Changing trends in the epidemiology of pyogenic vertebral osteomyelitis: the impact of cases with no microbiologic diagnosis.

Objectives: The observed higher incidence of pyogenic vertebral osteomyelitis (PVO) may entail an increasing number of patients with no microbiologic diagnosis. The true incidence of these cases, how exhaustive the etiologic diagnostic efforts must be, and the usefulness of an empirical antibiotic therapy are not well defined.

Methods: Retrospective analysis of all cases of vertebral osteomyelitis in our center (1991-2009) and retrospective analysis of cases of PVO (2005-2009). Clinical data, diagnostic procedures, treatment, and outcome were reviewed. A comparative analysis between microbiologically confirmed PVO (MCPVO) and probable PVO (PPVO) was performed.

Results: Increasing incidence of PVO (+0.047 episodes/100,000 inhabitants-year). During the last decade, there was an increase of PPVO (+0.059 episodes/100,000 inhabitants-year) with stable incidence of MCPVO. During 2005-2009, there were 72 patients [47 (65%) MCPVO and 25 (35%) PPVO]. 60% men; mean age was 66 years. Bacteremia was found in 59%. Computed tomographic guided vertebral biopsy, positive in 7/36 (19%), was more successful among patients with bacteremia. Among MCPVO, there was an increasing proportion of less virulent bacteria. Cases of MCPVO presented more frequently with sepsis, fever, and high acute-phase reactants, and PPVO cases were mostly treated with oral fluoroquinolones plus rifampin. No differences were found between both groups in outcome (93% success, 22% sequelae).

Conclusions: An epidemiologic change of PVO is suggested by a higher incidence of PPVO and the isolation of less virulent microorganisms among MCPVO. In this setting, the availability of an oral and effective empirical antibiotic therapy may challenge an exhaustive prosecution of the etiology.