Stanley L. Marks, BVSc, PhD, DACVIM (Internal Medicine, Oncology), DACVN
Associate Professor, University of California, Davis, School of Veterinary Medicine
Davis, CA, USA
The clinical documentation of Clostridium perfringens, Clostridium difficile, Campylobacter spp. and Salmonella spp. as causes of diarrhea in dogs and cats is clouded by the presence of these organisms as a normal component of the indigenous intestinal microflora. Diagnosis of C. perfringens and C. difficile-associated diarrhea is traditionally made by detection of toxin(s) in the feces in association with clinical signs of enteritis, colitis, or enterocolitis. Diagnosis of Campylobacter and Salmonella-associated diarrhea is made on the basis of a positive fecal culture in association with clinical signs of enteritis. Stained fecal smears can be examined for the presence of gram-negative, curved, or "seagull-shaped" rods consistent with Campylobacter spp., however, caution should be exercised in interpreting these findings as many healthy dogs are asymptomatic carriers of Campylobacter spp.
Clostridium perfringens is an anaerobic, spore-forming, gram-positive bacillus that has been associated with outbreaks of acute, often severe diarrhea in humans, horses, dogs, and cats. The elaboration of four major toxins (a, ß, γ, and e) is the basis for typing the organism into five toxigenic phenotypes, A-E. Each type may also express a subset of at least 10 other established toxins, including C. perfringens enterotoxin (CPE), a well-characterized virulence factor whose production is co-regulated with sporulation. Clostridium perfringens enterotoxin has been shown to induce fluid accumulation and diarrhea in a dog model, and several recent studies have shown an association between detection of CPE in canine fecal specimens and the presence of diarrhea. Dogs with C. perfringens-associated diarrhea frequently exhibit large-bowel diarrhea characterized by increased frequency of bowel movements with tenesmus, fecal mucus and hematochezia; however, clinical signs of enteritis or enterocolitis are also commonly seen.
Current diagnosis of C. perfringens-associated diarrhea in dogs and cats is made based on detection of CPE in fecal specimens in conjunction with clinical signs of disease. The value of quantitative fecal culture and fecal spore counts have been shown to be of poor diagnostic value, as the organism is isolated from more than 75% of healthy dogs and there is no correlation between spore counts and detection of enterotoxin. There is only one commercially available ELISA kit (Techlab Inc., Blacksburg, VA) for detection of CPE in fecal specimens; however, the performance characteristics of this assay have not been validated in the dog or cat to date.
There is mounting evidence to support a causal relationship between elaboration of CPE and presence of diarrhea. In a recently completed study, we compared the detection of CPE in conjunction with the isolation of sporulated enterotoxigenic fecal isolates in 32 diarrheic and 100 non-diarrheic dogs. Clostridium perfringens enterotoxin was detected via ELISA in 12%, 14%, and 34% of fecal specimens collected from non-diarrheic outpatients, non-diarrheic inpatients, and diarrheic patients, respectively. Enterotoxigenic C. perfringens was isolated following heat shock of fecal specimens in 12% of non-diarrheic outpatients, 14% of non-diarrheic inpatients, and 41% of diarrheic dogs. Although there was a high incidence of CPE positive fecal specimens in the two non-diarrheic groups of dogs (12% and 14%, respectively), a very strong association was shown between the simultaneous detection of CPE and enterotoxigenic C. perfringens isolates and the presence of diarrhea. Furthermore, all but one diarrheic dog that had fecal specimens positive for enterotoxin or enterotoxigenic C. perfringens isolates had no other identifiable contributing causes for their diarrhea. We have documented similar findings in the cat, in which 51 healthy cats were all negative for CPE, whereas 9/63 (14.3%) diarrheic cats were positive for CPE.
The fact that in the abovementioned study, toxigenic strains were more likely to be found in a diarrheic specimen provides further evidence that CPE is associated with diarrhea in the dog and cat. This could suggest either a causative role for C. perfringens enterotoxin, or that the intestinal milieu in diarrheic patients selects for enterotoxigenic strains. It is plausible that in some animals, enterotoxigenic C. perfringens may be a part of the normal flora. In asymptomatic dogs and cats, sporulation may be occurring at a low rate, so the amount of enterotoxin being released by these commensal enterotoxigenic strains is too low to cause disease, or released enterotoxin is destroyed by proteases produced by the normal commensal bacterial population in the intestine. Diarrhea may then result from some unknown change in the intestinal environment (antibiotic exposure, dietary change, pH changes) that either induces increased sporulation of commensal C. perfringens (and a subsequent release of a large amount of enterotoxin by enterotoxigenic strains), or alters intestinal flora in such a way that destruction of released enterotoxin is dramatically reduced, or both.
