• Open Access

Corynebacterium ulcerans Bronchopneumonia in a Dog


Corresponding author: Jane E. Sykes, BVSc(Hons) PhD DACVIM, Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California, 2108 Tupper Hall, Davis, CA 95616; e-mail: jesykes@ucdavis.edu.


polymerase chain reaction


Veterinary Medical Teaching Hospital

An 11-year-old male neutered Labrador Retriever mixed-breed dog weighing 38 kg was referred to the William R. Pritchard VMTH at the University of California, Davis, with a 10-month history of coughing and decreased exercise tolerance. Thoracic radiographs taken by the referring veterinarian 2 months after the onset of illness showed a mild bronchointerstitial pattern and consolidation of the right middle lung lobe. The dog was treated with enrofloxacin, 3.6 mg/kg PO every 12 hours, for 14 days with minimal clinical improvement. Thoracic radiographs were repeated and showed resolution of the lobar consolidation. A CBC and serum biochemistry panel disclosed mild monocytosis (2,028 cells/μL; reference range, 150–1,350 cells/μL). Reported findings on bronchoscopic evaluation performed at the referring veterinarian's clinic consisted of bronchiectasis in the left caudal lung lobe and airway collapse. Cytology of a bronchoalveolar lavage specimen indicated mild-to-moderate mixed mononuclear and lymphocytic inflammation with 10% eosinophils. Aerobic and anaerobic bacterial, mycoplasma, and fungal cultures of the lavage fluid were negative. Doxycycline 5.3 mg/kg PO every 12 hours was prescribed, with no response. Six months after the onset of coughing, the dog was treated with prednisone, 0.5 mg/kg PO every 12 hours for 5 days tapered to 0.5 mg/kg PO every 24 hours. Inhaled fluticasone, 440 μg every 12 hours, and inhaled albuterol, 180 μg every 12 hours, were also prescribed, but were administered inconsistently. Partial clinical improvement occurred initially, but the cough returned 1 month later. Thoracic radiographs showed alveolar infiltrates within the right middle lung lobe. The prednisone was tapered and discontinued, and treatment with enrofloxacin 4.5 mg/kg PO every 12 hours initiated. The cough did not improve, and the owner discontinued all medications.

Six weeks later, at the time of referral to the VMTH, the owner described the cough as occasionally productive, and occurring up to 15 times a day. The owner also reported that over the preceding months there had been a change in the dog's bark, which was noticeably weaker. Other medications included tramadol (1.3 mg/kg PO q12h) for osteoarthritis, and monthly milbemycin oximea and fipronil.b The dog had been vaccinated annually for distemper, adenovirus, parvovirus, parainfluenza virus, and Borrelia burgdorferi and every 3 years for rabies virus. Six years before presentation, the dog had developed transient respiratory distress and subcutaneous emphysema after anesthesia and treatment of a puncture wound in the inguinal area, presumably as a result of a tracheal laceration. The owner operated a dog walking business in the San Francisco area, and the dog was in close contact with >70 dogs on a daily basis. The dog's travel history was confined to northern California. The owner was reportedly immunocompetent, although the owner's mother, who spent time in the household, had recently died from mesothelioma, which was associated with numerous hospital visits, and multiple infectious complications.

On physical examination at the VMTH, the dog was bright, responsive, and well hydrated, with a body condition score of 6/9. Rectal temperature was 101.4°F. The dog was panting, had pink mucous membranes and a capillary refill time of 1 second, and a moderate amount of dental calculus was noted on oral examination. Heart rate was 80 beats/min, with strong, synchronous pulses, and harsh lung sounds were heard on auscultation of the thorax. Three coughing episodes occurred during the examination, which were paroxysmal and nonproductive. No other abnormalities were detected.

Results of a CBC and serum biochemistry panel were unremarkable. On urinalysis, specific gravity was 1.042, pH was 8.5, and trace proteinuria was present. Serology for Coccidioides spp. antibodies was negative. Right and left lateral and dorsoventral thoracic radiographs showed a mild increase in interstitial density throughout the lungs consistent with the age of the dog. Slight dynamic change in mainstem bronchi size was noted when comparing the left and right lateral projections.

