Epidemiology of Febrile Neutropenia in Children With Central Nervous System Tumor: Results From a Single Center Prospective Study Elio Castagnola, MD,* Maria Luisa Garre`, MD,w Luisella Bertoluzzo, PhD,z Sara Pignatelli, MD,w Marco Pavanello, MD,y Ilaria Caviglia, PhD,* Silvia Caruso, PhD,z Francesca Bagnasco, PhD,z Cristina Moroni, MD,* Angela Tacchella, MD,* and Riccardo Haupt, MDz Summary: Data regarding the epidemiology febrile neutropenia during chemotherapy for pediatric central nervous system neopla- sia are scarce. Data retrieved from a prospective study performed from January 2002 to December 2004 at G.Gaslini Children Hospital, Genoa, Italy, where analyzed to evaluate proportions, rate for 1000 neutropenic days and etiology of fever in neutropenic children receiving gentle, standard, or peripheral blood stem cell transplant (PBSCT) therapy for central nervous system tumor. During the study duration, 243 periods of neutropenia (granulocyte count <1000/cmm), accounting for 3544 patient-days at risk, were documented in 62 children. A total of 72 febrile episodes were observed in 66 (27%) neutropenic periods, for a rate of 20.31. A primary febrile episode was observed in 10% of neutropenic periods after gentle chemotherapy, in 30% after standard chemotherapy, and in 48% after PBSCT (P<0.0001). The rate of primary febrile episodes was 6.19 after a gentle chemotherapy, 27.02 after standard treatment, and 31.02 after PBSCT (P<0.0001). In a multivariable regression model, the type of chemotherapy (gentle vs. standard and PBSCT) and the thresholds of granulocyte count at neutropenia onset (999-501/cmm and 500- 101/cmm vs. r100/cmm) were the only factors significantly associated with the development of febrile neutropenia. Key Words: febrile neutropenia, central nervous system neoplasia, children (J Pediatr Hematol Oncol 2011;33:e310–e315) C hemotherapy has been becoming more important in the multimodal approach to children with central nervous system (CNS) tumors. Its use either before or after surgery or radiotherapy has allowed performing less demolitive interventions, longer survival, and in some circumstances, reduction of early and late complications. CNS tumor patients, however, might be at increased risk for complications during chemotherapy-induced neu- tropenia, either because of previous surgery or, in some circumstances for the presence of CNS peritoneal or external derivation. Despite the frequency of pediatric CNS tumors, 1 no specific study deals with infections in this patients’ population, especially during neutropenic phases, when there is the highest risk of infectious complications. 2 As part of a prospective, single center, observational study 3 on febrile neutropenia in children receiving anti- neoplastic chemotherapy, we evaluated the epidemiology of febrile episodes during chemotherapy-induced neutropenic periods in children with CNS tumors. PATIENTS AND METHODS To be eligible for this study, children (age 0 to 14 y) diagnosed with a CNS tumor, had to be admitted for treatment at the Department of Hematology and Oncology or of Neurosurgery of the G. Gaslini Children’s Hospital, in Genoa, Italy, and to have experienced at least 1 neutropenic period between January 1st, 2002 and Decem- ber 31st, 2005. A previous institutional study, performed between 2002 and 2004 on the incidence of febrile neutropenia in children with cancer, included a subgroup of patients also enrolled in this study. 3 Neutropenia was defined as an absolute granulocyte (PMN) count <1000 cells/mm 3 . The duration of the neutropenic periods was calculated as the difference between the date of the last day of neutropenia (or date of death) and the date of the first day of neutropenia plus 1. For this purpose, 2 trained investigators prospectively reviewed the daily white blood cell counts of inpatients and outpatients and registered any new neutropenic period that was then monitored for any febrile complication. Because the evaluation of white blood cell count during chemotherapy-induced neutropenia was performed at least every 3 days both for inpatients and outpatients, neutropenic periods of a duration <3 days were removed from the analysis; these periods were considered to be the result of individual or instrumental variability. Similarly, consecutive neutropenic periods separated by r3 days were pooled. 3 Fever was defined as 1 oral or axillary temperature measurement Z38.51C or 2 measurements Z38.01C separated by at least 1 hour. 2 Protocols of antineoplastic chemotherapy were those adopted by the Italian Association of Pediatric Hematology and Oncology or by the SIOPE CNS tumors committee, and were all approved by the institutional ethics committee. Because of the lack of an internationally approved toxicity grading of therapeutic protocols phases of antineoplastic therapy were arbitrarily classified 3–5 in 3, mutually exclusive categories (i) conditioning regimens with high doses of alkylating agents and platinum compounds followed by Received for publication November 19, 2010; accepted June 7, 2011. From the *Infectious Diseases Unit; wSection of Neuro-Oncology; zEpidemiology and Biostatistics Section, Scientific Directorate; and yUnit of Neurosurgery; “G. Gaslini” Children Hospital, Genoa, Italy. Supported partially by grants from “Associazione Neuroblastoma,” “Associazione per la Ricerca sui Tumori Cerebrali del Bambino,” and from “G. Gaslini” Children Hospital. The authors declare no conflict of interest. Reprints: Elio Castagnola, MD, Infectious Diseases Unit, Department of Hematology and Oncology, “G. Gaslini” Children’s Hospital, Largo G. Gaslini, 5, 16147, Genoa, Italy. E-mail: eliocastagnola @ospedale-gaslini.ge.it). Copyright r 2011 by Lippincott Williams & Wilkins ORIGINAL ARTICLE e310 | www.jpho-online.com J Pediatr Hematol Oncol  Volume 33, Number 7, October 2011