Prevalence of Bordetella bronchiseptica in dogs with respiratory disease

B. bronchiseptica is a bacterial pathogen frequently involved in respiratory disease in various animal species, also sometimes affecting immunosuppressed humans (Echeverri-Torro et al. 2015, Rampelotto et al. 2016). In dogs, the agent has traditionally been associated with the canine infectious respiratory disease complex (CIRD), in which it is thought to play a key role as a primary pathogen, frequently in conjunction with viral co-infections (Decaro et al. 2004, Keil and Fenwick 1998, Mochizuki et al. 2008). In experimental studies, it could be shown that the organism is able to induce clinical respiratory disease in dogs as a single agent (Thompson et al. 1976). Clinical signs can range from mild respiratory disease to severe pneumonia and death, depending on immune status, vaccination history, and viral and bacterial co-infections (Keil and Fenwick 1998). Dogs can exhibit acute or chronic signs of respiratory disease, and clinical courses of up to two years have been described (Canonne et al. 2016). B. bronchiseptica has also been detected as a commensal bacterial organism in the upper respiratory tract of healthy shelter (Lavan and Knesl 2015) and privately owned dogs (Schulz et al. 2014), and could be isolated from the lower airways of healthy shelter dogs (Chalker et al. 2003). In addition, in a recent study using quantitative PCR it could be shown, that B. bronchiseptica could also be isolated from several dogs with eosinophilic bronchopneumopathy (EBP) and chronic bronchitis (Canonne et al. 2018) and several privately owned asymptomatic dog (Canonne et al. 2016, 2018). Canonne and coworkers (2018) were able to show, that moderate and severe B. bronchiseptica-loads were only present in dogs with EBP, in which a positive detection was associated with a higher disease severity index and higher neutrophil counts on bronchoalveolar-lavage-fluid (BALF) (Canonne et al. 2018).

To date, it is still unknown why some dogs develop clinical disease associated with B. bronchiseptica-infection and others just seem to be asymptomatic carriers of the organism. A recent study using quantitative PCR for detection of B. bronchiseptica in bronchoalveolar lavage fluid (BALF) demonstrated high bacterial DNA levels of the organism in dogs with clinical disease in contrast to healthy control dogs (Cannone et al. 2016). 

A study using a lipopolysaccharide (LPS) antigen-based enzyme-linked immunosorbent assay to measure the immune response in B. bronchiseptica-infected shelter dogs with and without respiratory signs showed that antibodies to LPS did not correlate with the development of clinical disease, disease severity, or colonization of the lungs (Chalker et al. 2003). Persistence or clearance of the bacterium from the respiratory tract probably depends on the host immune response and pathogenic properties of the bacterium. Furthermore, certain characteristics of the bacterium itself, including adhesion molecules, such as fimbriae and filamentous hemagglutinin causing ciliostasis and epithelial damage, seem to play an important role in the pathogenesis of infection (Anderton et al. 2004, Bemis et al. 1977). 

The aim of the study was to investigate the involvement of B. bronchiseptica in upper and lower respiratory samples of dogs with various types of respiratory disease and in dogs without evidence of respiratory disease using PCR and culture as two different detection methods.

All participating dogs were patients presented to the Clinic of Small Animal Medicine of the Ludwig-Maximilians-Universität Munich, Germany. Pharyngeal swabs of dogs with respiratory disease (RD) were taken with the owners` consent. Samples included BALF leftover material from the diagnostic work-up. Samples of the control dogs were obtained shortly after euthanasia for reasons not related to the study.

The study population included 26 dogs that were presented for various types of upper or lower respiratory conditions [bronchopneumonia (8), chronic bronchitis (4), CIRD (3), laryngeal paralysis (3), airway collapse (one dog with tracheal collapse, one dog with tracheal collapse and diffuse bronchomalacia), pulmonary fibrosis (2), eosinophilic bronchopneumopathy (2), oronasal fistula (1), and reverse sneezing (1)]. All patients underwent bronchoscopy and BAL as part of their diagnostic work-up. Exclusion criteria were evidence of cardiac disease, neoplasia, or pleural effusion based on the diagnostic work-up. Besides physical examination, for most dogs with RD thoracic radiographs in right lateral and ventrodorsal position (22/26 dogs) and various other investigations were performed, based upon assessment of the primary clinician to obtain a diagnosis. 

