Prevalence of Enteric Pathogens in Dogs of North-Central Colorado
To evaluate the prevalence of enteric pathogens in dogs of north-central Colorado, fecal samples were obtained from client-owned dogs presented to the Veterinary Teaching Hospital at Colorado State University for evaluation of acute small-bowel, large-bowel, or mixed-bowel diarrhea (n=71) and from age-matched, client-owned, healthy dogs (n=59). Infectious agents potentially associated with gastrointestinal disease were detected in 34 of 130 (26.1%) fecal samples. Agents with zoonotic potential were detected in feces from 21 (16.2%) of 130 dogs and included Giardia spp. (5.4%), Cryptosporidium parvum (3.8%), Toxocara canis (3.1%), Salmonella spp. (2.3%), Ancylostoma caninum (0.8%), and Campylobacter jejuni (0.8%). Positive test results occurred in dogs with or without gastrointestinal signs of disease. Dogs, particularly those in homes of immunocompromised humans, should be evaluated for enteric zoonotic agents.
Introduction
Many gastrointestinal bacterial, viral, and parasitic infections are important to practicing veterinarians worldwide, owning to disease-inducing potential in dogs and zoonotic risk for humans. Diarrhea is one of the most common presenting complaints from dog owners; for example, over 100 dogs with diarrhea are assessed at the Colorado State University Veterinary Teaching Hospital yearly. However, dogs with and without signs of intestinal disease can harbor most of the enteric infectious agents.
In the past decade, the increasing prevalence of the acquired immunodeficiency syndrome (AIDS) in humans has led to increased concern about zoonotic disease. The canine intestinal zoonoses (in particular, Giardia spp., Cryptosporidium spp., Salmonella spp., and Campylobacter spp.) are considered to be of utmost importance.1–5 With an estimated 55 million dogs housed in 47.5% of the homes in the United States, potential for contracting zoonotic disease may be high.6 While many immunosuppressed humans are told to not own pets, the incidence of enteric zoonoses in client-owned dogs is essentially unknown. Since it is well documented that pet ownership improves the mental well-being of chronically ill people, veterinarians need more information concerning the zoonotic risk of dogs so clients can be accurately counseled.17
The purpose of this study was to define the prevalence of select infectious agents in the gastrointestinal tracts of dogs with and without diarrhea that were evaluated at a university clinic in north-central Colorado.
Materials and Methods
Experimental Design
Samples were collected from dogs between September 1997 and December 1998. Seventy-one consecutive canine admissions to the Colorado State University Veterinary Teaching Hospital with a complaint of acute diarrhea (median, 2 days; mean, 5.6 days) were entered into the study. For the majority of diarrhea cases entered into the study (n=59), the primary clinician submitted feces from an age-matched control without signs of gastrointestinal disease. Approximately 4 grams of fecal material were collected from each dog for assessment of potential enteropathogens. Assays were performed on feces collected within 24 hours; the majority of the assays were performed within 3 hours of collection. Samples not assayed on the same day were stored at 4°C.
Fecal Parvo Enzyme-Linked Immunosorbent Assay (ELISA) and Electron Microscopy
Fecal parvovirus antigen testing was performed using an ELISA test assay as recommended by the manufacturer.a Feces were prepared and examined for enteric viruses by use of scanning electron microscopy.b
Staining of Fecal Smears
A thin fecal smear was made, stained,c and examined microscopically under oil-immersion (1,000X) for the presence of spore-forming rods that were morphologically consistent with Clostridium perfringens (C. perfringens).b The term “significant number of spores” was reported by the reference laboratory if microscopic evaluation revealed >9 spores per high-power, oil-immersion field. A monoclonal antibody-based immunofluorescent antibody assay (IFA) for the detection of Cryptosporidium parvum (C. parvum) and Giardia spp. was performed as recommended by the manufacturerd on one thick smear made from each sample.
Fecal Flotation
Following zinc sulfate flotation of a Lugal’s iodine-stained fecal sample, microscopic examination was performed to identify parasitic ova, cysts, and oocysts.b
Fecal Culture
Approximately 1 gram of feces from each submission was cultured using tetrathionate broth, MacConkey’s agar, Hektoen agar, anaerobic blood agar, and Campylobacter agar with 10% sheep’s blood.8b,e
For Salmonella spp., two drops of iodine were added to the tetrathionate broth, and the samples were incubated for 12 hours at 42°C. The next day, the broth was plated onto Hektoen, brilliant green with sulfadiazine, and xylose lysine tersitol (XLT) four agar plates.e The plates were streaked for isolation and were incubated for 24 hours at 35°C. Isolated colonies were then subcultured on blood agar for 24 hours at 35°C, then were identified using a Micro ID.e Once Salmonella was confirmed, grouping serum was used to define the group of Salmonella present.f
For Clostridium spp., samples were cultured anaerobically using Centers for Disease Control (CDC) anaerobic agar.e Typical-appearing Clostridium spp. colonies were then plated to egg-yolk agar. Lecithinase-positive colonies were considered to be C. perfringens.
