Suspected Paraspinal Abscess and Spinal Epidural Empyema in a Dog

Case Report

A 4-month-old, 35-kg, intact female mastiff presented for evaluation of pelvic-limb ataxia that quickly progressed to paraparesis. The owner reported that the dog had been very listless, had developed a swelling on her back, and seemed to experience pain in the hind end. The dog had been completely normal until 2 days prior to presentation and had spent the previous 24 hours at the referral veterinary hospital where she was treated with intravenous (IV) fluids and enrofloxacin^a (150 mg IV twice daily) for the back swelling and pyrexia (temperature, 104.5°F). Past history included a bite wound that occurred via a dog fight 3 weeks prior to presentation, identified by the owner near the swelling that was seen on presentation.

Upon presentation, the dog was lethargic and nonambulatory, with a firm swelling that extended from approximately the tenth thoracic (T₁₀) to the sixth lumbar (L₆) vertebra. A bite wound was not identified at this time. A neurological examination demonstrated conscious proprioceptive deficits in the pelvic limbs, paraparesis with preservation of superficial and deep pain sensation, and hyperesthesia on digital palpation of the thirteenth thoracic (T₁₃) and first lumbar (L₁) spinal column section. All spinal reflexes were exaggerated in the pelvic limbs. No other physical or neurological abnormalities were detected. Neuroanatomical localization was assessed as a third thoracic (T₃) to third lumbar (L₃) myelopathy. The differential diagnoses included: meningitis secondary to extension of a local bacterial disease process, spinal epidural empyema, trauma (e.g., spinal fracture/luxation), congenital abnormalities such as hydromyelia and syringomyelia, arachnoid cysts, and neoplasia.

Diagnostic evaluation included a complete blood count (CBC), serum biochemical profile, and urinalysis. Laboratory findings included a neutrophilia (14.11 × 10³ cells/μL; reference range, 2.06 to 10.60 × 10³ cells/μL), a monocytosis (1.49 × 10³ cells/μL; reference range, 0 to 0.84 × 10³ cells/μL), and a mild normocytic, normochromic anemia (hematocrit, 33.7%; reference range, 36% to 60%). Serum biochemical profile abnormalities included an elevated alkaline phosphatase (166 IU; reference range, 5 to 131 IU) and hyperphosphatemia (6.7 mg/dL; reference range, 2.5 to 6.0 mg/dL). Both of these abnormalities were attributed to the skeletal immaturity of the dog. Urinalysis was normal. Survey radiographs of all cervical, thoracic, and lumbar vertebrae were unremarkable.

After premedication with oxymorphone^b (0.05 mg/kg body weight, intramuscularly [IM]) and atropine sulfate^c (0.02 mg/kg body weight, IM), the dog was induced with diazepam^d (0.5 mg/kg body weight, IV) and thiopental^e (7 mg/kg body weight, IV), intubated, and maintained on isoflurane^f (1.25%) in oxygen at a flow rate of 1 L per minute through a Bain circuit. Lactated Ringer’s solution was administered perioperatively at 22 mL/kg body weight, IV for 3 hours. A broad-spectrum antibiotic, cefazolin^g (22 mg/kg body weight, IV), was also administered perioperatively. Myelography performed via a lumbar puncture demonstrated a thinning of the dye column at the body of the L₁, consistent with an extradural compressive lesion both laterally and dorsally [Figures 1, 2]. A computed tomography (CT) scan of the affected vertebrae demonstrated a mass effect with a hypodense area within the vertebral canal. Cerebrospinal fluid (CSF) analysis, obtained via a lumbar puncture, was consistent with a mixed pleocytosis having an increased cellularity and protein content consisting of 183 nucleated cells/μL (reference range, 0 to 4 cells/μL); the predominant cells were neutrophils comprising 52% of the total number of cells. The remainder of the cells were mononuclear cells. The protein level was 270 mg/dL (reference range, <35 mg/dL). No organisms or atypical cells were identified.