Sixteen dogs that were euthanized for different diseases that were not affecting the respiratory tract [neoplasia (7), neurological disease (6), renal failure (2), haematological disease (1)] were included in the group of dogs without signs of RD. Procedures performed in these dogs were physical examination and post mortem sampling (pharyngeal swabs and BAL). Exclusion criteria were respiratory signs reported within the last six months, clinical respiratory signs upon presentation, and abnormalities on BAL-fluid (BALF) analysis (cytology and culture). 

While samples of the dogs with RD were taken during routine work-up, samples of control dogs were obtained within ten minutes after euthanasia. PCR for B. bronchiseptica was performed on BALF samples of all dogs, pharyngeal swabs for B. bronchiseptica PCR were taken in 22/26 dogs with RD and in all 16 control dogs. For pharyngeal sampling, sterile cotton swabs (Sarstedt AG & Co., Numbrecht, Germany) were gently rotated in the pharyngeal area and frozen at -20 °C afterwards until analysis. 

In dogs with RD, BALF was taken endoscopically over the working channel of the endoscope. Endoscopy was performed using two different types of flexible bronchoscopes depending on the size of the dog (type 60001 VL1 and type 60002 VB1, Storz GmbH and Co. KG, Tuttlingen, Germany), that were either passed into the airways over a sterile endotracheal tube or a sterile laryngeal mask. All procedures were performed wearing sterile gloves. For BAL, 2–5 ml of warm saline (0.9% NaCL) solution were instilled into two different lung lobes, depending on location and type of respiratory disease. For collection of the fluid, a sterile container (Lukens Schleimprobenbehälter^®, Olympus GmbH, Hamburg, Germany) was used, that was interconnected between working channel of the endoscope and a mechanical suction apparatus. Recovered BALF of both locations was pooled after collection for each dog.

BALF material in dogs without signs of RD was collected over a sterile catheter (Braun, Melsungen AG, Melsungen, Germany), that was inserted into a sterile endotracheal tube and gently forwarded into the lower airways until resistance was noted. For lavage, 5–10 ml of sterile saline solution were instilled and recovered by manual suction with a sterile syringe. 

BALF of all dogs was submitted for bacterial culture, B. bronchiseptica culture and PCR, and cytological examination. 

All BALF cytology samples were evaluated by the same person (S. L.). Cytocentrifuge-concentrated smears (5 minutes at 1000 rpm) were prepared immediately after collection and stained with a modified Wright’s stain after air drying. A light-optical microscope (Olympus BX 51, Olympus Deutschland GmbH, Hamburg, Germany) was used for cytological evaluation. Interpretation of BALF cytology was performed based upon published reference values (Hawkins et al. 1990, Rajamäki et al. 2001). BALF cytology was interpreted as normal, if predominating cells were alveolar macrophages, and the remaining cells consisted of 15 per cent lymphocytes, 10 per cent neutrophils, 5 per cent eosinophils, and 1 per cent epithelial cells (Hawkins et al. 1990, Rajamäki et al. 2001). Dogs were excluded from the study, if BALF showed cytological signs of oropharyngeal contamination such as presence of Simonsiella spp. or squamous cells.

All bacterial cultures were performed at the Institute for Infectious Diseases and Zoonoses of the Ludwig-Maximilians-Universität Munich. For aerobic and facultative anaerobic cultures, Gassner and Rambach agar and sheep blood agar with colistin and nalidixin acid were used and incubated at 37 °C for 48 hours. Bordet-Gengou agar was used as a culture plate for detection of B. bronchiseptica and was incubated under the same conditions as described above. Results were provided using a semiquantitative description [(+) to +++].

Freshly collected BALF samples and pharyngeal swabs with a maximum storage time of 4 hours at 4 °C were used for DNA extraction as described before (Schulz et al. 2015). Briefly, 1.5–2.0 ml of BALF were treated with 30 µl acetylcystein (ACC^® Injekt Injektionslösung 300 mg/3ml, Hexal AG, Holzkirchen, Germany) for several minutes to reduce viscosity. The sample was centrifuged for 2 minutes at 5000 rpm, and 1 ml phosphate buffered saline (PBS) (phosphate buffered saline^®, Sigma Life Science, St. Louis, USA) was added to remove the acetylcystein. After centrifugation, 0.8 ml were removed and DNA was extracted from the cell pellet in the remaining volume using the QIAamp^® DNA Mini Kit according to the manufacturer’s instructions (Blood or Body Fluid Spin Protocol). DNA from the swabs was extracted according to the QIAamp^® DNA Mini Kit swab protocol. DNA of BALF and pharyngeal swabs was eluted in 100 µl AE buffer and stored at -20 °C. 

