Article Text

Download PDFPDF

Short Communications
Successful medical treatment of Erysipelothrix rhusiopathiae-induced lumbosacral diskospondylitis in a dog
  1. L. Golini,
  2. J. P. Morgan,
  3. T. Glaus and
  4. F. Steffen
  1. Cardiology Service, Clinic for Small Animal Internal Medicine, VetSuisse Faculty, University of Zurich, Winterthurerstrasse 260, 8050 Zurich, Switzerland
  1. E-mail for correspondence lgolini{at}

Statistics from

THE lumbosacral disc is a common site for the settlement of local or systemic bacterial infections, leading to lumbosacral diskospondylitis (Gilmore 1987, Burkert and others 2005 and others 2005). More commonly isolated organisms include Staphylococcus species, Brucella canis, Streptococcus species and Escherichia coli; additional less common organisms include Pasteurella multocida, Actinomyces viscosus, Nocardia species, Mycobacterium avium, Proteus species and corynebacterium species (Betbeze and McLaughlin 2002, Burkert and others 2005 and others 2005). In human medicine, pyogenic vertebral osteomyelitis is mainly due to Streptococcus aureus, followed by E coli, coagulase-negative staphylococci and Propionibacterium acnes (Zimmerli 2010). Erysipelothrix rhusiopathiae is rarely recognised as zoonotic agent of osteomyelitis in immunodeficient patients but it is easily treatable. Whereas this organism is a very important pathogen in pigs, where acute disease is characterised by sudden death or general signs of septicaemia, the subacute form is characterised by classical diamond-skin light pink to reddish lesions or a chronic form can result with signs of local arthritis or proliferative pathological changes in the heart (Wang and others 2010). To the authors' knowledge, E rhusiopathiae has only been described in three fatal canine cases, of which two dogs had diskospondylitis as part of the disease (Houlton and Jefferies 1989, Burkert and others 2005 and others 2005, Seelig and others 2010). The present case report illustrates the clinical and radiological features; diagnostic procedure, clinical cure and long-term clinical follow-up of a dog diagnosed with E rhusiopathiae-induced lumbosacral diskospondylitis.

The dog, a 13-year-old-male labrador retriever, was presented to the Neurology Service of the Veterinary Teaching Hospital Faculty of Veterinary Medicine University of Zurich for a second opinion about a long-term problem in failure to jump and walk normally. The dog had been living in Ibiza, Spain, since it was born where it was regularly vaccinated, dewormed and during the warm season was wearing a deltamehtrin-based collar to reduce the risk of Phlebotomus bites. Since two months before presentation the dog had shown apathetic behaviour, progressive worsening of difficulties to walk normally and jump into the car, with occasional tenesmus and uncontrollable pain during petting the lumbosacral area. At clinical examination, the dog was apathetic, febrile (40.3°C) and panting, the mucous membranes were mildly reddish and the heart rate was 150 bpm. The heart auscultation was mildly arrhythmic without murmurs. The dog had no neurological deficits, however, he showed pain on palpation of the caudal abdomen and in the lumbosacral area as well as pain upon extension of the hips (lordosis test) and trans-rectal palpation of the ventral aspect of the sacrum. On rectal palpation, the prostate was moderately enlarged but not painful. Lumbosacral disease was clinically suspected and an infectious cause was presumed; degenerative, neuropathic or neoplastic disease involving the cauda equina or the bone and joints structures of this anatomical area were considered less likely.

Radiographs of thorax, caudal abdomen, lumbosacral area and hips of the dog were performed under sedation. Thorax, abdomen and hips were unremarkable and the vertebrae between C6 and L6 showed no obvious pathological changes. A heavy pattern of spondylosis deformans was present around L6-7 and L7-S1 ventrally and laterally. The endplates of the vertebrae L7-S1 were sclerotic with associated intervertebral disc-space collapse. Lytic/destructive lesions in the mid portion of the endplates were also observed (Figs 1, 2). The radiological diagnosis supported the clinical suspicion of diskospondylitis at the lumbosacral joint superimposed to chronic spondylosis deformans.


