Cefoperazone–sulbactam versus cefepime and piperacillin–tazobactam for initial empiric treatment of febrile neutropenia in patients with hematologic malignancies: A retrospective matched-cohort study

OBJECTIVES: Empiric antipseudomonal β-lactams are the standard of care for febrile neutropenia (FN), but comparative data for cefoperazone-sulbactam are limited. We compared cefoperazone-sulbactam with cefepime and piperacillin-tazobactam as initial empiric therapies for FN in adults with hematologic malignancies. METHODS: We conducted a retrospective matched-cohort study at a tertiary center in Taiwan (2021-2023). Adults with hematologic malignancies who received cefoperazone-sulbactam, cefepime, or piperacillin-tazobactam as the initial empiric therapy for FN were included (53 episodes/regimen). The episodes were matched by calendar time to reduce confounding factors related to changes in empiric antibiotic use. The primary outcome was treatment success, defined as defervescence without modification of the initial empiric β-lactam regimen and day 5 survival. RESULTS: Among the 159 included episodes, treatment success occurred in 120/159 (75.5%) episodes and did not differ across regimens (cefoperazone-sulbactam, 79.2%; cefepime, 71.7%; piperacillin-tazobactam, 75.5%; P = 0.70). Neutrophil recovery by day 5 independently predicted treatment success (adjusted odds ratio, 3.32; 95% confidence interval, 1.15-9.63; P = 0.027), but empiric β-lactam selection did not. The median duration of empiric β-lactam therapy was similar across regimens. Bleeding (3/159, 1.9%) and prothrombin time/international normalized ratio prolongation (2/46, 4.3%) were uncommon and similar across regimens. CONCLUSIONS: No significant differences in early clinical effectiveness or overall safety were observed among the three regimens. Neutrophil recovery, rather than empiric regimen, independently predicts treatment success. These retrospective findings should be interpreted cautiously given potential residual confounding by indication and warrant confirmation in larger studies.