PCR reactions were run on a 7500 Real Time PCR System^® (Applied Biosystems, Applera Deutschland GmbH, Darmstadt, Germany) in a total volume of 30 µl using 15 µl QuantiTect^®SYBR^®Green PCR MasterMix (QIAGEN, Hilden, Germany), 11.6 µl RNAse free water, 0.2 µl forward primer (25 pmol/µl), 0.2 µl reverse primer (25 pmol/µl), and 3 µl template. The PCR was performed at an annealing temperature of 60 °C for 40 cycles followed by melt curve analysis to determine specific amplification.
A PCR reaction for the canine MRPS7 gene was performed to confirm successful DNA extraction using the primers MRPS7-F and MRPPS7-R. For detection of B. bronchiseptica, the primer pair Bordetell_bpp_F and Bordetell_bpp_R was used. 

GraphPad Prism^® Version 5.00 (GraphPad Software, San Diego, USA) was used for all comparisons. Age between groups was compared by non-paired t-test. Comparison of the prevalence of B. bronchiseptica between patients with and without RD was performed using Fisher`s exact test. P-values  0.05 were considered significant for all tests.

While the median age of dogs with RD was 6 years (0.4–14.2 years), the median age of dogs without RD was 9 years (3.1–17.4 years) (p = 0.101). Dogs with RD included 10 male and 16 female dogs. The group of dogs without RD consisted of 12 male and four female dogs.
Clinical signs of the dogs with respiratory disease included nasal discharge (7/26, 26.9%), increased respiratory sounds (22/26, 84.6%), dyspnea (5/26, 19.2%), and cough (17/26, 65.4%). Thoracic radiographs in dogs with RD were interpreted as unremarkable in 5/22 (22.7%), 4/22 (18.2%) dogs showed an alveolar, 4/22 (18.2%) a bronchial, 2/22 (9.1%) an interstitial, and 7/22 (31.9%) a mixed lung pattern.

In the group of dogs with RD, 7/26 (26.9%) showed a neutrophilic inflammation, 5/26 (19.3%) a neutrophilic-eosinophilic inflammation, 2/26 (7.7%) an eosinophilic inflammation, 6/26 (23.1%) had increased amounts of mucous and/or Curschmann`s spirals, and 6/26 (23.1%) showed a physiological BALF cytology. Dogs with a physiological BALF included patients with reverse sneezing (1), CIRD (1), laryngeal paralysis (2), tracheal collapse and diffuse bronchomalacia (1), and oronasal fistula (1). In the group of dogs without signs of respiratory disease, all 16 dogs had a physiological BALF cytology (inclusion criterion).
In dogs diagnosed with bronchopneumonia (8), five had a neutrophilic inflammation and three a neutrophilic inflammation with eosinophilic component. Three B. bronchiseptica-culture-positive dogs showed degenerative neutrophils and bacteria on BALF cytology.

Results of B. bronchiseptica PCR and culture for both groups are displayed in table 1. When dogs with RD and dogs without signs of RD were compared, significantly more dogs with RD (21/26; 80.8%) were PCR-positive in their BALF for B. bronchiseptica than dogs without RD (5/16; 31.3%) (p = 0.003). All four culture positive dogs were also PCR positive for B. bronchiseptica in their BALF, and three of them were PCR positive for the organism in pharyngeal swabs as well. Dogs with RD with a PCR positive BALF for B. bronchiseptica included patients diagnosed with bronchopneumonia (6/21), chronic bronchitis (4/21), eosinophilic bronchopneumopathy (EBP) (3/21), CIRD (2/21), laryngeal paralysis (2/21), and one each with a diagnosis of idiopathic pulmonary fibrosis, airway collapse, reverse sneezing of unknown origin, and oronasal fistula. In the group of dogs with RD, all seven dogs with a PCR positive pharyngeal swab were also PCR positive in BALF, while twelve dogs were PCR positive in BALF, but negative in pharyngeal swab (two dogs without pharyngeal sampling). In the control group, of eight pharyngeal-positive dogs, four were positive in BALF-PCR. One BALF-PCR positive dog tested negative in the pharyngeal sample. The four dogs with a positive culture for B. bronchiseptica were diagnosed with bronchopneumonia (3) and EBP (1). The dog with EBP had an eosinophilic-neutrophilic inflammation on BALF cytology. Of the eight dogs diagnosed with bronchopneumonia, three were PCR positive for B. bronchiseptica in pharyngeal swabs, and seven were positive in BALF.

