Abstract
Case series summary
This case series describes an outbreak of multidrug-resistant (MDR) Bordetella bronchiseptica in 16 shelter-housed cats with infectious respiratory disease. Four cats presented with acute dyspnea on the same day, each with a history of previous upper respiratory disease that had resolved with treatment. Early diagnostic testing and culture and sensitivity allowed for targeted antimicrobial therapy and environmental interventions. A case definition based on exposure and clinical signs identified 12 additional presumptive cases, including the likely index case. Comprehensive outbreak management included diagnostic testing, risk assessment, vaccination, use of isolation and quarantine, increased surveillance and review of biosecurity practices. The outbreak resolved in 26 days.
Relevance and novel information
Management of an MDR B bronchiseptica outbreak in shelter-housed cats has not been previously described. Along with standard population and environmental measures, early and appropriate use of necropsy, PCR and bacterial culture allowed rapid and appropriate use of effective, second-line antibiotics. Shelters are resource-challenged population centers. Veterinarians working in animal shelters can play an important role in helping to develop cost-efficient and effective antimicrobial stewardship practices for companion animal settings. Outbreak management expertise and funding for diagnostic testing, as well as application of the principles of antimicrobial stewardship, are essential components of shelter medicine practice.
Introduction
Upper respiratory disease (URD) in shelter-housed cats threatens health, welfare and adoptability. Disease results in extended lengths of stay, a higher daily census and reduced life-saving capacity.1,2 Clinical signs of URD include serous-to-mucopurulent ocular discharge, nasal discharge, conjunctivitis, sneezing, fever and inappetence. 2 The most common agents implicated in feline URD are feline calicivirus (FCV) and feline herpesvirus (FHV). Other prevalent agents include Chlamydia felis, Bordetella bronchiseptica and Mycoplasma species.2–4
Bacterial agents are usually considered secondary pathogens in feline URD; however, B bronchiseptica has been documented as a primary respiratory pathogen in cats.3,5 The reported prevalence in shelter-housed cats is 3–14%.6–9 Risk factors include high-density environments, poor animal husbandry, comorbidities and exposure to clinically affected dogs.2,7,10,11 B bronchiseptica-associated feline URD usually causes clinical signs, including fever, sneezing, ocular discharge and coughing. 2 Severe manifestations such as development of lower respiratory disease (LRD), pneumonia, dyspnea or death are more common in kittens <10 weeks of age.2,3
Several reports describe the epidemiology of B bronchiseptica in dogs and cats, including shelter populations.6–9,12–15 B bronchiseptica isolates have generally demonstrated susceptibility to amoxicillin/clavulanate, tetracyclines, doxycycline, trimethoprim–sulfamethoxazole and gentamicin.2,3,5–8 Doxycycline is recommended as empirical, first-line treatment because it is effective against B bronchiseptica, Mycoplasma species and C felis.16–18 Doxycycline is commonly used in animal shelters as the standard treatment for URD when bacterial infection is suspected. 2 However, B bronchiseptica isolates have demonstrated resistance to multiple classes of antibiotics, most commonly penicillins and cephalosporins.7,19–22 Resistance to doxycycline and other tetracyclines has also been identified.16,20–23
A bacterial pathogen demonstrating resistance to three or more classes of antibiotics is considered to be multidrug-resistant (MDR). MDR plasmids have been identified in B bronchiseptica isolates from healthy cats and dogs, and those with URD. Studies have focused primarily on the molecular epidemiology and genetics of resistance of the isolates rather than clinical impact or management.21,23
Infectious disease surveillance in shelters is critical for the timely recognition and management of outbreaks.24,25 Diagnostic testing for URD is not the standard protocol in shelters; limited resources are allocated to prevention and treatment measures.2,26 When a shelter experiences higher incidence or higher mortality than usual, diagnostic testing is advised.24,27 For URD, PCR testing is available for common respiratory pathogens, but results can be difficult to interpret in the face of recent vaccination, or in determining a definitive causative agent from commensal or carrier state.15,28,29 If an affected patient dies or is euthanized during an outbreak, a necropsy followed by ancillary testing of tissues is recommended as this is more likely to provide a more conclusive diagnosis. 30 Cytology and/or histopathology can confirm bacterial invasion and subsequent inflammation in affected tissues. 31 Culture and sensitivity help in identifying the causative agent and selecting the correct antimicrobial therapy. 32 However, animal shelters are often resource challenged, both in financial resources for diagnostics and shelter medicine expertise.