A dorsal approach to the affected vertebrae was made, and a dorsal laminectomy was performed over the body of L₁. The dorsal lamina was removed to the cranial and caudal ends of L₁. The vertebral canal contained a large amount of firm and hyperemic epidural fat. Upon closer inspection, a 2-mL pocket of tenacious, white, purulent-appearing fluid in the epaxial musculature was found communicating with the transverse process of L₁. The transverse process was soft and irregular; however, the spinal cord appeared grossly normal. Normal tissue within the fluid pocket was debrided, and aerobic cultures and a biopsy of the epidural fat were obtained. Unfortunately, cytopathological evaluation of the fluid was not performed. After copious lavage, the laminectomy site was closed in a routine manner, and a 5-mm, flat, closed-suction surgical drain^h was placed and connected to the suction reservoir. Biopsy results were consistent with a chronic-active steatitis that was hemorrhagic and multifocal. Completed aerobic culture results were negative (i.e., no growth occurred within a 5-day period). Anaerobic cultures were not performed. Medical therapy consisted of cefazolin^g (22 mg/kg body weight, IV), oxymorphone^b (0.05 mg/kg body weight, IM), and nursing care. Postoperative instructions included amoxicillin/clavulanic acidⁱ (500 mg orally twice daily for 14 days), bladder expressions, and restriction of activity.

The dog presented for suture removal at 2 weeks postoperatively. She was still nonambulatory but could support weight in the hind limbs. At the 4-week follow-up, the dog was fully ambulatory and had no neurological deficits.

Discussion

The authors’ findings are consistent with spinal epidural empyema as an extension from a paraspinal abscess. Bacterial infections of the central nervous system (CNS) are rarely reported.¹² Potential causes of spinal epidural empyema include direct extension from a nearby contaminated area, such as suspected in this case, or through hematogenous bacterial spread.^1–4

The most frequently reported clinical signs include fever, spinal hyperesthesia, and spinal cord dysfunction, which often rapidly progresses to para- or tetraparesis.¹² Based on the human literature, the authors can extrapolate that an early, accurate diagnosis increases the likelihood of successful treatment. There have been only two previous reports in the veterinary literature. In one report of two cases, both dogs were euthanized at the owner’s request, due to the extensive nature of the disease processes and the poor clinical status.¹ The second report was the only documented case in which complete neurological recovery occurred following decompressive surgery.⁷ The most common etiological agents reported for spinal epidural empyema in the dog and human include Staphylococcus and Streptococcus species.¹⁵⁶ A number of other organisms have also been identified. Skin infections are thought to be a common source of the bacteria involved.¹⁶ Neurological dysfunction in this case was caused by direct compression of the spinal cord from the steatotic epidural fat and potentially was a result of insult to the cord from vasoactive substances released by bacteria.¹⁵⁶ Cerebrospinal fluid analysis was consistent with findings in previous reports and consisted of a mixed pleocytosis with an elevated protein level.¹⁵⁶ It is also possible for CSF analysis to be normal, because the disease is confined to the epidural space.² Historically, myelography has been the most widely used diagnostic tool. Computed tomography scanning or magnetic resonance imaging (MRI) may be better diagnostic modalities, due to their noninvasive nature.^12568–10

Differing opinions exist concerning medical versus surgical management of spinal epidural empyema.¹²⁵⁶ Because of the severe neurological dysfunction and spinal cord compression, a decompressive laminectomy was performed in the case of this study. Most clinicians are in favor of surgical drainage in combination with antibiotic therapy.¹²⁵⁶ Nonsurgical management has been successful in humans.^10–12 Medical management may be reserved for patients with minimal neurological deficits, patients who are poor anesthetic candidates, large lesions where surgical decompression would lead to vertebral instability, and complete paralysis or loss of deep pain sensation for >72 hours.^13–15

A definitive source of infection was not identified in this case. The paraspinal abscess most likely formed from the bite wound that occurred in the area 3 weeks prior to presentation, although no external puncture wounds or scars were found on physical examination. The negative aerobic culture results may have been influenced by the use of broad-spectrum antibiotics preoperatively. Spinal epidural empyema has been associated with immunosuppression and immaturity of the immune system, which may have been true in this case.¹

Conclusion

Spinal epidural empyema should be considered as a differential diagnosis for any dog that presents with a fever and a rapidly progressive myelopathy. Aggressive surgical decompression and drainage, in combination with appropriate antibiotic therapy (based on culture and sensitivity results), are indicated when there is evidence of marked spinal cord compression and severe neurological dysfunction.