If BALF-samples of both groups were combined, B. bronchiseptica was detected significantly more frequently by PCR (26/42, 61.9%) than by culture (4/42, 9.5%) (p  0.001). 

Additional bacteria cultured in dogs with bronchopneumonia were Pasteurella sp. (2), Staphylococcus epidermidis (1), Neisseria sp. (1), and Enterobacter cloacae (1). In addition, 5/8 dogs were PCR positive for Mycoplasma sp.

B. bronchiseptica is a well-known respiratory pathogen associated with CIRD, chronic bronchitis, and bronchopneumonia in dogs (Canonne et al. 2016, Decaro et al. 2016, Johnson et al. 2013, Keil and Fenwick 2000, Mitchell et al. 2017, Mochizuki et al. 2008, Taha-Abdelaziz et al. 2016). However, recent studies have shown that the organism can also be detected in upper respiratory samples of healthy dogs. A study investigating different respiratory pathogens in nasal and pharyngeal swabs of dogs with CIRD and healthy control dogs showed a PCR prevalence of 45.6% for B. bronchiseptica in healthy in contrast to 78.7% in diseased dogs (Schulz et al. 2014). In a study performed at shelters in different parts of the USA, 19.5% out of 503 asymptomatic dogs were PCR positive in ocular and oronasal swabs for B. bronchiseptica (Lavan and Knesl 2015). Similarly, in the present study 50% of dogs without evidence of RD and 31.8% of dogs with RD were PCR positive for B. bronchiseptica in pharyngeal swabs, underlining the role of the bacterium as a commensal organism in this area. These findings indicate that results of PCR testing of upper airway samples for B. bronchiseptica should be interpreted with caution in dogs with RD, since oropharyngeal carriage of the organism is common and a positive PCR result could lead to unnecessary antibiotic treatment in such a case. PCR detection rates of 80% in BALF were significantly higher for B. bronchiseptica compared to 30% in pharyngeal swabs in this study, indicating, that upper airway sampling alone might miss an infection or colonization of the lower airways and therefore cannot be recommended for testing in symptomatic dogs. 

The present study shows that privately owned dogs with different types of respiratory disease were more commonly PCR positive for B. bronchiseptica in BALF samples than dogs without evidence of RD; however, still 31.3% of dogs without RD were positive for the organism in their lower airways. In a recent study by Canonne and coworkers (2016) investigating diagnostic tests in 24 dogs with chronic B. bronchiseptica infection, a low level of B. bronchiseptica DNA was detected by quantitative PCR in only one out of ten healthy control dogs from private households. In contrast to that, in a group of 152 dogs euthanized in a rehoming center in the UK, 39% of dogs without signs of RD were PCR positive for B. bronchiseptica in lung washings performed post mortem in comparison to a detection rate of 52% in dogs with RD (Chalker et al. 2003). It can be assumed that the prevalence of colonization of the lower airways with the bacterium probably depends on individual (age, immune status, concurrent diseases) and environmental factors (housing, population density, infection pressure, stress). It would have been interesting to perform a quantitative PCR for B. bronchiseptica to evaluate differences in copy numbers between dogs with and without RD in the present study. Considering the results of Canonne and coworkers (2016), it can be assumed, that dogs with infectious or inflammatory RD probably harbor higher numbers of the bacterium in their lower airways than dogs without inflammatory airway disease. Potentially airway disease impairs the normal respiratory defense mechanisms, thus facilitating colonization of mucous membranes with B. bronchiseptica. However, sampling techniques in the present study for diseased and control dogs could not be standardized, which affected quantity of fluid retrieval and potentially also quality of samples. For that reason, conventional PCR might have been a better method to compare samples of both groups than quantitative PCR, which relies very much on the ability to standardize sampling procedures to ensure comparability.

