Introduction

Brevundimonas vesicularis, formerly known as Pseudomonas vesicularis, is a non–spore-forming nonfermenting aerobic Gram-negative bacillus¹ that is ubiquitous in the environment and is commonly found in soil and water.² It is the main representative of Brevundimonas spp. and is considered an opportunistic pathogen with the potential to cause severe and invasive infections.³ This group of bacteria can survive in aquatic environments, including hospital water, purified water, and aircraft water, and is becoming an emerging cause of nosocomial infections in humans.⁴B vesicularis has been reported as a rare causative agent of pneumonia, tonsillitis, peritonitis, septic arthritis, bacteremia, ⁵ infective endocarditis,¹ pneumosepsis,² and meningitis⁶ in humans. In a review of 49 Brevundimonas spp. infections in humans, the majority (71%) were due to B vesicularis, and bacteremia was the most common clinical manifestation, identified in 35% of patients.⁴ Reports of Brevundimonas spp. isoloation in the veterinary literature are sparse and limited to a report of a urinary tract infection in a dog from Italy³ and fibrinous pericarditis in a cat from Japan.⁷

This report describes B vesicularis bacteremia in a dog presenting with multisystemic inflammatory processes, including urinary tract infection, polyarthritis, endocarditis, and spinal empyema. Given the concurrent Escherchia coli infection, the specific pathogenesis of B vesicularis is unclear, but it may represent an emerging pathogen in dogs.

Case Report

A 7 yr old, 31 kg, spayed female Labrador retriever presented to a primary care veterinarian for a 12 hr history of shaking, lethargy, and weakness. The dog lived on a large ranch in an agricultural area in Colorado with free access to pools and ponds of standing water and also had a travel history to Arizona. On initial exam, the dog was febrile at 104.5 °F, tachypneic (reference = 62 bpm), and weakly ambulatory with moderate lumbar spinal hyperpathia. Complete blood count and serum biochemistry profile were within normal limits apart from lymphopenia (0.45 K/μL; reference range [RR]: 0.83–4.91 K/μL), monocytopenia (0.1 K/μL; RR: 0.14–1.97 K/μL), hyperglycemia (135 mg/dL; RR: 75–125 mg/dL), and hypertriglyceridemia (171 mg/dL; RR: 30–130 mg/dL). Three-view thoracic radiographs were within normal limits. The dog was hospitalized overnight and treated with IV fluids (Lactated Ringers Solution,^a correcting for 7% dehydration), maropitant citrate^b 1 mg/kg IV q 24 hr, pantoprazole^c 1 mg/kg IV q 12 hr, lidocaine^d 30 μg/kg/min constant rate infusion [CRI] for analgesia, and hydromorphone^e 0.05 mg/kg subcutaneously q 6–8 hr as needed. The dog was persistently febrile and progressed to become nonambulatory by the next morning. The patient was subsequently transferred to a specialty hospital for further evaluation.

On presentation to the specialty hospital 24 hr after initial evaluation, the dog was quiet, responsive, and nonambulatory tetraparetic with bilateral stifle and tarsal pain and effusion. Rectal temperature was 104.1 °F. The remainder of the general physical exam was unremarkable. Initial diagnostics included left carpal, tarsal, and stifle radiographs, joint fluid cytology with culture, urinalysis with culture, and Rickettsial infectious disease testing.

