Febrile Neutropenia Syndromes in Children: Risk Factors and Outcomes of Primary, Prolonged, and Recurrent Fever

Abstract

Supplemental Digital Content is available in the text. Background: The approach to recurrent febrile neutropenia (FN) in children with cancer has not been sufficiently addressed and was cited as a research gap in the International Pediatric Fever and Neutropenia (IPFNP) Guideline 2017. Methods: Retrospective medical record review for all pediatric cancer patients with a diagnosis of FN was performed. Variables were collected at 2 different time sets (at day 1 and day 4 of presentation). Three FN syndromes have been defined based on the duration and time course of the fever: (1) primary: fever resolved before 96 hours and did not follow with recurrent fever; (2) prolonged fever: episodes failing to defervesce after at least 96 hours of antibacterial therapy; (3) recurrent fever: a new episode of fever >72 hours after resolution of the initial fever when a patient remained neutropenic and on antibiotics or if a fever developed within 1 week after antibiotic discontinuation. These entities were compared with define risk factors and adverse outcomes associated with recurrent fever. Results: A total of 633 FN episodes (FNEs) were identified in 268 patients. Each FNE was classified as primary (n=453, 71.5%), prolonged (n=119, 18.7%), or recurrent (n=61, 9.7%). In multivariable analysis, acute myelogenous leukemia (odds ratio [OR]=4.6, 95% confidence interval [CI]: 2.95-7.24), allogeneic stem cell transplant (SCT) (OR=4.9, 95% CI: 2.61-7.35), absolute lymphocyte count <300/mm3 (OR=3.8, 95% CI: 1.30-5.02), prior neutropenia of ≥10 days, (OR=3.95, 95% CI: 1.70-5.93) and hypotension (OR=3.65, 95% CI: 1.30-5.86) on day 1 of presentation were all associated with an increased risk of recurrent fever when compared with primary fever. In subset analysis for only the high-risk FN group, hypotension (OR=3.2, 95% CI: 1.80-4.96), prior neutropenia ≥10 days (OR=2.55, 95% CI: 1.40-6.22), and absolute lymphocyte count <300/mm3 at presentation (OR=2.6, P=0.03, 95% CI: 2.65-7.12) were associated with an increased risk of recurrent fever when compared with high-risk FN not developing recurrent fever. Allogeneic SCT (OR=5.9, 95% CI: 2.65-7.12) and prior neutropenia ≥10 days (OR=2.11, 95% CI: 1.25-9.32) were significantly associated with recurrent fever when compared with prolonged fever. Invasive fungal disease was a more common etiology with recurrent fever compared with primary and prolonged fever (P=0.001 and 0.01, respectively). Recurrent fever episodes were more likely to be admitted to the pediatric intensive care unit (OR=3, 95% CI: 1.27-6.23) and had a higher 30-day mortality (OR=8, 95% CI: 1.87-71.85) when compared with primary fever. Conclusions: Knowledge of risk factors for recurrent fever may enable the early detection infection-related complications of this high-risk group, and possible improved approaches to treatment resulting in decreased morbidity and mortality.