Introduction

A 10-year-old 21.5-kg spayed female Staffordshire Terrier was evaluated because of pain and unwillingness to ambulate. The owners reported that they had tripped over the dog several days earlier and the dog had seemed to be in pain ever since. The dog had a history of multiple mass excisions (mast cell tumors, melanocytomas, and several benign masses) and was routinely rechecked by a board-certified veterinary oncologist. On presentation through the emergency service the day after being tripped over, the dog had signs of pain during palpation of the lumbosacral region and was reactive to tail-jacking. Survey thoracolumbar spinal radiographs were considered unremarkable by a board-certified veterinary radiologist. Results of point-of-care testing (BUN, creatinine, total protein, and glucose concentrations and alkaline phosphatase and alanine aminotransferase activities) were within reference intervals (RIs). The dog was discharged with instructions for rest, tramadol (5.8 mg/kg, PO, q 8 to 12 h), and carprofen (4.6 mg/kg, PO, q 24 h).

• Download Figure
• Download figure as PowerPoint slide

View Full Size

• Download Figure
• Download figure as PowerPoint slide

• Download Figure
• Download figure as PowerPoint slide

The owners reported no improvement when the dog was examined by the neurology service 5 days later. A CBC and serum biochemical panel at that time revealed leukocytosis (21,300 WBCs/µL; RI, 4,000 to 15,500 WBCs/µL) characterized by mature neutrophilia (19,596 neutrophils/µL; RI, 3,000 to 12,500 neutrophils/µL) and high alkaline phosphatase (154 U/L; RI, 5 to 150 U/L) and creatine kinase (427 U/L; RI, 5 to 380 U/L) activities. Analysis of a urine sample obtained via cystocentesis showed a specific gravity of 1.048, rare (0 to 5) WBCs per hpf, and no bacteria. Bacterial culture of the urine sample did not yield any growth.

Assessment

Diagnostic Plan

The diagnostic plan included 3-view survey thoracic radiography to screen for evidence of pulmonary metastasis and MRI for evaluation of the spinal cord and vertebrae for evidence of a mass lesion or disk protrusion. Analysis of a CSF sample would be performed, if deemed appropriate, to evaluate cell count and morphology as well as for bacterial culture and susceptibility testing.

Diagnostic Test Findings

Three-view survey thoracic radiographs were evaluated by a board-certified veterinary radiologist; the lungs showed an unstructured interstitial pulmonary pattern, but findings were otherwise normal.

Magnetic resonance imaging was performed with a 1.5T unit (Siemens Medical Solutions). Multiplanar T2-weighted, STIR, and pre- and postcontrast (gadodiamide) T1-weighted images were acquired in the transverse, sagittal, and dorsal planes, including fat saturation, of the lumbar portion of the vertebral column. Material was seen in the dorsal aspect of the spinal cord extending from L5 to the sacrum and causing ventral and leftward deviation of the caudal spinal cord and cauda equina, with a mass-like lesion surrounding the spinal cord at L5 (Figure 1). The mass was ill-defined, had an undulating margin, surrounded the spinal cord, was mildly hyperintense on T2-weighted and STIR images and hypointense on T1-weighted images, and was contrast enhancing with loss of the normal spinal cord architecture in the region. Probable extradural compression of the cauda equina due to intervertebral disk protrusion at L7-S1 with lateralized spondylosis deformans as well as multiple noncompressive annulus fibrosus extrusions and desiccated intervertebral disks were also present. Also noted was hyperintensity on T2-weighted and STIR images within the paravertebral and iliac musculature that was also contrast enhancing, suggestive of spinal epidural empyema.

Transverse MRI image at the level of L5 (A) and dorsal reconstruction of the lumbar spinal column (B), both obtained after contrast administration, of a dog evaluated because of acute pain and unwillingness to ambulate. There is material causing ventral and leftward deviation of the caudal spinal cord and cauda equina in the dorsal aspect of the spinal cord, with an ill-defined mass circumferentially around the spinal cord causing loss of normal spinal cord architecture.