Urinalysis showed a specific gravity of 1.008, mild proteinuria (30 mg/dL), hematuria (1–5 red blood cells/high-power field), pyuria (6–10 white blood cells/high-power field), 3+ mixed bacteria, and a few amorphous crystals. Urine culture was submitted, but the sample was misplaced by the laboratory and could not be performed. Left stifle, tarsal, and carpal radiographs showed degenerative joint disease in the left carpus and metacarpophalangeal joints with proliferative changes in the ventral accessory carpal bone and mild effusion of the left stifle. The radiographic changes were thought to be representative of a more diffuse polyarthorpathy. Arthrocentesis was performed on the left carpus, left stifle, and left tarsus under sedation (methadone^f 0.15 mg/kg IV, dexmedetomidine^g HCL 3 μg/kg IV). During the procedure, the patient developed an arrhythmia consistent with an accelerated idioventricular rhythm with a heart rate of 130 bpm that resolved after atipamezole^h HCL (0.03 mg/kg intramuscularly reversal. The dog was admitted to the intensive care unit, and initial treatments included IV fluids (Lactated Ringers Solution at 60 mL/kg/day), dexamethasone sodium phosphateⁱ 0.2 mg/kg IV once, methadone 0.15 mg/kg IV q 6 hr, and pantoprazole 1 mg/kg IV q 24 hr.

The dog became progressively obtunded and tetraparetic throughout the day and was transferred to the neurology service for further diagnostics and treatment. Neurologic examination revealed moderate to severe obtundation, nonambulatory severe tetraparesis with marked cervical ventroflexion, reduced conscious proprioception (paw placement) in all four limbs that was worse on the right side, and normal thoracic and pelvic limb spinal reflexes. Cranial nerve deficits included reduced menace response on the right, miosis in the left eye, and a mildly reduced gag reflex. Moderate multifocal spinal hyperpathia was noted, which was most severe in the cervical spine. The neurological exam suggested multifocal central nervous system involvement raising concern for a more diffuse systemic infectious or immune-mediated inflammatory process, and magnetic resonance imaging (MRI) was recommended.

MRI of the brain and cervical spine was obtained using a 1.5T MRI^j under general anesthesia. Mannitol^k 0.5 G/kg IV was given slowly over 20 min before anesthesia. General anesthesia was induced using diazepam^l 0.2 mg/kg IV and propofol^m titrated to effect and maintained with 1–2% isofluraneⁿ in 100% oxygen via an endotracheal tube. MRI sequences of the cervical spine included fast spin echo T2-weighted sequences in sagittal and axial planes, short tau inversion recovery sequences in sagittal and axial planes, and spin echo T1-weighted sequences taken before and after the IV injection of gadobenate dimeglumine^o 0.2 mg/kg in sagittal, axial, and dorsal planes. T2-weighted sagittal and axial and postcontrast sagittal and axial images were also obtained of the brain. There was evidence of T2/short tau inversion recovery sequence hyperintense, T1 iso/hypointense, T1 postcontrast-enhancing irregular material in the dorsal epidural space causing a loss of subarachnoid filling extending from C2–C4. The brain was structurally normal with no obvious inflammatory lesions.

Under anesthesia. the patient had a single and couplet ventricular premature complexes conducted as singlets and couplets. Hyperventilation and hyperthermia were also noted, but recovery was uneventful. Anaerobic and aerobic blood culture specimens were collected before antibiotic administration from the left dorsal pedal vein and right dorsal pedal vein 1 hr apart.

Cerebrospinal fluid was collected from the cerebellomedullary cistern, and analysis showed a total nucleated cell count of 0 cells/μL, red blood cell count of 18 cells/μL, and increased protein (53.7 mg/dL; RR 0–30 mg/dL). Cytocentrifuged preparation showed 34 nucleated cells including 20 (59%) large mononuclear cells, 10 (29%) small lymphocytes, and 4 (12%) nondegenerate neutrophils. Culture of the spinal fluid revealed no growth.

Joint fluid analysis available 24 hr after collection showed marked suppurative inflammation with no overt infectious agents. In each joint fluid sample, almost all cells were nondegenerate neutrophils with rare large mononuclear cells or small mononuclear cells. Some neutrophils contained variably sized, blue to purple, particulate matter in their cytoplasm, which was interpreted as either cellular debris or ragocytes. Joint culture revealed no growth after 5 days of incubation.

