Friday, February 5, 2010

Epidural Abscess

We have seen a number of cases of epidural abscess this week. Reviews here (Lancet Neurology), and here (NEJM). Not to be confused with the coolest name in medicine Pott's puffy tumour as described here.

Epidemiology:
  • Rare - about 0.2 to 2 per 10,000 admissions
  • Often underlying condition:
    • Diabetes
    • Alcoholism, Injection Drug Use
    • HIV
    • Chronic spine disease
    • Previous spine surgery
    • Indwelling catheters
Pathogenesis:
  • Contiguous spread of infection from soft tissue/vertebrae/manipultion (~30%)
  • Hematogenous seeding ~50%
  • Unknown 15%
  • Can injure spinal cord leading to paralysis due to compression (local) or small septic emboli causing infarction.
Microbiology:
  • Most ~60% are Staphylococcus aureus including MRSA
  • Can be CNST if hardware in situ
  • E. coli from bacteremic urinary tract infection
  • Pseudomonas (often injection drug use)
  • Other: enterococcus, viridans group streptococcus, mycobacteria, fungus (candida, other) etc.
Diagnosis:

History:
  • Pain is common, initially localized, then radicular and referred (85%)
  • Increasing weakness as cord is involved and sensory loss
  • Bladder and bowel involvement occur later then fullparalysis
  • Fever may be absent! Especially if NSAIDS or acetominophen for pain
Labs:
  • Normal WBC in up to 40%!
  • Most have elevated CRP and ESR
  • Bacteremia in up to 60%
Imaging:
  • MRI is test of choice.
  • Concomittant osteomyelitis seen in ~80%. This may be seen on CT; but the abscess can be missed with plain CT.
Treatment:
  • If stable and no imminent neurologic comprimise, it is best to get a definitive diagnosis. Blood cultures should be sent and operative specimens. Then antibiotics can be started tailored to the etiologic agent.
  • Usually combined medical and surgical management is required. The deficits can progress exceptionally quickly (i.e. complete paralysis within hours) and can be irreversible (particularly if paralysis has been for more than 36 hours!
  • If empiric coverage is required, coverage for Staphylococcus aureus, including MRSA as well as gram negatives should be started. Antipseudomonal coverage may be required if there is a risk for pseudomonas -- i.e. IDU, known pseudomonas elsewhere.
    • One reasonable approach would be to start VANCOMYCIN (MRSA) and either CEFEPIME, PIPERACILLIN-TAZOBACTAM, or CEFTAZADIME
Common pitfalls - See figure from NEJM below:

Prognosis:
  • Paralysis will develop in 4-22%. Paralysis is likely to be irreversible after 24-36 hours
  • With prompt surgery most patients have as good or better neurologic status than they did before the OR.
  • Death does occur in about 5%

No comments:

Post a Comment