Citation: Journal of the American Veterinary Medical Association 260, 6; 10.2460/javma.20.10.0580

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Transverse MRI image at the level of L5 (A) and dorsal reconstruction of the lumbar spinal column (B), both obtained after contrast administration, of a dog evaluated because of acute pain and unwillingness to ambulate. There is material causing ventral and leftward deviation of the caudal spinal cord and cauda equina in the dorsal aspect of the spinal cord, with an ill-defined mass circumferentially around the spinal cord causing loss of normal spinal cord architecture.

Citation: Journal of the American Veterinary Medical Association 260, 6; 10.2460/javma.20.10.0580

• Download Figure
• Download figure as PowerPoint slide

Transverse MRI image at the level of L5 (A) and dorsal reconstruction of the lumbar spinal column (B), both obtained after contrast administration, of a dog evaluated because of acute pain and unwillingness to ambulate. There is material causing ventral and leftward deviation of the caudal spinal cord and cauda equina in the dorsal aspect of the spinal cord, with an ill-defined mass circumferentially around the spinal cord causing loss of normal spinal cord architecture.

Citation: Journal of the American Veterinary Medical Association 260, 6; 10.2460/javma.20.10.0580

• Download Figure
• Download figure as PowerPoint slide

Treatment

Given the MRI findings, primary differential diagnoses for the mass lesion consisted of neoplasia (eg, sarcoma or lymphoma), empyema, and neoplasia with secondary empyema. A standard right hemilaminectomy was performed at the level of L5-L6. Inspissated fat and liquid compressing the spinal cord ventrally and laterally were removed, and a durotomy was performed. Samples of the inspissated fat and dura were submitted for cytologic and histologic examination and for bacterial culture and susceptibility testing. Results of cytologic examination at the time of surgery were consistent with a sarcoma with septic inflammation; however, the possibility of a primary abscess with secondary reactive fibroplasia or other reactive tissue could not be excluded. The final histologic diagnosis for the dura was mild, diffuse, suppurative meningitis; the final histologic diagnosis for the epidural fat was pyogranulomatous to mixed steatitis and fasciitis. No neoplastic cells were seen on histologic examination, and results of special staining for fungal and mycobacterial elements were negative. Bacterial culture was positive for Staphylococcus pseudintermedius that was resistant only to penicillin G.

Treatment with amoxicillin–clavulanic acid (23.2 mg/kg, PO, q 12 h), gabapentin (13.9 mg/kg, PO, q 8 h), carprofen (2.3 mg/kg, PO, q 12 h), and tramadol (4.7 mg/kg, PO, q 8 h, as needed) was initiated. Recheck MRI was performed 47 days after surgery, and findings were consistent with resolution of the dorsal spinal epidural empyema in the caudal lumbar region and a reduction in the severity of the caudal lumbar meningitis. However, lumbosacral diskospondylitis with lumbosacral instability or subluxation that had not been seen previously was now evident. Radiography of the lumbosacral region was performed, and the interpretation by a different board-certified radiologist was progressive worsening of lumbosacral diskospondylitis, compared with initial radiographs. Results of a neurologic examination performed prior to the repeated MRI and lumbosacral radiography were normal.

Comments

Bacterial meningomyelitis and spinal epidural empyema are uncommon in dogs. Meningomyelitis, defined as inflammation of the spinal cord and surrounding meninges, can be of infectious or noninfectious etiology.^1,2 In a 2008 review³ of 28 dogs with meningomyelitis, bacterial (Streptococcus spp) meningomyelitis localized to L4-Cd5 was diagnosed in 1 dog by means of bacterial culture of a CSF sample. More common is the diagnosis of spinal epidural empyema, which is defined as bacterial infection within the extradural space.^4,5 Historically, this condition has been viewed as requiring surgical treatment; however, more recent reports^5–8 describe success with medical management. The decision to proceed with medical management versus surgery may depend on the location of infection, extent of disease on diagnostic imaging, clinical extent of neurologic signs and pain, and financial constraints of the owners. In the case described in the present report, concerns about a potential neoplastic process, owner desire for a definitive diagnosis, and severity of patient pain were primary factors driving the decision to proceed with surgical intervention.

