Optimal duration of antibiotic therapy in vertebral osteomyelitis.

Objectives: To compare the risk of relapse of vertebral osteomyelitis (VO), according to the duration of antibiotic therapy (< or =6 weeks versus >6 weeks).

Methods: We performed a 10-year retrospective study to assess the risk of VO relapse and to verify that this risk was not enhanced in patients who received 6 weeks of antibiotic therapy (Group 1) as compared with those who received a longer treatment (Group 2). VO was diagnosed based on clinical manifestations, magnetic resonance imaging and/or computed tomography findings, and isolation of a pyogenic organism in blood cultures and/or a discovertebral biopsy. Relapse was diagnosed based on isolation of the same organism in blood cultures and/or a discovertebral biopsy. Outcome was evaluated 6 months post-treatment and in December 2004.

Results: Group 1 included 36 patients (mean age, 58 +/- 15 years) and Group 2 included 84 patients (mean age, 67 +/- 15 years) (P = 0.003). Clinical data and microorganisms were comparable in the 2 groups. In the first 6 months, 6 (5%) patients died (Group 1, n = 2; Group 2, n = 4), and 5 (4%) in Group 2 relapsed, 2 with recurrent VO and 3 with recurrent bacteremia. In 2004, 91 patients were evaluated (mean follow-up, 40.6 +/- 31 months): 77 (85%) were cured, 13 (14%) died (Group 1, n = 3; Group 2, n = 10), 1 had VO due to a different microorganism (Group 2), and no long-term relapses occurred.

Conclusion: Our results suggest that antibiotic therapy of VO could be safely shortened to 6 weeks without enhancing the risk of relapse.