Empirical antibiotics against Gram-positive infections for febrile neutropenia: systematic review and meta-analysis of randomized controlled trials

Objectives: To assess the value of empirical anti-Gram-positive antibiotics for the treatment of febrile neutropenia. Methods: Systematic review and meta-analysis of randomized controlled trials comparing antibiotics with anti-Gram-positive spectrum to control or placebo, in addition to the same baseline antibiotic regimen in both arms. We searched MEDLINE, EMBASE, LILACS, the Cochrane Library, conference proceedings, and references. No restrictions on inclusion were imposed. Two reviewers independently applied selection criteria, carried out quality assessment, and extracted the data. Relative risks with 95% confidence intervals were pooled using the fixed effect model. The primary outcome assessed was all-cause mortality. Results: Thirteen studies met inclusion criteria, including 2392 participants. Glycopeptides were assessed in nine trials. Empirical anti-Gram-positive antibiotics were assessed for the initial treatment in 11 studies, and for persistent fever in two. No significant difference in all-cause mortality was seen [RR 0.86 (0.58-1.26), seven studies, 852 participants]. Overall failure at end of therapy occurred equally [RR 1.00 (0.79-1.27), six studies, 943 participants]. Failure associated with treatment modifications was more frequent in the control arm when empirical initial glycopeptides were assessed [RR 0.70 (0.61-0.80), five studies, 1178 participants]. Bacterial superinfections, mainly Gram-positive, were detected less frequently in the intervention arm. Adverse events were significantly more common with the additional antibiotic, and nephrotoxicity was significantly more common with additional glycopeptides [RR 1.88 (1.10-3.22), six studies, 1282 participants]. No significant heterogeneity was present in these comparisons. Conclusions: The use of glycopeptides can be safely deferred until the documentation of a resistant Gram-positive infection.