Previously reported routes of infection in dogs with spinal epidural empyema are hematogenous spread, extension from an adjacent site, penetrating trauma, and migration of foreign material.^5,7–9 Pathogens reported as causative agents include Bacteroides spp, Staphylococcus spp, Streptococcus canis, Escherichia coli, Pasteurella spp, Actinomyces spp, Fusobacterium spp, Klebsiella spp, and Salmonella spp.^9–14 In the present report, the underlying cause of meningomyelitis and spinal epidural empyema was not determined, but the authors speculate that the origin was either hematogenous spread or extension from an adjacent site. The patient was not a hunting or active outdoors dog and had not sustained trauma or recent regional medical interventions, making penetrating trauma or migrating foreign material less likely. In addition, with the abrupt onset of clinical signs, a more-distant penetrating trauma that might have been forgotten by the owners seemed less likely.

A diagnosis of bacterial meningomyelitis or spinal epidural empyema can be made, or at least suspected, on the basis of results of a CBC, serum biochemical panel, analysis and bacterial culture of a urine sample, survey spinal radiography, analysis and bacterial culture of a CSF sample, and MRI.^3–17 Abnormalities in affected patients may include leukocytosis characterized by mature neutrophilia on a CBC, neutrophilic pleocytosis on CSF analysis, diskospondylitis on survey spinal radiographs, and positive bacterial culture results. Serum biochemical abnormalities may include high alkaline phosphatase and alanine transaminase activities, hyperglycemia, abnormal electrolyte concentrations, and high creatine kinase activity.¹⁴ Various PCR assays may aid in the diagnosis of an infectious etiology (eg, fungal, protozoal, rickettsial, or viral).^13,17 Reported MRI findings include compression of the spinal cord and narrowing of the vertebral canal, spinal cord abnormalities such as hyperintense gray matter on T2-weighted images, epidural lesions appearing as hyperintense or mixed-signal masses compatible with areas of inflammation on T2-weighted images, and diskospondylitis.^13,16 In the dog described in the present report, initial diagnostic testing revealed leukocytosis characterized by mature neutrophilia and high alkaline phosphatase and creatine kinase activities. Urinalysis revealed rare WBCs and no bacteria, with negative bacterial culture results. The origin of the empyema was considered to be diskospondylitis on the basis of survey radiographic findings and MRI abnormalities.

Clinical signs in patients with bacterial meningomyelitis and spinal epidural empyema are dependent on location and the severity of disease. Patients may have nonspecific signs such as lethargy, anorexia, vomiting, respiratory tract signs, pain or paraspinal hyperesthesia, and neurologic dysfunction related to the location of the lesion.^3,13,14 Some dogs may be febrile, but this finding is not consistent.^3,13,14 Visible external infection may also be evident.¹⁴ In the dog described in this report, the only clinical signs were pain and unwillingness to ambulate, and the patient did not have a rectal temperature > 39.2 °C.

In the present report, neurolocalization based on neurologic examination findings was the lumbosacral region owing to the location of hyperesthesia and lack of tail movement and sensation. Magnetic resonance imaging identified a lesion at L5-L6, but the neurolocalization was still accurate because the sacral spinal cord segments are located at the level of L5 in average-sized dogs. These nerves then travel to the lumbosacral region as nerve roots in the cauda equina. Although MRI was able to delineate the region of abnormality, a definitive diagnosis could not be reached without surgery and diagnostic testing of specimens obtained intraoperatively. Presumably, concerns about possible neoplasia resulting from intraoperative cytologic examination of specimens stemmed from severe inflammation and disruption of normal cell morphology. Despite the lack of evidence of neoplasia on histologic examination and the positive bacterial culture result, the owners were cautioned at the time of discharge that neoplasia with secondary abscess formation or empyema remained possible. Because the MRI findings had normalized 47 days after surgery, the chances of a neoplastic component appeared minimal. The radiographic abnormalities 47 days after surgery were not completely unexpected, in that radiographic resolution of diskospondylitis is often delayed, compared with the clinical response to antimicrobial treatment.

This case highlighted the importance of a complete diagnostic work-up and reevaluation when owners’ goals are definitive diagnosis and appropriate treatment. Although surgery was elected in this case, many dogs with spinal epidural empyema have been successfully managed medically, and medical management may be successful in similar cases when owners are unwilling or unable to proceed with surgery.