Shelter medicine, as a population-centered, preventive practice, relies on standard protocols, environmental interventions and biosecurity, in addition to pharmaceutical therapy, when managing infectious disease outbreaks. Every shelter should have written infectious disease and outbreak-management protocols, approved by their veterinarian, that address individual and population management, environmental controls and biosecurity. 24
This case series describes the clinical presentation and management of an infectious respiratory disease outbreak of MDR B bronchiseptica, which was resistant to doxycycline in shelter-housed cats. Although morbidity and mortality were higher than with more typical forms of URD, rapid recognition and management likely reduced the severity and impact of the outbreak.
Case series description
On 16 July 2020, a private, open-admission animal shelter encountered acute respiratory distress in four unrelated kittens aged 2–4 months. The shelter had an annual intake of 25,000 animals and employed one full-time veterinarian and one contracted medical director, who was also a veterinarian. Prior to 2020, the shelter’s live release rate approximated 65% for cats. Care for complex medical cases was greatly limited by shelter staffing and available medical resources, although both veterinarians had advanced shelter medicine training and expertise.
The four kittens were either owned by, or recently adopted from, the shelter. Each kitten had received a parenteral modified-live feline viral rhinotracheitis calicivirus panleukopenia vaccine upon intake. Each had been revaccinated at 2-week intervals, as recommended for shelter-housed kittens, 25 and had received at least two vaccines. Each kitten had previously been treated for URD in the shelter’s isolation ward. The clinical signs of the URD had resolved with doxycycline (10 mg/kg PO q24h for 7–10 days). On 16 July 2020, each presented in acute respiratory distress.
Cat A presented dyspneic, hypothermic (92.2°F) and hypoglycemic, with severe purulent nasal discharge. Supportive care included oxygen, intravenous fluids and dextrose. Cefazolin (22 mg/kg IV) and enrofloxacin (10 mg/kg IV) were administered for presumptive pneumonia and sepsis. Owing to the severity of cat A’s clinical signs, no response to stabilization and the inability to house oxygen-dependent cats, cat A was euthanized after 4 h. No necropsy or additional diagnostics were performed.
Cat B presented 4 h after cat A, dyspneic and febrile (104.2°F), with severe purulent nasal discharge. Oxygen saturation was 85%. Thoracocentesis did not yield fluid or air. After stabilization and antibiotics, cat B remained oxygen-dependent at the end of the day, was euthanized and a necropsy was performed in-house. Post-mortem examination revealed locally extensive dark-red, firm lung lobes affecting 75% of the lungs in a cranioventral pattern (Figure 1). Discharge was collected from the nasal cavity and cut surface of the lungs and submitted for PCR testing (IDEXX Feline Upper Respiratory Disease RealPCR Panel). Affected lung tissue was submitted for bacterial culture (IDEXX Reference Laboratories).

Necropsy images from cat B. (a) Left lateral view of thoracic and abdominal cavities demonstrating dark-red, mottled lung lobes, consistent with consolidation. The light-pink tissue of the right caudal lung lobe appears grossly normal. (b) Dorsal view of the lungs removed from the thorax, with consolidation of all lung lobes (dark red) except the right caudal lung lobe, which appears grossly normal
Cat C then presented, dyspneic and hypothermic (92.0F), with mild purulent nasal discharge and crackles in all lung fields. Cat C also remained oxygen-dependent, despite treatment, and was euthanized. In-house post-mortem examination revealed dark-red, firm lung lobes, affecting 50% of the tissue in a cranioventral pattern (Figure 2). Affected lung tissue was submitted for bacterial culture.

Necropsy images from cat C. (a) Left lateral view of thoracic and partial abdominal cavities demonstrating dark-red, mottled lung lobes consistent with consolidation. The light-pink tissue of the right caudal lung lobe appears grossly normal. (b) Dorsal view of the lungs removed from the thorax demonstrating dark-red lung lobe consolidation, with multifocal areas of grossly normal (pink) tissue in the right and left caudal lung lobes
Cat D presented while the other cats were receiving treatment. The kitten was bright, alert and responsive, with severe purulent nasal discharge and non-productive retching. It was prescribed azithromycin (5 mg/kg PO q24h), a nasal swab was collected for PCR and it was sent home with the foster parent given that there was no overnight care at the shelter or resources to refer to emergency care. The pet returned to the clinic 4 days later with improved but unresolved clinical signs.
Cat B’s PCR returned positive for FCV, B bronchiseptica and Mycoplasma felis. Cat D was positive for FCV, FHV-1, B bronchiseptica and M felis.