Campylobacter spp. samples were incubated at 35°C for 48 hours in a microaerophilic environment (10% carbon dioxide [CO2]; 2.5% oxygen [O2] with the balance in nitrogen). Suspect colonies were evaluated microscopically for morphological characteristics of Campylobacter spp.
Results
Of the 130 dogs, there were 12 with diarrhea and five controls <6 months of age, the age group usually associated with parvovirus enteritis. Infectious agents purported to cause gastrointestinal tract disease in dogs were identified in feces from 84 (64.6%) of 130 dogs [see Table]. Clostridium perfringens was cultured from 76 (58.5%) of the 130 fecal samples; spore-forming rods morphologically consistent with C. perfringens were reported as significant (>9 spores per high-power field) in 18 (13.8%) of the 130 samples; and C. perfringens enterotoxin was identified in 14 (10.8%) of the 130 samples.
Enteric viruses were detected in two dogs. Parvovirus was detected in one 14-week-old, unvaccinated puppy. This animal was coinfected with Toxocara canis (T. canis). Coronavirus was detected in feces from one healthy, 10-year-old Gordon setter. Dogs with toxocariasis ranged in age from 5 weeks to 5 years.
If only the presence of C. perfringens enterotoxin was considered significant, the overall prevalence of enteropathogens was 34 (26.2%) of 130 samples that included 13 (22.0%) of 59 nondiarrheic dogs and 21 (29.6%) of 71 dogs with diarrhea.
Multiple enteropathogens were identified in seven (5.3%) of 130 samples. Of these, six were from dogs with diarrhea and one was from a dog with normal stool.
Those agents with zoonotic potential were detected in feces from 21 (16.2%) of 130 dogs [see Table] and included Giardia spp. (5.4%), C. parvum (3.8%), T. canis (3.1%), Salmonella spp. (2.3%), Ancylostoma caninum (0.8%), and Campylobacter jejuni (0.8%). Most of the agents were detected in dogs with or without diarrhea. Of the three Salmonella isolates, a group E was cultured from a control dog, and a C-1 and an unknown type were cultured from dogs with diarrhea.
Discussion
Overall, enteropathogens were detected in 29.6% of dogs with diarrhea and in 22% of nondiarrheic dogs. In a similar study of humans, enteropathogens were detected in 56% of patients and in 16% of controls.9 Since only a single fecal sample was assessed and feces were stored for 3 to 24 hours prior to being assayed, it is possible the prevalence of some infectious agents was higher, and some agents may have been missed completely. For example, sensitivity of fecal examination following zinc sulfate centrifugation for detection of Giardia spp. is increased from 70% in infected dogs tested once to 93% in infected dogs tested twice.10 Some enteric bacteria are detected more readily using specialized techniques; in one study, Campylobacter spp. were demonstrated more frequently in human immunodeficiency virus (HIV)-positive patients when a membrane filter technique was used.11 Other enteric bacteria capable of infecting humans and dogs, such as Yersinia enterocolitica and Shigella spp., are not isolated on the media used here. Results may also vary based on geographical location, housing type, and population of animals tested. For example, while Salmonella spp. and Campylobacter spp. infections were uncommon in dogs described here, both puppies and adult dogs from a shelter commonly cultured positive.12
Enteropathogens with zoonotic potential were detected in 16.2% of the dogs in this study. Potentially zoonotic agents were detected in dogs with and without diarrhea; all seasons and all dog age groups were represented. north-central Colorado is semi-arid with subfreezing temperatures in the winter; it is likely that dogs from other climates have at least as high or higher prevalence of enteric zoonotic agents. The results of this study emphasize that fecal flotation, fecal evaluation for C. parvum, and fecal culture should be performed on samples from dogs in homes of immunosuppressed humans.
Giardia spp., C. parvum, T. canis, Baylisascaris procyonis, Uncinaria stenocephala, Ancylostoma caninum, Stronglyloides stercoralis, Echinococcus multilocularis, and Echinochoccus granulosa are the most significant enteric parasitic agents that infect both dogs and humans. Giardia spp. and Cryptosporidium spp. were the most prevalent infectious agents with zoonotic potential detected in this study. Both dogs and humans can develop diarrhea when infected with Giardia spp. or C. parvum; disease is most severe in immunosuppressed individuals.2 However, the zoonotic potential of canine isolates of Giardia and Cryptosporidium is currently unknown. Some Giardia1314 and Cryptosporidium1516 isolates from dogs vary genetically from human isolates. However, the canine Cryptosporidium genotype has been detected in immunosuppressed humans.17 Isolates of Giardia from humans sometimes fail to cause infection in dogs.14 In one study of HIV-infected humans with cryptosporidiosis, a statistical association with dog ownership was not detected.18 It is more likely that humans infected with Giardia or Cryptosporidium are infected from contaminated water or fomites rather than from direct contact with dogs. However, since both agents are associated with infection and disease in dogs and humans, and since microscopic examination cannot determine zoonotic potential of isolates, it seems prudent to assume all infected dogs are a potential zoonotic risk.