It has been shown in studies investigating pathogens associated with CIRD that infection with B. bronchiseptica is commonly associated with coinfections with respiratory viruses and Mycoplasma sp. (Mitchell et al. 2017, Schulz et al. 2014). Five of the eight dogs diagnosed with bronchopneumonia in this study were PCR positive for Mycoplasma sp. In BALF, and seven dogs were positive for B. bronchiseptica, in addition, several other bacterial pathogens were detected. In a study by Canonne and coworkers (2016), PCR detection for Mycoplasma cynos did not differ between groups of dogs tested positive for B. bronchiseptica in BALF and healthy control dogs. It would have been interesting to evaluate viral pathogens in the present study as well; however, it is unlikely that they played a role in patients with chronic respiratory disease. 

A positive culture result for the organism was associated with a neutrophilic inflammation in BALF cytology in the four culture-positive dogs in the present study, and interestingly, one of the four had an eosinophilic-neutrophilic inflammation and was diagnosed with EBP. In a study looking at quantitative B. bronchiseptica loads by PCR in dogs with EBP, chronic bronchitis and healthy dogs, only EBP was associated with a high copy number of the organism, and BALF from B. bronchiseptica positive EBP patients had higher numbers of neutrophils than BALF from negative ones, possibly indicating an association between EBP and B. bronchiseptica infection (Canonne et al. 2018). While the patient with EBP was a three year old dog with chronic respiratory signs, the three other culture positive dogs were younger than one year and diagnosed with acute bronchopneumonia based on clinical, laboratory and radiographic results. In other studies, B. bronchiseptica pneumonia was also associated with young age and severe clinical signs (Batey and Smits 1976, Radhakrishnan et al. 2007). In the present study, total cell counts and differential cell counts of BALF were not performed in most dogs. It would have been interesting to compare inflammatory cell ratios between groups of dogs with and without PCR detection of B. bronchiseptica to evaluate, if carriage of the organism can be associated with an inflammatory response potentially triggering chronic inflammatory diseases.

In addition to dogs with inflammatory or infectious airway disease, also dogs with laryngeal paralysis, airway collapse, reverse sneezing, and oronasal fistula and one third of the control dogs were PCR-positive for B. bronchiseptica in BALF samples, representing a significantly higher number than published for healthy control dogs in the study by Canonne and coworkers (2016). Although in none of the dogs oropharyngeal contamination was indicated by the presence of squamous cells or Simonsiella spp. on BALF cytology, aspiration of oropharyngeal secretions ante or post mortem potentially could explain the relatively high number of positive samples. Although this has not been shown for B. bronchiseptica so far, a study looking at the relation between oropharyngeal contamination and detection of Mycoplasma organisms was able to show that contamination with upper airway secretions does facilitate detection of these organisms (Chan et al. 2013). Diseases such as laryngeal paralysis and oronasal problems with hypersecretion could potentially lead to silent aspiration of oropharyngeal bacteria as well. The control group chosen in this study consisted of dogs, that did not show signs of RD on history and physical exam and that had a physiological cell population on BALF cytology; however, these dogs were suffering from different other systemic diseases potentially influencing the immune response of the respiratory tract and facilitating colonization with bacteria such as B. bronchiseptica. Furthermore, pretreatment with immunosuppressive agents and antibiotics could have influenced the detection rates of B. bronchiseptica in both groups. Therefore, results obtained from this population of sick dogs probably do not represent the real prevalence of B. bronchiseptica in healthy dogs. 

Of the four dogs with upper airway disease, two had nasal disease and were PCR negative for B. bronchiseptica in BALF. While in one dog with reverse sneezing no pharyngeal sample for PCR had been taken, another dog with nasal disease (oronasal fistula) tested positive in the pharyngeal swab. However, it should be noted, that in case of upper airway sampling in dogs with nasal disease, the nasal cavity would be the preferred sampling site instead. 