Vector borne SNAP4DX Plus^p was negative (Lyme disease, Ehrlichia spp., Anaplasma spp., heartworm). Vector-borne disease polymerase chain reaction and serology testing^q reported 14 days later were negative for Anaplsma spp., A platys, A phagoctophilum, Babesia spp., Apicomplexa, Bartonella spp., Ehrlichia spp., hemotropic Mycoplasma spp., and Rickettsia spp.). Bartonella PCR^r and culture were both negative.

Overnight treatment included IV fluid supplementation (90 mL/kg/day Lactated Ringers Solution), ampicillin/sulbactam^s (30 mg/kg IV q 8 hr), enrofloxacin^t (10 mg/kg IV q 24 hr), metronidazole^u (10 mg/kg IV q 12 hr), dexamethasone (0.1 mg/kg IV q 24 hr), fentanyl^v CRI (2 μg/kg/hr), ketamine^w CRI (2 μg/kg/min) pantoprazole (1 mg/kg IV q 12 hr) and 24 hr telemetry monitoring. The dog’s mentation, cervical ventroflexion, and ambulation improved overnight. The pyrexia resolved, but single ventricular premature complexes were still present. On cardiology consultation the following day, a grade I/VI left apical systolic murmur was auscultated. Echocardiography revealed mild thickening and prolapse of the anterior mitral valve leaflet with a fibrinous strand attached to the atrial surface (see Figure 1) and mild mitral regurgitation (MR velocity: 4.48 m/s). There was also thickening of the left coronary cusp of the aortic valve (AV) and a fibrinous mobile strand attached to the valve with trivial aortic regurgitation. There was no evidence of left ventricular outflow tract obstruction (AV velocity: 1.2 m/s). Mild pulmonary arterial hypertension was also present (tricuspid regurgitation velocity: 3.05 m/s). The echocardiogram findings were suspicious for mitral and AV endocarditis versus nonbacterial thrombotic endocarditis. Following the echocardiogram, clopidogrel^x 37.5 mg (1.04 mg/kg per os [PO] q 24 hr) was added to the treatment regimen. No specific anti-arrhythmic therapy was recommended, and the ventricular arrhythmia resolved within 24 hr of hospitalization.

View Figure

• Download as Powerpoint
• Download Figure

After the first 48 hr of hospitalization, IV analgesia was discontinued, and the dog was transitioned to gabapentin^y 9.9 mg/kg PO q 8 hr. Oral appetite stimulation was initiated with mirtazapine^z 0.95 mg/kg PO q 24 hr and then transitioned to capromorelin^aa 3 mg/kg PO q 24 hr on day 3 of hospitalization. Gastrointestinal support was provided with maropitant citrate 1 mg/kg IV q 24 hr and pantoprazole 1 mg/kg IV q 24 hr. The dog’s comfort, appetite, mentation, and ability to ambulate improved steadily throughout 5 days of hospitalization. A recheck echocardiogram on day 5 revealed mild thickening of the mitral valve with very mild prolapse of the anterior leaflet; however, the previously noted fibrinous strand attached to the mitral valve leaflet was not visualized and regurgitation quantity was reduced to trivial. Trivial aortic regurgitation remained stable without visualization of the previously noted fibrinous strand. The mild pulmonary arterial hypertension was resolved. The dog was discharged from the hospital with metronidazole^bb 15.8 mg/kg PO q 12 hr, enrofloxacin 8.5 mg/kg PO q 24 hr, amoxicillin/clavulonic acid^cc 23.8 mg/kg PO q 12 hr, gabapentin 9.5 mg/kg PO q 8 hr, clopidogrel 2.3 mg/kg PO q 24 hr, maropitant citrate 2 mg/kg PO as needed, capromorelin 30 mg/kg PO q 24 hr as needed, and prednisone^dd 0.6 mg/kg PO q 24 hr for 3 days, then 0.3 mg/kg PO q 24 hr until recheck.

