RESUMO
BACKGROUND: There is no practice standard regarding antibiotic duration in children with cancer and unexplained febrile neutropenia (FN). We hypothesized that absolute monocyte count (AMC) and absolute phagocyte count (APC= ANC + AMC + bands) are more sensitive, earlier, and safe markers of antibiotic cessation compared with absolute neutrophil count (ANC). METHODS: A retrospective review of FN episodes (FNEs) in pediatric oncology patients was conducted between 2009 and 2016. Included patients were afebrile for 24 hours and without an identified infectious source at antibiotic cessation. Primary endpoints, including recurrent fever, readmission, bloodstream infection, microbiologically documented infection, and adverse outcomes, were assessed 10 days after antibiotic cessation and compared among different bone marrow recovery parameters (ANC, AMC, APC). Secondary endpoints included length of FN stay, antibiotic-free days, and cost. RESULTS: Three hundred ninety-one FNEs in 235 patients were included. Three groups were compared based on ANC (cells/µL) at the time of antibiotic cessation: < 200 in 102 (26%), 200 to 500 in 111 (28%), and >500 in 178 (46%). No statistically significant differences in primary endpoints were identified among the 3 ANC groups; however, a trend toward unfavorable outcomes in the ANC ≤200 cells/µL group compared with the ANC >200 cells/µL was observed. Primary endpoints based on AMC >100 cells/µL at the time of antibiotic cessation showed statistically significant favorable outcomes compared AMC ≤100 cells/µL (80%, 88%, 90%, 89%, and 93% risk reduction in recurrent fever, readmission, new bloodstream infection, new microbiologically documented infection, and adverse events, respectively). Similar favorable results were seen when APC >300 cells/µL was used as a threshold for antibiotic cessation. The median length of stay for FN if discharged when AMC >100 cells/µL was 3 days shorter and associated with fewer unfavorable outcomes, thus resulting in fewer hospital days, fewer antibiotic days, and decreased cost. CONCLUSION: Our results suggest that AMC >100 cells/µL (regardless of ANC) or APC >300 cells/µL may be safe thresholds for empiric antibiotic cessation and result in reduced unfavorable clinical outcomes within 10 days postdischarge, reduced antibiotic days of therapy and reduced health care costs. Further prospective studies are needed to validate AMC as an accurate surrogate marker for antibiotic cessation in FNEs in children with cancer.