Patient outcomes associated with carbapenem-sparing management of febrile neutropenia

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Background Febrile neutropenia (FN) guidelines recommend that local epidemiology inform empiric antibacterial treatment, but that efforts be made to restrict carbapenem administration to unstable patients. At our institution most stable patients are treated with cefepime. We examined the clinical consequences of this regimen in the context of rising rates of antimicrobial resistance (AR). Methods This retrospective single-center study included patients <22 years of age with cancer or HCT who had received immunosuppressive therapy within 3 months, had an absolute neutrophil count (ANC) <500 cells/μL, a laboratory confirmed Gram-negative bloodstream infection (GNBSI) between 11/1/21 and 2/28/25 and who received empiric cefepime for FN. Demographic and clinical data were abstracted from medical records. Results A total of 142 episodes of GNBSI were identified among 100 patients with neutropenia. Of these, 79 episodes were caused by cefepime-susceptible (CS) isolates, and 63 by cefepime-nonsusceptible (CN) isolates. Compared with CS episodes, CN episodes were associated with older age (median 12.4 vs. 8.1 years; p = 0.025) and Hispanic ethnicity (51% vs. 22%; p < 0.001). Patients with CN were more likely to have leukemia, whereas solid tumors were more common among patients with CS episodes (p<0.001) (Table 1). The median ANC was 10 cells/μL in both groups. A similar proportion of patients in each group were initially admitted to the intensive care unit (ICU). Patients with CN isolates were more likely to have focal infections related to their GNBSI. Compared with CS episodes, CN episodes were associated with significant delays in the administration of effective therapy (median 17 vs. 0 hrs, p < 0.001). Empiric therapy for CN infections was more likely to be modified because of poor clinical response (98% vs. 44%, p < 0.001). The proportion of episodes requiring ICU transfer after admission and the duration of bacteremia and fever were similar. CN episodes had longer times to clinical resolution (median 4 vs. 2 days) and longer hospitalizations (12 vs. 5 days). Complications were more common in CN episodes (46% vs. 22%) but mortality at hospital discharge and 30-day all-cause mortality were similar to those with CS episodes. Conclusion In this setting, a carbapenem-sparing strategy for management of FN was associated with a significant delay in effective antimicrobial therapy and poorer clinical outcomes in patients with CN infections. Refinements in risk stratification and the rapid identification of resistant bacteria are needed to optimize the management of GNBSI while limiting the development of AR.