A laryngeal examination under light anesthesia with thiopental and after administration of doxapram (0.9 mg/kg IV) indicated left laryngeal paresis and right laryngeal paralysis. The dog was intubated and anesthesia maintained with isoflurane. Thoracic computed tomography showed only mild thickening of the left cranial bronchus. A bronchoscopic examination showed widespread mucosal erythema, edema, and irregularity with copious quantities of thick yellow-green mucus throughout the airways, particularly the trachea and mainstem bronchi. Severe collapse of the left mainstem bronchus was noted, and a mucous plug was present at the opening of the left caudal bronchus. Samples were obtained from the right middle and left cranial lung lobes with bronchoalveolar lavage. Cytology of the lavage fluid indicated marked neutrophilic inflammation with intracellular bacteria in both washes. The left cranial bronchoalveolar lavage sample contained 2,140 total nucleated cells/μL, with 82% neutrophils, 7% lymphocytes, and 11% macrophages. A few neutrophils contained moderate numbers of mixed bacteria including cocci, coccobacilli and rods, and few mild-to-moderately vacuolated macrophages. The right middle bronchoalveolar lavage sample contained 6,580 total nucleated cells/μL, with 98% neutrophils and 1% each of lymphocytes and macrophages. Both degenerate and nondegenerate neutrophils and occasional intracellular cocci, coccobacilli, and rods were observed. Culture for aerobic bacteria yielded Pseudomonas aeruginosa that was susceptible only to amikacin, gentamicin, and imipenem. Although the isolate was resistant to enrofloxacin and orbifloxacin with minimum inhibitory concentrations of 4 and >4 μg/mL, respectively, an expanded panel showed susceptibility to ciprofloxacin (<0.50 μg/mL) and levofloxacin (2 μg/mL). The isolate was resistant to tetracyclines, all cephalosporins, ticarcillin-clavulanate, trimethoprim-sulphamethoxazole, chloramphenicol, and linezolid. Cultures for anaerobic bacteria and mycoplasmas yielded no growth.

The dog initially was treated with enrofloxacin (7.2 mg/kg PO q24h), amoxicillin (21 mg/kg PO q12h), and coupage. After results of the culture and susceptibility, the amoxicillin was discontinued and replaced with nebulizedc gentamicin (160 mg q24h for 10 minutes), together with saline nebulization for 4 weeks. This was associated with resolution of the cough. However, beginning 2 weeks after initiation of this therapy, the cough reappeared and became progressively worse over the next 2 weeks. Physical examination at that time was unchanged from the previous examination, except that the dog had lost 3.4 kg. Thoracic radiographs showed moderate diffuse bronchointerstitial infiltrates, which were slightly increased from the previous examination. A transtracheal wash disclosed mildly to moderately increased numbers of inflammatory cells, composed primarily of mildly to moderate degenerate neutrophils, with fewer foamy activated macrophages, scattered eosinophils and small lymphocytes. Two large septate branching fungal hyphae with adherent neutrophils were observed. Cultures for aerobic and anaerobic bacteria and mycoplasmas yielded no growth. Fungal culture yielded 3 colonies of Aspergillus fumigatus. The dog was treated with itraconazole (5.8 mg/kg PO q24h) for 2 months, which resulted in resolution of the cough. Two weeks after initiating treatment, the dog became lame on the left pelvic limb, and a comminuted fracture of the 1st phalynx of the left 5th digit was diagnosed, presumably as a result of trauma. This was treated with exercise restriction and firocoxibd (5 mg/kg PO q24h for 2 weeks) and healed uneventfully.

One month after discontinuing the antifungal treatment, the cough returned. Thoracic radiographs showed mild, diffuse interstitial infiltrates. A transtracheal wash indicated suppurative septic inflammation with variably degenerate neutrophils, many of which contained slender long rods and cocci, although the overall cellularity was low. Aerobic bacterial culture yielded small numbers of Corynebacterium spp., which were identified as Corynebacterium ulcerans based on colony morphology, partial sequence analysis of the 16S rRNA gene,1,e and a bacteriological identification kit.f Partial sequence analysis of the 16S rRNA gene showed 99% homology to C. ulcerans and C. pseudotuberculosis. Further differentiation between these 2 species was made on the basis of positive glycogen fermentation.2

