommendations. J Am Vet Med Assoc. In press.