A second study recently completed in the Companion Animal Gastrointestinal Laboratory at UC Davis evaluated fecal enteric panels performed on 260 diarrheic dogs, and showed a strong association between detection of CPE via ELISA and acute hemorrhagic diarrheal syndrome (AHDS). CPE was detected in the feces of 8/12 dogs (67%) that had clinical signs consistent with AHDS. Cases were excluded from a diagnosis of AHDS if there was any potential concurrent disease or historical finding that could reasonably be expected to cause AHDS, or if there were signs of colitis only. In addition, of the four dogs that had peracute symptoms and died as a result of the disease, all had fecal specimens positive for CPE.
The antimicrobials of choice for managing C. perfringens-associated diarrhea include the macrolides (particularly tylosin), ampicillin, and metronidazole. Tetracylcin is not recommended due to the high incidence of in vitro resistance documented against this antibiotic.
Clostridium difficile is a gram-positive, anaerobic spore-forming bacillus, and is the major cause of antibiotic-associated pseudomembranous colitis in human patients. C. difficile has also been associated with diarrhea and enterocolitis in foals and adult horses, as well as diarrhea in dogs. Two toxins, toxin A and toxin B, are thought to be primarily responsible for disease associated with the organism, although other toxins may also play a role. Current diagnosis of C. difficile-associated diarrhea is primarily made based upon detection of toxin A or toxin B in fecal specimens via ELISA. Isolation of the organism alone is not sufficient for diagnosis due to the presence of nontoxigenic strains. In addition, previous studies have reported no significant difference in the isolation of C. difficile from diarrheic and non-diarrheic dogs, although a recent study revealed detection of toxin A via ELISA to be significantly associated with the presence of diarrhea. This finding is underscored by the findings of Weese et al., who found that 21% of diarrheic dogs were positive by ELISA for toxins A and B, in contrast to 7% in nondiarrheic dogs. Multiplex PCR performed on fecal isolates for C. difficile toxin A and B genes proved to be unreliable for determining an association between toxigenic strains and the presence of diarrhea. Similar to C. perfringens, a strong association was found between the detection of C. difficile toxin A and the presence of AHDS, although no association was found between the prevalence of C. difficile toxin A and antibiotic administration in any of the populations. Although C. difficile-associated diarrhea is predominantly associated with hospitalization and antibiotic administration in humans, no significant association was found between antibiotic administration and C. difficile culture in a study conducted at UC Davis.
In summary, the strong correlation between the detection of toxin A produced by C difficile and clinical diarrhea, and the absence of the toxin in 103 dogs with normal feces supports the notion that C difficile is associated with diarrhea in dogs and may be a cause of it. Clostridium difficile seems to be less prevalent in cats compared to dogs. In a recently completed study in healthy and diarrheic cats at UC Davis, all 51 healthy cats were negative on both culture and ELISA for C. difficile, whereas 3/61 diarrheic cats (4.9%), and 4/63 diarrheic cats (6.3%) were positive for C. difficile on ELISA and culture, respectively.
The antimicrobial of choice for therapy of C. difficile-associated diarrhea is metronidazole. Vancomycin is an expensive antibiotic, and is only advocated for metronidazole-resistant cases.
Diarrhea produced by this small, curved, motile, microaerophilic, gram-negative rod is seen primarily in younger animals, although it has been seen in animals of all ages. It can be isolated from the feces of a high percentage (approximately 40%) of healthy animals that have been kenneled, particularly in animal control facilities. Poultry and poultry products are a major source for human beings, as well as unpasteurized milk (the organism can be shed in milk). Puppies and kittens are a source for humans, however, human beings may also be a source of infectious organisms for dogs and cats. Campylobacter can survive for days in surface water and as long as 4 weeks in feces. The duration of excretion in infected dogs and cats can be as long as 4 months and infected animals should be quarantined away from children during this period.
Campylobacter jejuni can colonize the jejunum, ileum, cecum, and colon; however, histologic changes are largely restricted to the colon. The organism adheres to the intestinal epithelium via an outer surface protein and produces an enterotoxin that results in a secretory diarrhea mediated by cyclic AMP. Campylobacter organisms also elaborate a cytotoxic enterotoxin that is probably responsible for the epithelial damage. Clinical signs range from mild transient diarrhea to mucous laden bloody stools with associated signs of colitis.