The isolate's identity to C. ulcerans was confirmed by PCR amplification and sequence analysis of the rpoB gene, as described previously,3,4 which revealed 98% sequence homology to the rpoB gene of C. ulcerans (GenBank accession number AY492271). In addition to the C. ulcerans, small numbers of Streptococcus viridans group and Streptococcus canis biotype 2 were isolated. Anaerobic, mycoplasma, and fungal cultures were negative. The C. ulcerans isolate was susceptible to all antibiotics tested with the exception of penicillin, and it had intermediate susceptibility to clindamycin. The dog was treated with doxycycline, 5.8 mg/kg PO every 12 hours for 10 weeks, which was associated with complete resolution of the cough. However, a dry cough returned after discontinuation of the antimicrobials. Thoracic radiographs showed focal bronchopneumonia of the right cranial lung lobe, and diffuse moderate bronchointerstitial infiltrates. A transtracheal wash again was compatible with suppurative inflammation and intracellular, rod-shaped bacteria were present. Aerobic bacterial culture, 16S rRNA and rpoB gene PCR and sequencing again indicated the presence of C. ulcerans. Small numbers of Serratia marcescens were isolated, with anaerobic, mycoplasma, and fungal cultures being negative. The C. ulcerans isolate was susceptible to all drugs tested except penicillin, and the S. marcescens was resistant only to cefazolin, cephalothin, and tetracycline. Treatment with enrofloxacin, 10 mg/kg PO every 24 hours, nebulized gentamicin (160 mg q24h for 10 minutes), saline nebulization, and coupage was initiated for 6 weeks, with resolution of the cough. One week after discontinuing the medication, a milder, paroxysmal cough returned. A transtracheal wash indicated moderate inflammation consisting of nondegenerate to mildly degenerate neutrophils, rare macrophages, and rare intracellular, large diploid bacteria. Aerobic culture and sensitivity yielded small numbers of Staphylococcus pseudintermedius, resistant only to chloramphenicol, clindamycin, and erythromycin, with intermediate susceptibility to enrofloxacin and tetracycline. Treatment with clavulanic acid-amoxicillin, 22 mg/kg PO every 12 hours for 3 weeks, was initiated, which resulted in complete resolution of the cough, and 16 months after initial presentation, the dog remains alive. The owner remained well throughout the course of her dog's illness.

In addition to rpoB gene PCR amplification, extracted DNA from each of the 2 canine C. ulcerans isolates was subjected to 2 conventional PCR assays for amplification of the A and B subunits of the diphtheria toxin (tox) gene, as described previously.5 The DNA also was subjected to a real-time PCR assay for the gene, as described in Schuhegger et al,6 with minor modifications to optimize primer annealing temperatures. The assay used the modified oligonucleotides Cdiphth-444f (5′-TATCAAAAGGTTCGGTGATGGTG-3′), Cdiphth-596r (5′-CCACGTTTTCCACGGGTTT-3′), and the probe UPL#111 (FAM-5′-CTCAGCCT-3′)g. Each reaction contained 20 × primer and probes with a final concentration of 400 nM for each primer and 80 nM for the probe, and a commercially available PCR mastermixh in a final volume of 12 μL. PCR was performed using 1 μL of extracted DNA. Samples were amplified in 384-well plates in an automated fluorometer using the manufacturers conditions.i The assay was validated using 10-fold dilutions of genomic DNA testing positive for the C. ulcerans diphtheria toxin gene. Dilutions were analyzed in triplicate and a standard curve plotted. The slope of the curve was used to calculate amplification efficiencies, using the formula E= 101/−s− 1. The sensitivity of the assay was determined in triplicate using serial log (0–7) dilutions of a plasmidj containing a 152 base pair amplicon of target DNA.

Positive controls included DNA from a toxigenic C. ulcerans isolate containing a missing nucleotide in the subunit B section, resulting in a frameshift mutation within the gene; DNA from a toxigenic C. ulcerans isolate lacking the frameshift mutation; and, DNA from a toxigenic C. diphtheriae isolate. The 1st positive control has produced false-negative results in some PCR assays previously.7 A negative control consisting of ultrapure water, which also was subject to the extraction process, was included.

The real-time PCR assay detected between 2 and 10 copies of diphtheria toxin gene DNA, with an efficiency of 96%. All positive controls were tox-positive using both conventional and real-time PCR assays. All negative controls were negative. Both C. ulcerans isolates from the dog tested negative for both subunits of the diphtheria toxin gene.

C. ulcerans is a nonmotile, non-spore–bearing, Gram-positive bacterium, which may carry a bacteriophage encoding the diphtheria toxin. Diphtheria toxin-producing C. ulcerans is an emerging zoonotic bacterial infection, classically associated with ingestion of unpasteurized milk from dairy cattle with C. ulcerans mastitis, or an association with farm animals.7,8 The frequency and severity of disease associated with the organism appear to be increasing worldwide.8,9 Women may be predisposed, and infections have frequently occurred in the elderly or otherwise immunocompromised.8,10 In humans, the organism can produce a disease that is indistinguishable from classic diphtheria, which is caused by Corynebacterium diphtheriae. Diphtheria may be manifested as pseudomembranous pharyngitis, which may be associated with respiratory obstruction, or cutaneous ulceration (Veld sore). Pseudomembrane formation within the upper respiratory tract may be associated with respiratory obstruction. Subsequent systemic toxin absorption may be associated with development of myocardial arrhythmias, or cranial and peripheral neuropathies.11 Erythromycin is recommended as the 1st line of treatment, although erythromycin and clindamycin-resistant strains have been reported.7,11

