Antibiotic susceptibility of Gram-negatives isolated from bacteremia in children with cancer. Implications for empirical therapy of febrile neutropenia

Background: Monotherapy is recommended as the first choice for initial empirical therapy of febrile neutropenia, but local epidemiological and antibiotic susceptibility data are now considered pivotal to design a correct management strategy. Aim: To evaluate the proportion of Gram-negative rods isolated in bloodstream infections in children with cancer resistant to antibiotics recommended for this indication. Materials & methods: The in vitro susceptibility to ceftazidime, piperacillin-tazobactam, meropenem and amikacin of Gram-negatives isolated in bacteremic episodes in children with cancer followed at the Istituto Giannina Gaslini, Genoa, Italy in the period of 2001-2013 was retrospectively analyzed using the definitions recommended by EUCAST in 2014. Data were analyzed for any single drug and to the combination of amikacin with each beta-lactam. The combination was considered effective in absence of concomitant resistance to both drugs, and not evaluated by means of in vitro analysis of antibiotic combinations (e.g., checkerboard). Results: A total of 263 strains were evaluated: 27% were resistant to piperacillin-tazobactam, 23% to ceftazidime, 12% to meropenem and 13% to amikacin. Concomitant resistance to beta-lactam and amikacin was detected in 6% of strains for piperacillin-tazobactam, 5% for ceftazidime and 5% for meropenem. During the study period there was a nonsignificant increase in the proportions of strains resistant to beta-lactams indicated for monotherapy, and also increase in the resistance to combined therapies. Conclusion: in an era of increasing resistance to antibiotics guideline-recommended monotherapy could be not appropriate for initial empirical therapy of febrile neutropenia. Strict local survey on etiology and antibiotic susceptibility is mandatory for a correct management of this complication in cancer patients.