Laterolateral radiograph of the lumbosacral region of a dog during fine-needle aspiration. Laterolateral radiograph of the lumbosacral joint with a heavy pattern of spondylosis deformans at L6-7 and L7-S1. A collapsed LS disc space is evident plus endplate sclerosis at this disc space. The centrally located end plate lysis is difficult to visualise because of the laterally positioned superimposed spondylosis. The tip of the needle is within the centre of the intervertebral disc


Ventrodorsal radiograph of the lumbosacral joint. The same features seen on Fig.1 are also evident as well as the bony response of the spondylosis deformans that bridges the L7-S1 disc space laterally. The bone lysis within the endplates is clearly identified on this view

With the dog still under anaesthesia, a fluoroscopically guided fine-needle aspiration (FNA) of the L7-S1 disc was performed. A 20G 90 mm long spinal needle was aseptically inserted perpendicularly through the surgically prepared skin into the disc (Fig 1), and gentle aspiration was performed with a sterile 10 ml syringe. The aspirated material was gently deposited on a sterile transport swab, suitable for both aerobes and anaerobes (COPAN Innovation). A second sample was obtained by inserting the needle on a different path from the first one and flushing the disc with 1 ml of sterile and pyrogen-free irrigation solution (NaCl 0.9 per cent, B. Braun Melsungen AG). The re-aspirated fluid was transported to the laboratory by means of another sterile transport swab similar to previous one. Finally, ultrasound examination of the heart was performed and did not reveal any abnormalities, specifically, all cardiac valves were free of any suspicion of endocarditis.

Routine haematological examination showed severe leucocytosis (65,000 WBC/μl; reference range: 4700 to 11,300 cells/μl) with a left shift (6000/μl bands; reference range: 0 to 800 cells/μl) and monocytosis (1500/μl; reference range: 200 to 920 cells/μl). Analysis of urine obtained by cystocentesis did not reveal pyuria or bacteria, but urine was cultured nevertheless. Ultrasound evaluation of the prostate gland and of the testicles did not reveal any abnormalities other than expected with older age, and a FNA of the prostatic tissue was cytologically unremarkable. Blood cultures for aerobic and anaerobic bacteria were taken, sampling 10 ml of whole blood hourly three times and putting the blood into a commercially available transportation/culture medium (BC0102M, Oxoid SIGNAL blood culture system medium; Oxoid).

Recovery from anaesthesia was uneventful and the dog was administered first-generation intravenous cefazolin (Kefzol, TEVA Pharma AG, Basel) at 22 mg/kg/three times a day. General physical condition, gait abnormalities and pain improved significantly by the following day. After 24 hours of intravenous antibiotic treatment, the dog was discharged with long-term oral antibiotic treatment at the same dosage (Cefaseptin forte, Vétoquinol AG), to be modified based on microbiological culture and sensitivity results.

No aetiological agent growth was obtained from either urine culture or blood cultures while from the disc material, a small Gram-positive bacillus was cultured from both the FNA of the disc and the disc wash. An amplification of 1600 base pairs with conventional PCR for the 16S rDNA of the bacterium was performed, according to the protocol developed by Medlin and others 1988, and using as forward primer the sequence 5'-CAG AGT TTG ATC CTG GCT CAG-3'and as reverse primer 5'-TAC GG(CT) TAC CTT GTT ACG ACT T-3'. Afterwards, the amplified gene was sequenced and compared with an online gene bank ( The sequenced gene showed a perfect homology with the available sequences of E rhusiopathiae. The same procedure was subsequently performed on the blood sampled previously for culture, leading again to the identification of DNA of E rhusiopathiae. The antibiogram showed that it was sensitive to cefazolin and therefore this treatment was continued for three months. Regular telephone follow-up confirmed the resolution of all clinical signs at 20 days after discharge, and after six weeks, the dog was jumping and running normally. After three months, the referring veterinarian stopped the treatment and at the time of writing, nine months after diagnosis, the dog has not been showing any recurring clinical signs.