Blood culture results available 6 days after discharge showed growth of B vesicularis and E coli with high antibiotic susceptibility for E coli and limited knowledge of susceptibility pattern for Brevundimonas spp (see Table 1). Because of limited information on antimicrobial sensitivity for B vesicularis, the patient was kept on broad-spectrum antibiotic therapy.

TABLE 1 Initial Blood Culture

View Fullscreen

When rechecked 16 days after discharge, the dog was afebrile, appetite and mentation were normal, and there was no evidence of paresis or ataxia. There was mild persistent joint stiffness and occasional lethargy but was otherwise behaving normally at home. Antibiotic therapy was de-escalated to amoxicillin/clavulonic acid 23.8 mg/kg PO q 12 hr alone. Gabapentin 9.5 mg/kg PO q 8 hr and clopidogrel 2.3 mg/kg PO q 24 hr were continued. Prednisone was tapered to 0.3 mg/kg PO q 48 hr for 2 wk then discontinued. A phone recheck was done 1 wk after discontinuing the prednisone, and she was doing very well at home.

During the recheck echocardiogram 35 days after discharge, the owners reported normal activity level and appetite and no syncopal events or episodes of weakness at home. No murmur or arrhythmia was auscultated on physical examination. Echocardiographic assessment revealed stable trivial MR and trivial aortic regurgitation with nonspecific, mild thickening of the mitral and AV leaflets previously described. Amoxicillin/clavulonic acid 23.8 mg/kg PO q 12 hr, gabapentin 9.5 mg/kg PO q 8–12 hr pro re nata (PRN), and clopidogrel 2.3 mg/kg q 24 hr were continued as previously prescribed.

Rechecks by the neurology service 2 and 3 mo after discharge were normal other than slightly reduced cervical range of motion without overt pain with direct palpation of the cervical spine. Owners reported suspected intermittent cervical pain for which she received Gabapentin 9.5 mg/kg PO PRN.

When rechecked by the cardiology service 3 mo after discharge, no murmur or arrhythmia was auscultated. Echocardiographic assessment revealed stable mild mitral valve thickening with very mild prolapse of the anterior leaflet and trivial MR. The AV was mildly hyperechoic with trivial aortic regurgitation. Pulmonary hypertension had resolved. Clopidogrel 1.1 mg/kg PO q 24 hr, and gabapentin 9.5 mg/kg PO q 8 hr were continued. Amoxicillin/clavulonic acid was discontinued after completing a 4 mo course.

Five months after discharge (1 mo after completing antibiotics), a recheck with the neurology service showed that she was overall doing well but had periods of stiffness when getting up. Examination revealed a stiff gait with left-sided thoracic and pelvic limb lameness and pain with palpation of the tarsi and extension of stifles bilaterally. Therapy was reinitiated with amoxicillin/clavulonic acid 23.8 mg/kg PO q 12 hr, gabapentin 9.4 mg/kg PO q 8–12 hr as needed, and a tapering dose of prednisone 0.6 mg/kg PO q 24 hr for 5 days, then 0.3 mg/kg PO q 24 hr for 5 days, then 0.3 mg/kg PO q 48 hr for five doses.

When rechecked 2 wk later, she was doing well at home aside from slipping on wet floors occasionally. On examination, she had bilateral prescapular lymphadenomegaly, a prominent right popliteal lymph node, multifocal spinal pain, and persistent lameness with multifocal joint pain. Concern for recurrence of the infection prompted urine culture, blood culture, echocardiogram, and aspiration of the prescapular lymph node. Echocardiogram revealed a stable MR with valve thickening, recurrent aortic regurgitation with new mild thickening of the left coronary cusp. Prescapular lymph node cytology was consistent with adipose tissue. Urine and blood culture revealed no growth after 5 days of incubation. Amoxicillin/clavulonic acid was continued as previously prescribed. Prednisone was increased to 0.3 mg/kg PO q 24 hr, and gabapentin 9.4 mg/kg PRN was continued for pain.