Lung cultures from cats B and C returned one isolate, B bronchiseptica, with identical resistance patterns (Table 1). The isolate showed resistance to cefpodoxime, ceftiofur, doxycycline and trimethoprim/sulfate, and intermediate susceptibility to gentamicin. The isolate demonstrated susceptibility to imipenem, chloramphenicol and multiple fluoroquinolones.
Antibiotic susceptibility profile of Bordetella bronchiseptica isolates from cats B and C
The minimum inhibitor concentration (MIC) breakpoints reported by IDEXX Refence Laboratories for this B bronchiseptica isolate are based on the Clinical and Laboratory Standards Institute breakpoints for doxycycline against Staphylococci species isolated from respiratory tissue. 33 They are reported as ⩽0.12 (susceptible), 0.25 (intermediate) and ⩾0.5 (resistant)
Not recommended or reported owing to low drug concentrations in the lungs per IDEXX Reference Laboratory
Not reported owing to a lack of standards for testing Gram-negative organisms in vitro as per IDEXX Reference Laboratory
NR = not reported
The three tested cats were diagnosed with MDR B bronchiseptica. A presumptive diagnosis for the first cat (cat A) was based on its history and clinical signs. A case definition was established based on clinical history, presence of URD or LRD, and elevated risk due to housing history and exposure to affected cats A–D. Because it is unknown whether the previous bouts of URD in cats A–D were related to the development of LRD, URD was included in the case definition. These criteria were used to perform a risk assessment for the remaining population.
During the outbreak, the facility housed 64 cats across nine rooms. Following standard guidelines for cat housing, the physical capacity of the facility is approximately 70 cats. 25 Analysis of housing over time demonstrated that all four affected cats had been housed in one of two rooms: a public-facing communal room and/or URD isolation containing eight kennels.
Using the case definition, 11 additional presumptive cases were identified, including a fifth cat (E) that had presented on 16 July (Table 2).
Clinical summary of 16 cats involved in an outbreak of multidrug-resistant bordetellosis
Age estimated by veterinary technicians using dentition
Diagnosed with pneumonia at a separate, non-affiliated clinic. Verbal conversation with treating DVM
Treatment with orbifloxacin (7.5 mg/kg PO q24h for 7 days) was initiated for the remaining affected cats. A twelfth cat (AA), the presumed index case, presented a week prior and was treated with the shelter’s doxycycline protocol, followed by azithromycin when the URD did not resolve. Azithromycin had been prescribed to three of the affected cats, including the index case and two cats that presented on 16 July 2020 (cats D and E) prior to the use of orbifloxacin. This medication was originally picked as an empirical choice given its efficacy against Mycoplasma species and use in previous studies.16,34 Once the bacterial culture was returned with the recommendation, and that azithromycin could not be reported (Table 1), treatment with azithromycin was discontinued.
Age estimates ranged from 2 months to 2 years, with a median age of 4 months (Table 2). All affected cats were domestic shorthair mixed-breed cats; three were females and 13 were males. Eight of 16 cats (50%) required some degree of inpatient care at the shelter’s clinic; the other eight were treated as outpatients, with adopters or at the shelter.
Movement of cats in affected rooms was halted. A clean break was established between exposed and unexposed populations. Exposed, asymptomatic cats were quarantined in the communal room. Clinical cases were hospitalized in the shelter’s clinic, placed in an URD isolation ward or treated outside the shelter with adopters. All rooms and enclosures were deep-cleaned with accelerated hydrogen peroxide cleaner (Rescue) in a two-step cleaning and disinfection process. 35 All asymptomatic cats exposed to test-positive cats were vaccinated with the feline intranasal attenuated live B bronchiseptica vaccine (Nobivac) 5 days after the initial cases presented.
Disease surveillance increased, with population rounds increasing to twice daily in all cat rooms. The shelter was already tracking the incidence of URD as part of their standard protocol. Because B bronchiseptica infections in cats are often a sign of poor husbandry,2,3 previously written handling and sanitation protocols in the shelter were reviewed. They were found to be appropriate for this outbreak, but staff compliance with the protocol was ensured. Supplies, including food, litter and clean laundry stored in the isolation room, were removed. Doors to isolation were kept locked to limit traffic by non-essential staff.
The shelter saw additional cases of URD during this period, in unexposed cats. These cats received the standard treatment with doxycycline, and all recovered without complication.