Diskospondylitis is an infection of the intervertebral disc that spreads through both endplates and proceeds slowly into the two contiguous vertebrae (Gilmore 1987). Although any intervertebral disc space may be affected, the lower cervical, the midthoracic, the thoracolumbar and the lumbosacral disks are more often involved in dogs (Kornegay and Barber 1980, Gilmore 1987, Burkert and others 2005 and others 2005, Kinzel and others 2005). Clinical signs range widely from pain alone to non-ambulatory para- or tetraparesis, either flaccid or spastic, depending upon the localisation of the inflammatory and compressive process (Kornegay and Barber 1980, Betbeze and McLaughlin 2002, Burkert and others 2005 and others 2005). Male dogs are twice as likely to be affected as females and the risk increases with age (Burkert and others 2005 and others 2005). Additional predisposing factors are previous surgery, treatment with immunosuppressive drugs, or systemic or local infections, most commonly located in the lower urinary tract (Burkert and others 2005 and others 2005). In entire male dogs, a common infectious route to affect the lumbosacral disc is through the lymphatic vessels that drain the prostate gland with subsequent arterial spread. Primary haematogenous spread is also a recognised route of infection (Gilmore 1987, Betbeze and McLaughlin 2002, Burkert and others 2005 and others 2005) Rarely, penetrating wounds, migrating foreign bodies, such as foxtail or other plant thorns, or even more seldom sterile procedures, as epidural punctures, can convey bacteria into the affected site (Corlazzoli and Pizzirani 1998, Betbeze and McLaughlin 2002, MacFarlane and Iff 2011).

For the aforementioned reasons, urine culture should be performed systematically in all dogs with lumbosacral discospondylitis even in absence of pyuria, even though urine cultures are negative in up to 40 per cent of the cases (Kornegay and Barber 1980, Gilmore 1987, Burkert and others 2005 and others 2005). The combination of urine and blood cultures yield a successful aetiological diagnosis in more than 70 per cent of the cases (Gilmore 1987, Betbeze and McLaughlin 2002, Burkert and others 2005 and others 2005, Kinzel and others 2005). Pretreatment of affected animals with antibiotics could lead to false negative culture results. Finally, as also performed in the case reported here, a direct biopsy of the inflamed tissue either by fluoroscopic guidance or CT and by direct surgical sampling in order to isolate and identify the responsible bacteria is an accurate diagnostic procedure in 76 to 100 per cent of the cases (Kornegay and Barber 1980, Gilmore 1987, Vignoli and others 2004, Kinzel and others 2005, Zimmerli 2010).

A multitude of bacteria are commonly isolated from dogs with diskospondylitis, however, E rhusiopathiae has only been described in two clinical reports with diskospondylitis in which dogs showed widespread joint inflammation, endocarditis and diskospondylitis at multiple sites (Houlton and Jefferies 1989, Seelig and others 2010). Furthermore, besides polyarthritis and aortic valve endocarditis, diskospondylitis was also found in eight of nine beagle dogs experimentally infected with E rhusiopathiae (Schütt and others 1978).

E rhusiopathiae is a facultative, non-spore forming, non-acid- fast, small, Gram-positive bacillus. Its distribution is worldwide and it can survive for long periods in the soil as well as in marine environment (Wang and others 2010). Interestingly, E rhusiopathiae is known to be a natural host of the slim of some fishes without causing any apparent diseases (Wang and others 2010). As for other cases, the route of infection in the present patient was not determined.

To the authors' knowledge, this is the first clinical report describing the successful treatment of lumbosacral diskospondylitis due to E rhusiopathiae. The initial antibiotic choice was a first-generation cephalosporin, which is suggested by several authors as drug of choice when the causing organism is unknown, and which then was proven to be effective in this case based on the sensitivity results and clinical course (Kornegay and Barber 1980, Gilmore 1987, Burkert and others 2005 and others 2005). The duration of antibiotic treatment in diskospondylitis does not have a fixed rule and has ranged from two to 130 weeks in various reports (Gilmore 1987, Betbeze and McLaughlin 2002, Burkert and others 2005 and others 2005). It is mainly guided by radiological evidence of bone healing or, if repeated radiographs can not be taken, it should last at least two weeks beyond the cessation of clinical signs (Gilmore 1987). In this case, in the absence of control radiographs, the authors advised a 12-week long treatment regimen.

In conclusion, although rarely encountered in clinical practice, E rhusiopathiae may cause diskospondylitis in dogs. The prognosis seems to be favourable in the absence of mulitfocal infection such as polyarthritis and endocarditis.

Correction notice

This article has been corrected since it was published Online First. An error was spotted in the title, it is ‘diskospondylitis’ and not ‘discospondylitis’.


View Abstract


  • Provenance not commissioned; externally peer reviewed

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.