One month later (7 mo after initial diagnosis), the neurology recheck examination was normal, and the owners reported normal gait and activity. Prednisone therapy was tapered to 0.16 mg/kg PO q 24 hr, and antibiotic therapy remained the same with gabapentin as needed. Echocardiogram at that time was stable with trivial MR and trivial aortic regurgitation. Long-term amoxicillin/clavulonic acid therapy was recommended given the rapid recurrence off antibiotics with recheck neurological exams every 2–3 mo and echocardiograms every 3–6 mo to guide recommendations for the length of therapy.

Discussion

This is the first report of B vesicularis bacteremia in a dog and highlights it as a potential causative pathogen for infectious endocarditis (IE), spinal empyema, and polyarthritis. The concurrent infection with E coli makes it difficult to determine to what degree each bacteria contributed most significantly to the multifocal inflammatory processes; however, given the large growth of B vesicularis on blood culture and the rarity of B vesicularis as a reported pathogen, description of this case was deemed important.

B vesicularis has only been cultured in the urine of one dog³ and, to the best of our knowledge, has never been cultured in the blood of a canine. Unfortunately, due to a lab error, urinary tract infection with B vesicularis could not be documented. Joint and spinal fluid showed no bacterial growth; however, culture from these fluids is often low yield.^8,9 There is limited information on the disease spectrum associated with B vesicularis infection in canines and no reported data on the optimal treatment regimen. Given the rarity of human Brevundimonas spp. infections, there is little documentation of best antimicrobial therapy.⁴ In humans, Brevundimonas spp. are grouped with other opportunistic pathogens causing nosocomial infections such as Acinetobacter baumannii, Burkholderia cepacian, and Pseudomonas aeruginosa.⁴ This group of pathogens has been shown to acquire resistance to many antibiotics, including fluoroquinolones and ß-lactams.⁴ Cephalosporins, penicillins, and aminoglycosides have been used most successfully.⁴ The B vesicularis urinary tract infection reported in a single dog was susceptible to aminoglycosides, fluoroquinolones, and sulfonamides, and treatment with marbofloxacin was successful.³ Because of a lack of Clinical and Laboratory Standards Institute interpretive criteria, the reference lab didn’t provide antimicrobial susceptibility data for B vesicularis in this case, and the patient was treated with broad-spectrum antibiotic therapy initially. Continuation of amoxicillin/clavulonic acid for 3.5 mo allowed for persistent marked improvement/resolution of clinical signs and echocardiographic abnormalities. Reintroduction of amoxicillin/clavulonic acid when the patient relapsed with similar clinical signs provided adequate control again.

IE is a rare disease in dogs, with a reported incidence of <0.1% of cases examined at veterinary teaching hospitals and 1% of cases managed by cardiologists.¹⁰ Diagnosis rarely occurs before severe lifethreatening complications such as septic thromboembolism, congestive heart failure, and glomerulonephritis. Arrythmias and conduction disturbances occur as a complication of IE in 45% of cases and were noted in the case reported here.¹⁰ Thromboembolic disease occurs more commonly in dogs, with IE affecting the mitral valve,¹⁰ which precipitated the use of clopidogrel in this case. Aggressive medical treatment with long-term broad-spectrum antibiotics is needed, and prognosis is grave for dogs with aortic IE and poor to fair in dogs with mitral valve IE.¹¹ Treatment of endocarditis involves long-term broad-spectrum or blood culture and susceptibility-guided antibiotic therapy. IV antibiotic therapy is recommended for 1–2 wk with long-term oral antibiotic therapy after.¹⁸ The dog in this report showed marked improvement in the vegetative lesions and clinical cardiac abnormalities after initiation of IV antibiotic therapy. The cardiac arrhythmia and quiet murmur were likely identified early in the course of disease and may have influenced the positive prognosis in this case. Recommendations for length of oral antibiotic administration suggest therapy for 6–8 wk or longer.¹⁴ The course of therapy should be guided by response to antibiotics, serial monitoring of echocardiograms, and other parameters such as blood work, urine or blood cultures, and recurrence of previous abnormal physical exam abnormalities.¹⁴ Antibiotic therapy in this dog was discontinued when her cardiac changes were stable, physical exam abnormalities were resolved, and after she had completed a 4 mo course. However, prompt recognition of relapse in clinical signs and physical exam abnormalities prompted further diagnostics and ultimately restarting prolonged oral antibiotic therapy because of a fear of recurrence. Lifelong antibiotic therapy in the face of negative culture results may contribute to the risk of increasing bacterial resistance and thus is a decision that should be made with this in mind.

