comunicación de congreso
A nomogram for predicting serious complications in patients with solid tumors and apparently stable febrile neutropenia: prospective data on 781 consecutive episodes from the FINITE study
Fecha
2014Registro en:
10.1200/jco.2014.32.31_suppl.165
Autor
Ghanem, Ismael
Rebollo, Maite Antonio
Garrido, Marcelo
Martínez, Jerónimo
Font, Carme
Ramchandani, Avinash
Biosca, Merce
Beato, Carmen
Martínez de Castro, Eva
Castanon, Eduardo
Virizuela Echaburu, Juan
Espinosa, Javier
Sevillano, Elena
Aragon Manrique, Isabel
Cardona, Merce
Mondejar, Rebeca
Baron, Francisco
Acevedo Claros, Francisco Nicolás
Jiménez-Fonseca, Paula
Carmona Bayonas, Alberto
Institución
Resumen
Background: An accurate estimate of the likelihood of serious complications in patients with otherwise apparently stable febrile neutropenia (FN) may assist in decision-making regarding individualized therapy. Our group has developed a prognostic score for predicting complications in patients with solid tumors and apparently stable episodes called CISNE (Clinical Index for Stable Febrile Neutropenia). The purpose of this study is to present a nomogram based on the previously mentioned index in a broader dataset of patients. Methods: FINITE is a prospective and multicenter study which aims to investigate prognostic factors and outcomes of FN episodes with clinical stability at first assessment, defined as events without acute organ dysfunction, vital signs abnormalities or major infections. We performed a nomogram based on the CISNE score which includes the following prognostic variables: ECOG PS≥2, chronic obstructive pulmonary disease, cardiovascular disease, mucositis NCI grade ≥2, monocytes <200/mm3 and stress-induced hyperglycemia. A calibration plot was used to analyze the accuracy of this multivariate nomogram. Results: From October 2012 to December 2013, 781 patients with apparently stable FN were recruited in 21 Spanish hospitals. The rate of infection-related complications and death was 15.6% (95% confidence interval [CI], 12.9-18.6%) and 1.7% (95% CI, 0.98%-3.01%). A nomogram was designed according to the CISNE score. The area under the ROC curve was 0.836 (95% CI, 0.808-0.861). The observed and predicted probabilities also matched closely. Conclusions: Our group has developed a user-friendly nomogram for predicting complications in patients with apparently stable FN. This nomogram may be particularly useful to prevent premature discharges of cancer patients starting inpatient management.