Spinal Epidural Abscess

Definition A collection of pus (purulent material) within the epidural space of the spinal canal.

Epidemiology

• Incidence is rising (aging population, IVDU, spinal procedures).
• Common in men (2:1).
• Peak age 50-70.

Pathophysiology

• Bacteria enter the epidural space via hematogenous spread (skin, UTI) or direct extension (discitis).
• The infection causes mechanical compression of the cord/cauda equina.
• It also causes septic thrombophlebitis of the epidural veins → Venous congestion → Cord ischemia → Infarction.

Epidural Space

• Potential space between the Dura Mater and the Periosteum/Ligamentum Flavum.
• Contains fat and the Batson's Venous Plexus.

Batson's Plexus

• Valveless venous system.
• Allows retrograde spread of infection from pelvic organs (e.g., during coughing/straining) to the spine.
• Explains why UTI is a common source.

Anterior vs Posterior

• Posterior: Most common in Thoracic/Lumbar spine (more epidural fat posteriorly).
• Anterior: Associated with Vertebral Osteomyelitis/Discitis (direct extension).

History

• Classic Triad: Fever + Back Pain + Neurology. (Only 10-15% sensivity).
• Back Pain: Most common symptom (greater than 75%). Severe, unrelenting, night pain.
• Fever: Only present in ~50%.
• History of: IVDU, Diabetes, Recent spinal injection, UTI.

Examination

• Spine: Focal percussion tenderness (Highly suspicious).
• Neurology:
  • Assess Power (Myotomes).
  • Assess Sensation (Fluid level? Saddle anesthesia?).
  • Assess PR Tone/Sensation (Cauda Equina).

Red Flags

• New onset back pain in an IV Drug User = SEA until proven otherwise.

Medical Management

• Reserved for:
  1. Neurologically intact patients.
  2. Known organism (Blood Cx or Biopsy positive).
  3. Too unfit for surgery.
  4. Complete paralysis greater than 48-72 hours (salvage unlikely).
• Antibiotics: Empiric (Vanc + Ceftriaxone) → Targeted.
• Monitoring: Daily neurological checks. Serial MRI if worsening.

Failure of Medical Management

• Defined as:
  • New neurological deficit.
  • Persistent fever/CRP elevation.
  • Increasing pain.
  • Enlargement of abscess on MRI.

The classification section notes that anterior and circumferential collections are "harder to access" and may need "corpectomy or a transpedicular approach," but the surgical technique above describes only laminectomy — so the key principle deserves stating: the approach must match the abscess location relative to the cord.

• Posterior collection (the majority) → posterior laminectomy and evacuation, as described above. This reaches the pus directly and is the default.
• Anterior collection (usually accompanying vertebral osteomyelitis/discitis) → laminectomy alone is inadequate (it cannot reach an anterior collection) and destabilising (especially if facets are sacrificed). These need an anterior approach (corpectomy, debridement and anterior-column reconstruction with a strut graft or cage) or a posterolateral route (costotransversectomy in the thoracic spine, or transpedicular drainage) — often with posterior instrumented stabilisation.
• Circumferential disease → may require combined anterior and posterior (360-degree) surgery.

The stabilisation caveat. Both anterior corpectomy/debridement and multilevel laminectomy can render the spine unstable, so titanium instrumentation may be required despite the active infection (the debate covered in the Stabilization tab) — contemporary series support fusion with adequate debridement and antibiotics.

Antibiotics

• Long term IV (usually 6-8 weeks).
• Oral suppression may be needed lifelong if implant retained (rare).

Rehabilitation

• Spinal Cord Injury protocol if deficit persists.
• Bladder/Bowel management.
• Pressure area care.

Prognostic Factors

• Pre-operative Neurology: The single most important factor.
• Duration of Deficit: Less than 36 hours has better prognosis.
• Age: Greater than 65 has worse outcome.
• Diabetes: Associated with higher mortality.

Recovery

• Complete recovery is rare if paralysis has set in.
• Early decompression (less than 24 hours) yields best results.

Global Epidemiology

• Incidence is rising worldwide: historically 0.2–2 per 10,000 hospital admissions (Reihsaus), now commonly quoted at 2–8 per 10,000 in tertiary centres, driven by an ageing population, diabetes, injection drug use and increasing spinal instrumentation.
• Male predominance (roughly 1.5–2:1); peak age 50–70 years.
• Staphylococcus aureus causes 60–70% of cases; the proportion of MRSA varies markedly by region and rises where community-acquired MRSA is endemic. Gram-negative organisms (e.g. E. coli) point to a urinary source; consider Mycobacterium tuberculosis (Pott disease) and Brucella in high-prevalence and limited-resource settings.

Guidance — Areas of Broad International Agreement

Where Practice Genuinely Varies

• Empirical regimen: vancomycin (MRSA cover) plus a broad Gram-negative agent (e.g. ceftriaxone, cefepime or an antipseudomonal in IVDU/healthcare-associated cases) is widespread, but local antibiograms and MRSA prevalence drive the exact choice. Where MRSA is rare, an anti-staphylococcal penicillin may suffice.
• Initial non-operative trial: more readily adopted in high-resource centres with rapid MRI access and the ability to monitor with frequent neuro-obs and serial imaging; risky where MRI and theatre access are limited, so earlier surgery is often favoured.
• Source-setting differences: in TB-endemic and limited-resource regions, tuberculous epidural abscess is a major differential and changes both imaging interpretation and drug therapy.

Registry & Outcome Notes

• SEA is not tracked in arthroplasty registries; the evidence base is retrospective cohorts and meta-analyses (Reihsaus; Curry; Kim) rather than RCTs. No high-quality randomised data compares medical vs surgical management — a key limitation acknowledged across guidelines.

The guidelines above flag tuberculosis (Pott disease) and Brucella as differentials, and a viva will ask how a TB-endemic background changes your thinking — so it is worth setting out, because the acute-pyogenic playbook will miss these organisms.

When to suspect it. A patient from a TB-endemic region, or who is immunosuppressed or HIV-positive, with an insidious course over weeks to months rather than the acute pyogenic picture, often with constitutional symptoms (night sweats, weight loss) and a relatively "cold" abscess (less florid local inflammation).

How it differs. Tuberculous spinal infection typically arises anteriorly from the vertebral body and disc (spondylodiscitis), with progressive bony destruction, relative early disc preservation, kyphosis/gibbus, and large paravertebral or psoas collections that track far from the vertebral focus (subligamentous spread to multiple levels); the thoracolumbar spine is the predilection site.

Why the pyogenic approach fails — and what to do:

• Routine bacterial cultures and Gram stain will not grow it: you need tissue (image-guided or operative) for AFB smear, mycobacterial culture, GeneXpert/PCR and histology (caseating granulomas).
• Empirical vancomycin-plus-Gram-negative cover does not treat it: treatment is prolonged multi-drug anti-tuberculous chemotherapy (months), not the 6-week pyogenic course. Surgery is reserved for neurological deficit, instability/deformity or a large abscess. Brucella similarly needs serology and prolonged combination antibiotics.
• (Detailed Pott-disease classification, deformity correction and the anti-TB regimen are developed in the tuberculosis-spine-potts topic; brucellar spondylitis in the brucellosis-spine topic.)