Medical and surgical management of spinal epidural abscess: a systematic review.

Object: Spinal epidural abscess (SEA) is a rare condition that has previously been treated with urgent surgical decompression and antibiotics. Recent availability of MRI makes early diagnosis possible and allows for the nonoperative treatment of SEA in select patients. The first retrospective review of medically and surgically managed SEA was published in 1999, and since that time several other retrospective institutional reports have been published. This study reviews these published reports and compares pooled data with historical treatment data.

Methods: A PubMed keyword and Boolean search using ("spinal epidural abscess" OR "spinal epidural abscesses" AND [management OR treatment]) returned 429 results. Filters for the English language and publications after 1999 were applied, as the first study comparing operative and nonoperative management was published that year. Articles comparing operative to nonoperative treatment strategies for SEA were identified, and the references were further reviewed for additional articles. Studies involving at least 10 adult patients (older than 18 years) were included. Case reports, studies reporting either medical or surgical management only, studies not reporting indications for conservative management, or studies examining SEA as a result of a specific pathogen were excluded.

Results: Twelve articles directly comparing surgical to nonsurgical management of SEA were obtained. These articles reported on a total of 1099 patients. The average age of treated patients was 57.24 years, and 62.5% of treated patients were male. The most common pathogens found in blood and wound cultures were Staphylococcus aureus (63.6%) and Streptococcus species (6.8%). The initial treatment was surgery in 59.7% of cases and medical therapy in 40.3%. This represented a significant increase in the proportion of medically managed patients in comparison with the historical control prior to 1999 (p < 0.05). Patients with no neurological deficits were significantly more likely to be treated medically than surgically (p < 0.05). There was no statistically significant difference overall between surgical and nonsurgical management, although several risk factors may predict failure of medical management.

Conclusions: Since the first reports of nonoperative treatment of SEA, there has been a substantial trend toward treating neurologically intact patients with medical management. Nevertheless, medical therapy fails in a fair number of cases involving patients with specific risk factors, and patients with these risk factors should be closely observed in consideration for surgery. Further research may help identify patients at greater risk for failure of medical therapy.