In total, 5/16 affected cats (31.3%) were euthanized or died due to pneumonia, including one kitten (K) that was seen by a community veterinarian and treated with cefovecin. The remaining 11 cats recovered and were adopted. The time from acute presentation of the four originally affected cats to when the last cat was released from treatment was 26 days. The epidemic curve demonstrated a point-source outbreak (Figure 3).

Epidemic curve of new cases over time in an outbreak of respiratory disease in shelter-housed cats
Discussion
While the literature provides multiple examples of B bronchiseptica isolates demonstrating antibiotic resistance in dogs, pigs and cats,19,20,23,36,37 this is the first report of managing an MDR B bronchiseptica infection outbreak in a shelter setting. The presentation of four cats with pneumonia on the same day, and following the previous treatment for URD, prompted concern for atypical infectious disease. Early recognition and diagnosis is critical to implementing disease-specific treatment protocols in populations, reducing morbidity and limiting mortality, and the duration of an outbreak.24,27
Diagnostic testing was employed early in this case. PCR testing is accessible to shelters experiencing URD outbreaks if they have the funds and expertise to collect samples. Interpreting results can be challenging, given that PCR is a highly sensitive testing modality and will recover agents which may not be actively causing infection at the time, or are a result of recent vaccination.15,28,29 In this outbreak, other common respiratory agents were recovered on PCR testing in addition to B bronchiseptica, and may have contributed to the progression of clinical signs.
More importantly, investigation in this case included necropsy and lower respiratory tract tissue sampling for culture. The growth of an MDR B bronchiseptica on lung culture allowed for identification of the organism in the lower respiratory tract and revealed its resistance to standard treatment (doxycycline). Ideally, histology would have been performed to confirm intracellular B bronchiseptica and its role as primary pathogen.
M felis was also identified on PCR. Mycoplasma species are fastidious slow-growing organisms that are difficult to culture, which is usually only available on special request. 38 This was not performed in this case. Mycoplasma species are susceptible to multiple fluoroquinolones, doxycycline and azithromycin; 16 therefore, M felis cannot be ruled out as contributing to this outbreak. However, M felis isolates are also commonly recovered from oropharyngeal swabs from healthy cats.2,17,28,38 Histopathology would likely have elucidated the role it played in this case.
Cats fitting the case definition received orbifloxacin after consideration of the culture and sensitivity report and based on the availability of product in the shelter’s clinic, the cost and its ease of administration. In retrospect, enrofloxacin would have been a more judicious choice given that the laboratory reported its minimum inhibitory concentration (MIC), and it has demonstrated both efficacy against B bronchiseptica and distribution to lung tissue in cats. 39 Selection strictly based on judicious-use principles would incorporate isolate-specific MIC data for each antibiotic, as well as pharmacodynamic and pharmacokinetic indices for the species being treated, to ensure efficacy and tissue penetration for the antibiotic at the site of infection. 40 Unfortunately, in veterinary medicine, data informing on antibiotic selection, even with culture and sensitivity testing, is often limited by a lack of knowledge of species-specific pharmacodynamics and pharmacokinetics (and uses human interpretive data at times), as well as limitations on the range of antimicrobials included in laboratory panels. Although the diagnostic laboratory provided an MIC for this isolate, this was based on interpretive data of doxycycline breakpoints for staphylococci in respiratory tissues. 33 The laboratory did not include orbifloxacin in its sensitivity panel. Fortunately, orbifloxacin proved to be clinically effective.
Vaccination was also utilized. The Bordetella species intranasal vaccine is non-core for shelter-housed cats.3,41 Only one affected cat (M) was vaccinated and then developed clinical signs; notably, this cat did not require advanced care or hospitalization. Four unaffected cats considered to be high risk were vaccinated; none developed clinical signs.
Outbreak management and antimicrobial stewardship incorporates more than pharmaceutical therapies. Unlike many shelter outbreaks, this shelter was not exceeding its capacity for care, at least in terms of physical housing capacity.42,43 The outbreak occurred in July 2020 when, due to the COVID-19 pandemic, the shelter only had 64 cats on site and intake was reduced to emergency only. During the same time frame in the previous 3 years, cats on site ranged from 89 to 298, exceeding physical capacity. The lower population likely contributed to the reduced morbidity and mortality. Being within the capacity for care in a shelter supports effective outbreak management and resolution, and minimizes the need for antimicrobial interventions.24,42
Additional control measures included performing a risk assessment, increased frequency of daily rounds, review of biosecurity practices with personnel and establishing a clean break between affected, exposed and unexposed populations. Outbreak management is a standard element in shelter medicine training; however, even with this expertise, population-level investigation and response can be challenging. Ongoing exposure of naive animals to infectious disease results in prolonged outbreaks, reoccurring with time. A point-source outbreak of short duration is the best possible outcome.