Up to 24% of dogs with bacterial endocarditis have concurrent neurological deficits,¹⁰ and inflammatory central nervous system lesions have been identified histopathologically in dogs with endocarditis due to infarction and suspected immune-mediated processes. In a retrospective study of 71 dogs diagnosed with possible or definitive infective endocarditis, only 6 dogs (8%) had necropsy confirmation of neurologic lesions.¹⁰ Lesions found on necropsy were suppurativelymphoplasmacytic meningitis (2 dogs), multifocal infarction (2 dogs), and suppurative-histiocytic meningitis (1 dog), and 1 dog had multifocal meningeal thrombosis with necrotizing vasculitis.¹⁰ Although there were no obvious changes noted on MRI of the brain in this case, the neurological examination findings of severe obtundation, menace deficit, and miosis suggest possible brain involvement, and subtle inflammatory lesions can often be missed on MRI. Given the marked improvement in this case, there was no histopathology available to look for brain lesions.

Conclusion

Spinal epidural empyema and IE are rarely diagnosed concurrently. To the authors’ knowledge, there is only one report of concurrent spinal empyema and IE in a dog, and Staphylococcus aureus bacteremia was isolated in that case¹² based on blood and epidural space culture results. Infection within the epidural space occurs most commonly because of extension from a local infectious process including discospondylitis, penetrating injury, and dermatitis/myositis/steatitis, and less commonly as a result of hematogenous bacterial spread from distant sites including pyothorax, endocarditis, prostatitis, and urinary tract infection.^12–17 Neutrophilic polyarthritis has been reported in 30% of dogs with infective endocarditis, with about 25% of those dogs having septic polyarthritis and the remainder having synovial changes consistent with immune-mediated disease.¹⁰ Bacterial growth can be difficult to elicit from joint fluid,⁹ so it is difficult to determine whether the polyarthritis in this case was due to a septic or secondary immune-mediated process. Given that the neutrophils were mature and not degenerative, it is more likely that this was a secondary immune-mediated process due to systemic inflammation; however, without joint histology, that could not be confirmed.

Although the diagnostic testing was relatively extensive for this dog, a urine culture would have provided valuable additional information and could have potentially linked urinary tract as the initial source of infection. The concurrent E coli infection could have contributed to systemic infection, and it is a commonly reported pathogen for urinary tract infection, IE, and spinal empyema.^{10,11,13,17–20} Histopathology may have provided additional information to help elucidate the degree to which B vesicularis contributed to the pathology in this patient but was not available given the dog’s resolution of signs with treatment.

The source of B vesicularis infection in this dog is unknown. Given the dog’s free access to agricultural land with various water sources, infection via contaminated water was assumed. Brevudimonas spp. have been isolated from multiple environments, including soil, deep subsea floor sediment, black sand, purified water, and tissue specimens of immunocompromised human patients.⁴Brevudimonas spp. have the ability to pass through sterilizing filters,⁴ which may allow the introduction of nosocomial infections in immunocompromised patients.

This case report introduces B vesicularis as a new source of bacteremia in dogs in the United States. Further cases are required to help elucidate the spectrum of pathology, antimicrobial sensitivity, and therapeutic response.