Prediction of Serious Complications in Patients With Seemingly Stable Febrile Neutropenia: Validation of the Clinical Index of Stable Febrile Neutropenia in a Prospective Cohort of Patients From the FINITE Study
Menée sur 1 133 patients atteints d'une tumeur solide et présentant des épisodes de neutropénie fébrile sans signe apparent de gravité, cette étude prospective multicentrique évalue la performance d'un système de score, basé sur des données cliniques (niveau sanguin des monocytes, présence d'une maladie cardiovasculaire chronique, ...), pour prédire le risque de complications sévères
Purpose : To validate a prognostic score predicting major complications in patients with solid tumors and seemingly stable episodes of febrile neutropenia (FN). The definition of clinical stability implies the absence of organ dysfunction, abnormalities in vital signs, and major infections.
Patients and Methods : We developed the Clinical Index of Stable Febrile Neutropenia (CISNE), with six explanatory variables associated with serious complications: Eastern Cooperative Oncology Group performance status ≥ 2 (2 points), chronic obstructive pulmonary disease (1 point), chronic cardiovascular disease (1 point), mucositis of grade ≥ 2 (National Cancer Institute Common Toxicity Criteria; 1 point), monocytes < 200 per
μL (1 point), and stress-induced hyperglycemia (2 points). We integrated these factors into a score ranging from 0 to 8, which classifies patients into three prognostic classes: low (0 points), intermediate (1 to 2 points), and high risk (
≥ 3 points). We present a multicenter validation of CISNE.
Results : We prospectively recruited 1,133 patients with seemingly stable FN from 25 hospitals. Complication rates in the training and validation subsets, respectively, were 1.1% and 1.1% in low-, 6.1% and 6.2% in intermediate-, and 32.5% and 36% in high-risk patients; mortality rates within each class were 0% in low-, 1.6% and 0% in intermediate-, and 4.3% and 3.1% in high-risk patients. Areas under the receiver operating characteristic curves in the validation subset were 0.652 (95% CI, 0.598 to 0.703) for Talcott, 0.721 (95% CI, 0.669 to 0.768) for Multinational Association for Supportive Care in Cancer (MASCC), and 0.868 (95% CI, 0.827 to 0.903) for CISNE (P = .002 for comparison between CISNE and MASCC).
Conclusion : CISNE is a valid model for accurately classifying patients with cancer with seemingly stable FN episodes.
Journal of Clinical Oncology , résumé, 2015