A noteworthy limitation of this response is that resources did not permit confirmatory testing for MDR B bronchiseptica in all affected cats. Cats fitting the case definition were elevated to treatment with second-line antibiotic therapy without confirmatory testing. In this outbreak, individual confirmatory testing was not pursued given the cost of diagnostics, increased handling of affected cats and availability of staff. In every outbreak, the case definition is commonly based on clinical signs and diagnostic testing, but the definition should be reviewed and revised as the outbreak progresses. 24
Antimicrobial resistance is an issue of concern in both veterinary and human medicine. Tackling it relies on applying optimal stewardship practices. Professional organizations and veterinary colleges have launched surveillance systems, research and guidelines to assist veterinarians in this pursuit.16,44–46
The US Food and Drugs Administration Center for Veterinary Medicine has made antimicrobial stewardship in veterinary medicine a priority, although initial efforts primarily target agricultural settings in terms of research, funding and regulation. 47 Supporting antimicrobial stewardship and judicious use principles remain a critical issue in both human and veterinary contexts. 48
Limited research has been performed in the USA on antimicrobial usage patterns in companion animal settings. These studies primarily frame antimicrobial selection in the context of highly individualized choices made by a veterinarian for individual patients under individual client pressures (eg, preferences and financial resources).45,46 A recent international literature review reported similar findings in other parts of the world. 49
There are increasing calls for a reframing of antibiotic use in veterinary practice as a product not solely of individual choices in an individual context, but as a result of systemic and structural pressures. These include economic conditions, infrastructure and supply chain issues, accessibility of veterinary care and disease-creating patterns in common animal husbandry practices. 49 As a population-based medicine, shelter practice balances individualized treatment within a population context and environmental factors. This reframing of antimicrobial stewardship applies to varied veterinary settings, but is highly relevant to animal shelters.
Heavily reliant on philanthropy, shelters are consistently challenged to provide veterinary care for populations of companion animals while under tremendous financial and staffing constraints. Municipal shelters have traditionally focused public funds on dog control and law enforcement. Municipal organizations often rely on support from non-profit partners and private donors to provide preventive medical care, or to address any issues related to feline health and welfare.50–54 Resources supporting antimicrobial stewardship in food animal production systems include subsidized testing programs and research on practice improvements. More resources and research are needed to support antimicrobial stewardship for companion animal populations, especially given their interface with human caretakers. Examples include elevating preventive medicine practices and increasing access to diagnostic testing and expertise.
Conclusions
Early recognition and response in this outbreak in shelter-housed cats combined limited diagnostic testing with population strategies to successfully manage an outbreak of an MDR B bronchiseptica in a highly vulnerable population. Doxycycline is the standard, and appropriate, first-line antibiotic for the treatment of feline URD when it warrants pharmaceutical intervention. However, this outbreak illustrates the importance of early and aggressive investigation when cases present with higher incidence and/or atypical clinical signs, especially in the face of previous treatment. This report also demonstrates how shelter medicine veterinarians employ comprehensive outbreak management as an essential component of treatment, reducing the incidence of disease and the need for additional pharmaceutical interventions. Veterinarians working in resource-challenged population centers, such as shelters, play an essential role in developing effective antimicrobial stewardship practices for companion animals.
Footnotes
Acknowledgements
The authors would like to thank Best Friends’ Animal Society for technical and financial support of the shelter.
Author note
The outbreak findings in this case series were presented as part of the Practitioner Case Presentations at the ABVP Symposium 2022.
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Ethical approval
The work described in this manuscript involved the use of non-experimental (owned or unowned) animals. Established internationally recognised high standards (‘best practice’) of veterinary clinical care for the individual patient were always followed and/or this work involved the use of cadavers. Ethical approval from a committee was therefore not specifically required for publication in JFMS. Although not required, where ethical approval was still obtained, it is stated in the manuscript.
Informed consent
Informed consent (verbal or written) was obtained from the owner or legal custodian of all animal(s) described in this work (either experimental or non-experimental animals, including cadavers) for all procedure(s) undertaken (either prospective or retrospective studies). For any animals or people individually identifiable within this publication, informed consent (verbal or written) for their use in the publication was obtained from the people involved.