Due to the small size and small numbers in canine feces, diagnosis of Giardia spp. and Cryptosporidium spp. can be difficult following microscopic examination of fecal matter after flotation. The monoclonal antibody-based procedures for documentation of Giardia cysts and C. parvum oocysts in feces are validated for use with human isolates. It is unknown whether this assay consistently detects the genetically divergent canine isolates. In this study, Giardia spp. cysts were detected in 3.1% and 5.4% of the samples assessed by flotation and IFA, respectively. Cryptosporidium spp. oocysts were detected in five dogs by IFA, but none by flotation. All Giardia spp. flotation-positive samples were positive by IFA. Of the seven Giardia spp. IFA-positive samples, three were negative by flotation. One of the three positive-Giardia spp. IFA-positive, flotation-negative dogs was nondiarrheic, while the other two samples came from dogs with diarrhea. There are two possible explanations for these results. Either the IFA is associated with false-positive results when used with canine feces, or it is more sensitive than microscopic examination for Giardia cysts or Cryptosporidium spp. oocysts after a single zinc sulfate fecal flotation and is superior for the detection of subclinical carriers. Further studies are needed to differentiate between these possibilities.
In the United States, 2.8% of the general population and between 4.6% and 7.3% of children from 1 to 11 years of age have antibodies against T. canis, the cause of visceral and ocular larva migrans.19 Since the eggs of T. canis must larvate to be infectious, it is unlikely that direct contact with infected dogs results in infection of humans. However, infected dogs contaminate the human environment. Results of this study show that even in a nonendemic area, some dogs are shedding T. canis ova into the environment; therefore, fecal analysis should be a part of routine diagnostics, and deworming for T. canis based on these diagnostics should continue to be a part of preventative health programs for dogs.
The enteric bacteria of significance in dogs include Salmonella spp., Campylobacter spp., Shigella spp., C. perfringens, enterotoxigenic Escherichia coli, and Yersinia enterocolitica; each of the agents, excluding C. perfringens, are zoonotic for humans. Salmonella spp. are considered important zoonotic pathogens in dogs, but the majority of the reports of Salmonella spp. infections in dogs have been case reports or examinations of the incidence in shelters of stray animals rather than pet dogs.20 The reported prevalence of different serotypes of Salmonella noted at that time ranged from 1% to 36%, with an overall incidence of 10% estimated nationally.21 Evaluation of client-owned cats with and without diarrhea in the northern Colorado region revealed one of 129 cats to harbor Salmonella spp., suggesting that trends in Salmonella spp. carriage by client-owned domestic pets are lower than previously estimated.8 Exposure to animals with diarrhea was associated with a fourfold increase in the risk of Campylobacter jejuni (C. jejuni) and Campylobacter coli infection in humans;22 but none of the 76 cats with diarrhea in northern Colorado were infected with C. jejuni.8 Similarly, Yersinia enterocolitica and Shigella spp. are both associated with enteric disease in humans and have been isolated from dogs, but the incidence of infection is undetermined.23 Salmonella spp. and Campylobacter spp. infections of dogs in this study were uncommon. Prevalence rates may have been higher if multiple fecal cultures had been performed, if all fecal samples had been cultured immediately, or if transport media had been used. Alternately, the lower percentages of infected dogs compared to other reports may reflect differences in husbandry, feeding practices, region, and populations studied.24 All dogs positive for Campylobacter spp. or Salmonella spp. should be considered zoonotic.
While not considered to be zoonotic, C. perfringens enterotoxin-producing strains have been suggested to cause diarrhea in dogs. Clostridium perfringens was commonly cultured from the feces of dogs in this study. However, percentages of positive cultures, percentages of samples with morphologically consistent spore-forming rods, and percentages of enterotoxin-positive results were similar in healthy dogs and dogs with diarrhea. These results emphasize that none of the currently available tests correlate consistently with clinical disease and that results from each should be interpreted cautiously.25
Intestinal viral infections are thought to be common causes of diarrhea in dogs but have not been shown to be zoonotic. Canine parvovirus, coronavirus, calicivirus, paramyxovirus, picornavirus, rotavirus, astrovirus, and reovirus particles have been demonstrated in the feces of individual or small numbers of dogs with diarrhea.26–31 Since in-house laboratory tests are available only for canine parvovirus, the incidence of the other agents is largely unknown. Viral infections are likely to be more common in young animals; the majority of dogs studied here were >6 months of age, which may explain this study’s low incidence of positive electron microscopic evaluation or parvovirus antigen results. The authors’ failure to identify coronavirus in the feces of any dog with diarrhea in this study suggests that it was not a significant pathogen in pet dogs of the age group and the region studied.
Canine parvovirus antigen test kit; IDEXX Corp., Portland, ME
Veterinary Diagnostic Laboratory; Colorado State University, Fort Collins, CO
Quick-Dip; Mercedes Medical, Sarasota, FL
Merifuor Cryptosporidium/Giardia direct immunofluorescence assay; Meridian Laboratories, Cincinnati, OH
Culture media and Micro ID; Baltimore Biological Laboratories, Baltimore, MD
Salmonella O polyvalent antisera A-E/Vi; Baltimore Biological Laboratories, Baltimore, MD
Contributor Notes