When both detection methods, culture and PCR, were compared regarding their ability to detect B. bronchiseptica in the lower respiratory tract of dogs, PCR proved to be more sensitive than culture for detection of the bacterium, which has already been shown for infection with B. bronchiseptica, B. pertussis, and B. parapertussis in humans with respiratory tract infections (Reizenstein et al. 1993). These results also confirm findings published by Canonne and coworkers (2016) who detected B. bronchiseptica by culture in 58% of dogs with infection confirmed by quantitative PCR and cytology. In this study, they were able to show that in dogs with high and very high DNA loads of B. bronchiseptica, cytological detection of pleiomorphic cocci or coccobacilli adhering to the cilia of the epithelial cells was positive in 9/9 dogs, while culture confirmed the infection in only 6/9 cases. On the other hand, while in case of a positive culture for B. bronchiseptica infection with the organism is likely, a positive PCR result can lead to misinterpretation and unnecessary antibiotic therapy in a case of sole colonization. 

There are several limitations of the study. Only small groups of dogs with certain respiratory diseases were included, and results might have been different in a larger population of affected dogs. Secondly, the control group consisted of clinically sick dogs without signs of RD instead of completely healthy dogs. Most of these dogs were systemically ill and potentially pretreated with immunosuppressive drugs or antibiotics, which could have influenced their B. bronchiseptica status. Pretreatment with antibiotics was also not defined as an exclusion criterion in dogs with RD and therefore could have decreased the numbers of culture or PCR positive dogs. In addition, vaccination status for B. bronchiseptica was not known for all dogs, and recent vaccination with a modified live intranasal vaccine could have led to a positive test result as well. However, dogs of the control group were severely diseased animals and many of the dogs with RD had a chronic history of disease, therefore recent vaccination was less likely in most dogs included.

In conclusion, the study shows that dogs with and without RD can carry B. bronchiseptica in their upper and lower airways and a positive PCR result from both locations therefore needs to be interpreted with caution, because it could also indicate colonization without clinically significant infection. Results should always be interpreted in the light of clinical findings, endoscopy results, and airway cytology.

There were no protected, financial, professional, or other personal interests on a product, service, and/or company that might influence the contents or opinions stated in the manuscript above.

While for BALF analysis leftover material was used, pharyngeal swabs were taken with owner`s consent. The study was conducted between 2010 and 2012 before an ethical committee was founded for veterinary studies at LMU University of Munich.

The authors declare that there is no conflict of interest.

This research received no grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conceptualization: B.S.S., K.R., K.H.
Data curation: B.S.S., K.R.
Formal analysis: B.S.S., K.R.
Investigation: B.S.S., K.R., S.L., K.W., K.H.
Methodology: B.S.S., K.R., S.L., K.W.
Project administration: B.S.S., K.H.
Resources: K.H.
Supervision: B.S.S., K.H.
Visualization: B.S.S., K.R., K.H.
Writing original draft: B.S.S.
Writing and editing: B.S.S., K.R., S.L., K.W., K.H.

PD Dr. Bianca Schulz
Medizinische Kleintierklinik der LMU München
Veterinärstr. 13
80539 München
B.Schulz@medizinische-kleintierklinik.de