Zoonotic transmission of toxigenic C. ulcerans has been suggested in several recent reports. Recently, a pig was shown to share a toxigenic isolate with a human who had contact with pigs.12 Investigation of a fatal case of diphtheria in an elderly woman resulting from toxigenic C. ulcerans infection in the United Kingdom was followed by isolation of toxigenic C. ulcerans with an identical ribotype in 2 farm dogs and a cat that were in direct or indirect contact with the woman.10 One of the dogs developed a unilateral serosanguinous nasal discharge and ulceration of the mucocutaneous junction of the same nostril, and the animals were treated with spiramycin and metronidazole to clear the infection. Since 2005, there have been reports of toxigenic C. ulcerans carriage in 7 cats with bilateral nasal discharge from the United Kingdom,9 and suspected transmission of toxigenic C. ulcerans between asymptomatic dogs and their owners in France.8,13,14 In 1 report, C. ulcerans was isolated from the tonsils, lip, and nose of a dog with chronic ulceration of the lower lip, sneezing and nasal discharge, and again was shown to have an identical ribotype to that isolated from the owner, who had been hospitalized with diphtheria.14 A toxigenic C. ulcerans also was isolated from an apparently healthy dog in Japan.15 Pneumonia caused by C. ulcerans has not been reported previously in dogs.

Human-to-human transmission of C. ulcerans infection has not been proven.8 Despite this, after documentation of human infection, throat swabs should be collected from close contacts, followed by prophylactic antimicrobial therapy, as for C. diphtheriae.8,16 At the time of isolation of C. ulcerans and before toxin gene PCR, it was recommended that the owner make an appointment with her physician, and a diphtheria toxin booster immunization was administered. Hand washing, attention to general hygiene, and, given the owner's occupation, minimizing contact with other dogs was recommended throughout the year after presentation, following isolation of the multiple resistant P. aeruginosa.

Not all C. ulcerans isolates that harbor a toxin gene produce diphtheria toxin, and it has been recommended that all isolates possessing the gene be tested for toxin gene expression using the Elek test, an immunoprecipitation test for diphtheria toxin, or cytotoxicity assays.17 Nevertheless, some strains testing negative using the Elek test have been associated with pseudomembranous pharyngitis, and it has been suggested that either the bacteria itself may contribute to pseudomembrane formation or there may be a mixture of toxin-expressing and non-toxin–expressing strains within a lesion.8 Because the toxin gene PCR assay was negative for the isolates reported here, further testing for toxin expression was not performed.

Recurrent bacterial bronchopneumonia and Aspergillus spp. infection in this dog suggested compromise of the immune system or impaired airway clearance mechanisms. Despite unremarkable thoracic radiographs and minimal changes on computed tomography, severe airway inflammation and mucus accumulation were visible during bronchoscopy at the time of infection with P. aeruginosa. It is possible that the history of indirect contact with the human hospital environment through a household family member contributed to the unusual bacterial organisms isolated from the dog in this report.

Detection of toxigenic C. ulcerans infections in companion and food-producing animals, as well as prevention of associated human illness, is dependent on a heightened awareness of this emerging zoonosis among veterinary clinicians. The use of this novel real-time PCR-based assay for diphtheria toxin gene detection should facilitate rapid screening for and early identification of tox-positive isolates. Subsequently, Elek testing or cytotoxicity assays should be performed on tox-positive isolates in order to document diphtheria toxin production.


aInterceptor, Novartis Animal Health Inc, Greensboro, NC

bFrontline, Merial Ltd, Duluth, GA

cJorgensen Laboratories, Loveland, CO

dPrevicox, Merial Ltd

eDavis sequencing, Davis, CA

fAPI 20 Coryne, BioMerieux, Durham, NC

gRoche Applied Science, Indianapolis, IN

hTaqMan Universal PCR Mastermix, Applied Biosystems, Carlsbad, CA

iABI PRISM 7900 HTA FAST, Applied Biosystems

jBlueScript II KS, Eurofins/MWG Biotech Inc, High Point, NC


We thank Dr Tejpratap Tiwari and Ms Pam Cassiday from the Pertussis and Diphtheria Laboratory, Centers for Disease Control and Prevention, for providing the DNA extracts from toxin-producing Corynebacterium isolates for use as control samples. We thank Dr Sandra Hazanow for her care of the patient in this report.