Reported rates of isolation of Campylobacter spp. from dogs with diarrhea range from 10.5% to 18.7%, whereas in these same studies the isolation rates from dogs with normal feces ranged from 0% to 11.1%. Unfortunately, the culture methods employed in many of these studies were aimed at primarily isolating C jejuni, a known enteric pathogen in humans that is catalase positive. Putative pathogenic species such as C upsaliensis could have been discarded on the basis of a negative catalase reaction. Several authors have suggested that Campylobacters may not be primary enteric pathogens in dogs because of the high isolation rates in nondiarrheic dogs. Genotypic studies are warranted in healthy and diarrheic dogs to facilitate differentiation of Campylobacter spp. It is highly plausible that certain Campylobacter spp. are more pathogenic than others.
We recently evaluated the prevalence of Campylobacter in healthy and diarrheic cats at UC Davis, and were surprised to find that 10/51 healthy cats (19.6%), and 7/63 diarrheic cats (11.1%) were positive for Campylobacter on fecal culture. These findings underscore the fact that many Campylobacter spp. are non-pathogenic and further genotypic studies are warranted to differentiate pathogenic from non-pathogenic species.
Antibiotic therapy should be used judiciously in animals with suspected Campylobacter-associated diarrhea, as many cases will self-resolve with supportive therapy. The antibiotics of choice are the macrolides (erythromycin azithromycin) and enrofloxacin.
Salmonella was infrequently isolated (< 2.3%) from non-diarrheic dogs presenting to the UC Davis School of Veterinary Medicine, and was also uncommon in diarrheic dogs (3 of 260 dogs or 1.2%). These results are very similar to those in cats in which the prevalence of Salmonella in 51 healthy and 63 diarrheic cats at UC Davis was 0%. In the dog study, one of the dogs was referred with a 1-year history of intermittent diarrhea that had recently become hemorrhagic, but the dog was healthy otherwise. Results of CBC were within reference ranges. Salmonella ser Anatum was cultured and the diarrhea resolved during a 3-week administration of trimethoprim-sulphadiazine and dietary management. The second dog was referred with a history of 3 episodes of vomiting and diarrhea during a 2-month period, the first 2 of which responded to conservative treatment. Bacteriologic culture of feces was performed at the time of the third episode and S Thompson was isolated. Results of CBC at that time were within reference ranges. No specific treatment was given and the vomiting and diarrhea resolved with dietary management. The third dog developed diarrhea within 12 hours of being discharged from the hospital after having been sedated for fine-needle aspiration of an area of pulmonary consolidation. Profuse, watery diarrhea developed in association with fever (rectal temperature, 40 C [104 F]), signs of depression, and a moderate leukocytosis with left shift and toxic changes. Salmonella Enteriditis was recovered from the feces in association with positive results for CPE via ELISA. The diarrhea became hemorrhagic, which coincided with clinical deterioration and decreased hematocrit and serum albumin concentration. After 48 hours of treatment including administration of enrofloxacin, ampicillin, metronidazole, and fresh frozen plasma, the dog was euthanatized because of continued deterioration. No necropsy was performed. Among the 72 control dogs with normal feces, no Salmonella spp were isolated from any sample.
The antibiotics of choice for managing patients with Salmonella-associated diarrhea include the fluoroquinolones (enrofloxacin and ciprofloxacin). Since Salmonella can survive in the phagocytic cell, the antimicrobial drug should be one that penetrates the cell. Examples of those that distribute in this manner include ampicillin, enrofloxacillin, trimethoprim-sulfonamides, and chloramphenicol.
In summary, the isolation of Clostridial spp., Campylobacter spp. and Salmonella in diarrheic dogs and cats must be interpreted cautiously. The diagnosis of Clostridial-associated diarrhea is facilitated by use of ELISA tests to determine the presence of enterotoxin of C. perfringens or toxin A and/or B of C. difficile. Campylobacter-associated diarrhea is difficult to prove as the organism is commonly carried asymptomatically in healthy dogs and cats. Further genotypic studies are warranted to help speciate the organism and elucidate the pathogenicity of these species. Although Salmonella is infrequently isolated in diarrheic dogs, the presence of fever, hemorrhagic diarrhea, and leukocytosis with toxic neutrophils or neutropenia in a patient warrants fecal culture for Salmonella.
References are available upon request.