Anderton TL, Maskell DJ, Preston A (2004): Ciliostasis is a key early event during colonization of canine tracheal tissue by Bordetella bronchiseptica. Microbiol 150: 2843.
Batey RG, Smits AF (1976): The isolation of Bordetella bronchiseptica from an outbreak of canine pneumonia. Aust Vet J 52: 186–186.
Bemis DA, Greisen HA, Appel MJ (1977): Pathogenesis of canine bordetellosis. J Infect Dis 135: 753.
Canonne AM, Billen F, Tual C, Ramery E, Roels E, Peters I, Clercx C (2016): Quantitative PCR and Cytology of Bronchoalveolar Lavage Fluid in Dogs with Bordetella bronchiseptica Infection. J Vet Intern Med 30: 1204–1209. 
Canonne AM, Peters I, Roels E, Desquilbet L, Clercx C (2018): Detection of specific bacterial agents by quantitative PCR assays in the bronchoalveolar lavage fluid of dogs with eosinophilic bronchopneumopathy vs. dogs with chronic bronchitis and healthy dogs. Vet J 232: 52–56.
Chalker VJ, Toomey C, Opperman S, Brooks HW, Ibuoye MA, Brownlie J, Rycroft AN (2003): Respiratory disease in kennelled dogs: serological responses to Bordetella bronchiseptica lipopolysaccharide do not correlate with bacterial isolation or clinical respiratory symptoms. Clin Diagn Lab Immunol 10: 352–356.
Chan CM, Ridgway MD, Mitchell MA, Maddox CW (2013): Association between Mycoplasma-specific polymerase chain reaction assay results and oral bacterial contamination of bronchoalveolar lavage fluid samples from dogs with respiratory tract disease: 121 cases (2005–2012). J Am Vet Med Assoc 243: 1573–1579.
Decaro N, Camero M, Greco G, Zizzo N, Tinelli A, Campolo M, Pratelli A, Buonavoglia C (2004): Canine distemper and related diseases: report of a severe outbreak in a kennel. New Microbiol 27: 177–181.
Decaro N, Mari V, Larocca V, Losurdo M, Lanave G, Lucente MS, Corrente M, Catella C, Bo S, Elia G, Torre G, Grandolfo E, Martella V, Buonavoglia C (2016): Molecular surveillance of traditional and emerging pathogens associated with canine infectious respiratory disease. Vet Microbiol 192: 21–25.
Echeverri-Toro L, Arango A, Ospina S, Agudelo C (2015): [Bordetella bronchiseptica recurrent bacteraemia in a patient with bone marrow transplantation]. Biomedica 35: 302–305.
Hawkins EC, Denicola DB, Kuehn NF (1990): Bronchoalveolar lavage in the evaluation of pulmonary disease in the dog and cat. State of the art. J Vet Intern Med 4: 267–274.
Johnson LR, Queen EV, Vernau W, Sykes JE, Byrne BA (2013): Microbiologic and cytologic assessment of bronchoalveolar lavage fluid from dogs with lower respiratory tract infection: 105 cases (2001–2011). J Vet Intern Med 27: 259–267.
Keil DJ, Fenwick B (1998): Role of Bordetella bronchiseptica in infectious tracheobronchitis in dogs. J Am Vet Med Assoc 212: 200–207. 
Lavan R, Knesl O (2015): Prevalence of canine infectious respiratory pathogens in asymptomatic dogs presented at US animal shelters. J Small Anim Pract 56: 572–576.
Mitchell JA, Cardwell JM, Leach H, Walker CA, Le Poder S, Decaro N, Rusvai M, Egberink H, Rottier P, Fernandez M, Fragkiadaki E, Shields S, Brownlie J (2017): European surveillance of emerging pathogens associated with canine infectious respiratory disease. Vet Microbiol 212: 31–38.
Mochizuki M, Yachi A, Ohshima T, Ohuchi A, Ishida T (2008): Etiologic study of upper respiratory infections of household dogs. J Vet Med Sci 70: 563–569.
Radhakrishnan A, Dobratz KJ, Culp WT, King LG (2007): Community-acquired infectious pneumonia in puppies: 65 cases (1993–2002). J Am Anim Hosp Assoc 230: 1493–1497.
Rajamäki MM, Järvinen AK, Saari SA, Maisi PS (2001): Effect of repetitive bronchoalveolar lavage on cytologic findings in healthy dogs. Am J Vet Res 62: 13–16. 
Rampelotto RF, Hörner A, Hörner C, Righi R, Hörner R (2016): Pneumonia caused by Bordetella bronchiseptica in two HIV-positive patients. Sao Paulo Med J 134: 268–272.
Reizenstein E, Johansson B, Mardin L, Abens J, Möllby R, Hallander HO (1993): Diagnostic evaluation of polymerase chain reaction discriminative for Bordetella pertussis, B. parapertussis, and B. bronchiseptica. Diagn Microbiol Infect Dis 17: 185–191.
Schulz BS, Kurz S, Weber K, Balzer HJ, Hartmann K (2014): Detection of respiratory viruses and Bordetella bronchiseptica in dogs with acute respiratory tract infections. Vet J 201(3): 365–369.
Schulz B, Raufeisen K, Weber K, Laberke S, Hartmann K (2015): Comparison of the prevalence of Mycoplasma species in dogs with and without respiratory disease. Berl Munch Tierarztl Wochenschr 128: 304–309.
Taha-Abdelaziz K, Bassel LL, Harness ML, Clark ME, Register KB, Caswell JL (2016): Cilia-associated bacteria in fatal Bordetella bronchiseptica pneumonia of dogs and cats. J Vet Diagn Invest 28: 369–376. 
Thompson H, McCandlish IA, Wright NG (1976): Experimental respiratory disease in dogs due to Bordetella bronchiseptica. Res Vet Sci 20: 16–23.