Prompt regression of paravertebral and epidural abscesses in patients withpyogenic discitis. Sixteen cases evaluated using magnetic resonance imaging

Objective. To determine whether spinal magnetic resonance imaging performed one month into anti-microbial therapy for pyogenic discitis demonstrated c hanges of value for predicting outcomes and making therapeutic decisions. M ethods. Prospective study of 16 patients with discitis. A physical evaluati on, laboratory tests for inflammation, plain radiographs centered on the af fected vertebral level, and magnetic resonance imaging with gadolinium inje ction were performed at baseline and on day 30. All 16 patients were reeval uated after three months and 15 after six months. Results. The 12 men and f our women had a mean age of 59 years. Fourteen patients had a paravertebral abscess (n=12) and/or an epidural abscess (n=6). On day 30, 14 of the 16 p atients were clinically improved and ten were radiologically improved. The C-reactive protein level was still elevated in five cases. Ail reevaluated patients were improved after three months (16/16) and six months (15/15). O nly two magnetic resonance imaging features improved during the first month : the size of paravertebral abscesses decreased in 11 of 12 patients, and t he size of epidural abscesses decreased in four of six patients. Conclusion . Paravertebral and epidural abscesses improve promptly under antimicrobial therapy. However, the presence of these lesions have no prognostic signifi cance. Magnetic resonance imaging does not add significantly to the follow- up of patients who respond clinically to antimicrobial therapy. However, in patients with an unsatisfactory clinical response, follow-up magnetic reso nance imaging can provide useful therapeutic orientation by showing whether the paravertebral and/or epidural abscesses have decreased in size. (C) 20 00 Editions scientifiques et medicales Elsevier SAS.