Adjunctive gentamicin did not improve outcome of enterococcal bacteraemia in neutropenic patients: a propensity scored matched study.

BACKGROUND

Patients with haematologic malignancies receiving chemotherapy have a high risk of developing febrile neutropenia and bloodstream infections. The benefit of adjunctive gentamicin treatment for enterococcal bloodstream infections is debated. In this study, we compare the treatment outcome of a cell wall-active antibiotic with and without gentamicin for enterococcal bacteraemia in patients with neutropenia.

METHODS

The observational study was performed from 1999 through 2016. Patients with bacteraemia due to non-high level gentamicin-resistant enterococci were included. Analyses were performed in two data sets of episodes with enterococcal bacteraemia. One data set consisting of all included episodes (full cohort, n = 154) and one with propensity score-matched episodes (n = 96). The primary endpoint was death within 30 days, and the secondary outcomes were defervescence and persistence of enterococcal bloodstream infection after initiation of anti-enterococcal therapy.

RESULTS

Episodes with gentamicin treatment (n = 82, full cohort; n = 48, propensity score-matched cohort) were comparable with episodes without gentamicin treatment (n = 72, full cohort; n = 48, propensity score-matched cohort) with regard to patient- and disease-related characteristics. Enterococcus faecium (40.9%) was the most frequently isolated organism. In the propensity score-matched cohort, there was no difference in 30-days mortality (14.6% in episodes with gentamicin versus 16.7% in episodes without gentamicin, p = 1), median time to defervescence (1 versus 2 days, p = .37) or persistence of enterococcal bloodstream infection for ≥72 h (9.4% versus 7.5%, p = 1).

CONCLUSIONS

In our study with neutropenic patients, treatment with a cell wall-active antibiotic without adjunctive gentamicin for episodes with enterococcal bloodstream infection was as effective as combination therapy with gentamicin.