• Patricia Torrella Esteban, Carlos Rodríguez Rojas, Sofía Wikström Fernández, Andrés Murillo Herrera, Pilar Garrido Orta, Esther Montoro Jorquera, Pablo Cerezuela-Fuentes, and de Guadiana-RomualdoLuis GarcíaLGServicio de Análisis Clínicos, Hospital Universitario Santa Lucía, Cartagena, Murcia, España..
• Servicio de Análisis Clínicos, Hospital Universitari Sant Joan d'Alacant, Alicante, España.
• Emergencias. 2024 Dec 1; 36 (6): 417424417-424.

ObjectiveTo analyze the usefulness of mean mid-regional pro-adrenomedullin (MR-proADM) level to stratify risk in emergency department patients with solid tumors attended for febrile neutropenia after chemotherapy. To compare risk prediction with MR-proADM to that of conventional biomarkers and scores on the Multinational Association for Supportive Care in Cancer (MASCC) score.MethodsProspective observational cohort study enrolling patients with solid tumors who developed febrile neutropenia after chemotherapy. We collected demographic and tumor variables, the suspected focus of infection, and other variables necessary for calculating the MASCC score. The following biomarkers were measured: C-reactive protein (CRP), procalcitonin, and MR-proADM. The main outcome was the development of serious complications that were not present when febrile neutropenia was detected.ResultsA total of 173 episodes were studied. The median patient age was 70 years, and 60.7% were women. Serious complications developed in 55 patients (31.8%). The strongest predictor was MR-proADM, with an area under the receiver operating characteristic curve of 0.90 (95% CI, 0.85-0.95), significantly greater than the areas for procalcitonin (0.83; 95% CI, 0.76-0.90), CRP (0.79; 95% CI, 0.71-0.86), or the MASCC score (0.74, 95% CI, 0.65-0.82). Yield was not improved by combining MR-proADM with other biomarkers or the MASCC score.ConclusionsRisk stratification in cancer patients with febrile neutropenia is essential for decision-making in the ED. MR-proADM was the best predictor of serious complications in these patients, and combining it with any of the other variables did not improve prediction.

41 keywords...

hide…