Abstract 2153P
Background
Febrile neutropenia (FN) is one of the most frequent and life-threatening complications in cancer patients. The most widely used model to identify low-risk FN episodes, candidates for outpatient management, is the Multinational Association for Supportive Care in Cancer (MASCC). However, its ability to predict serious complications is not optimal, occurring in up to 15% of episodes classified as low risk. The identification of circulating biomarkers may refine these indexes. We aimed to evaluate the role of C-reactive protein (CRP) and procalcitonin (PCT) in predicting the development of serious complications in low-risk FN episodes.
Methods
A retrospective study including FN episodes in patients with solid tumors presenting to the Emergency Department (ED), classified as low risk (MASCC score ≥ 21). In all episodes, a blood sample was collected on admission to the ED for analysis, including CRP and PCT. Outcome was defined as Intensive Care Unit (ICU) admission and/or death and/or other serious complications. To evaluate the discriminatory ability of biomarkers for outcome, Receiver Operating Characteristic (ROC) curve analysis was used and optimal cutoffs were calculated through Youden index.
Results
During the study period, 132 low-risk episodes were included (mean age: 58.4; SD: 11.5 years; 83 (62.9%) female). Serious complications were developed in 52 (39.4%) episodes. Median CRP and PCT levels were significantly higher in patients developing complications: 11.8 mg/dL (IQR: 5.21-19.6) vs 5.3 mg/dL (IQR: 2.2-9.0), p < 0.001; 0.42 ng/mL (IQR: 0.14-1.09) vs 0.12 ng/mL (0.06-0.23), p < 0.001, respectively. In ROC AUC analysis, both biomarkers achieved a similar predictive accuracy (CRP: 0.733 and PCT: 0.744). Optimal cutoffs were 9.8 mg/dL for CRP (Sensitivity (S): 61.5%, Specificity (Sp): 80%, positive predictive value (PPV): 66.7% and negative (NPV): 76.2%) and 0.27 ng/mL for PCT (S: 61.5%, Sp: 81.3%, PPV: 68.1%, NPV: 76.5%).
Conclusions
In chemotherapy-associated FN patients, circulating CRP and PCT levels might be useful to improve the risk stratification of low-risk episodes, contributing to decision making in the management of these patients.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
2150P - Clinical and molecular characterisation of cancer-associated venous thromboembolism (VTE)
Presenter: Pau Mascaró Baselga
Session: Poster session 07
2151P - Incidence patterns and clinical implications of venous thromboembolism (VTE) in patients (pts) with neuroendocrine tumors (NET)
Presenter: Jorge Hernando
Session: Poster session 07
2152P - Prognostic importance of GRIM score on sorafenib side effects, hospitalization and survival in hepatocellular cancer patients
Presenter: Duygu Ercan Uzundal
Session: Poster session 07
2154P - Total cholesterol as an independent prognostic factor in cancer patient survival: A single institution experience
Presenter: Maria Alameda-Guijarro
Session: Poster session 07
2156P - Use of taxanes during pregnancy: A systematic review
Presenter: Ana Ferrigno Guajardo
Session: Poster session 07
2157P - A phase II randomised controlled trial comparing the cardiotoxicity of capecitabine and S-1
Presenter: Sally Clive
Session: Poster session 07
2158P - A self-assessment survey to identify the risk of patient’s screen fail in phase I clinical trials (SCITEP trial)
Presenter: Kaissa Ouali
Session: Poster session 07
2159P - Machine learning algorithms to predict the risk of chemotherapy-related toxicity in older Asians
Presenter: Mingwei Li
Session: Poster session 07
2160P - Sociodemographic disparities associated with financial toxicity in stage IV non-small cell lung cancer survivors
Presenter: Mitchell Parma
Session: Poster session 07
2161P - Organization of hospital pharmaceutical consultations for cancer patients receiving oral anticancer drugs: A nationwide cross-sectional study
Presenter: Florian Slimano
Session: Poster session 07