Early Risk Identification of Sepsis During Induction Chemotherapy in Newly Diagnosed AML: A Nomogram-Based Tool.
Patients with acute myeloid leukemia (AML) experience severe infections frequently during and after chemotherapy. The purpose of this study was to develop an early risk assessment tool to identify newly diagnosed AML patients at high risk of developing sepsis during the early phase of induction chemotherapy.
The records of 192 newly diagnosed AML patients from February 2017 to December 2023 at our institution were retrospectively reviewed. Kaplan-Meier survival analysis was performed to explore the prognosis difference between groups with or without sepsis. Mann-Whitney U test, Chi-square test, and multivariable logistic regression analysis were used to identify independent risk factors and establish a nomogram for early risk stratification of sepsis. The predictive accuracy of the nomogram was evaluated using the area under the curve (AUC), calibration curves, and concordance index (C-index). The nomogram is intended for application on days 3-5 of induction chemotherapy.
Multivariate logistic regression identified TP53 mutation, elevated aspartate aminotransferase (AST) and C-reactive protein (CRP) at diagnosis, increased procalcitonin (PCT), and higher urea levels post-induction as independent risk factors for sepsis progression. The bootstrap-corrected AUC was 0.892 (C-index 0.892). The bias-corrected calibration curve was close to the ideal line, demonstrating strong consistency between actual observations and predictions. Patients with sepsis experienced a poorer prognosis (p < 0.001) both at 1-year and 5-year survival rates. The incidence rates of pneumonia and invasive fungal infection were higher in the sepsis group.
We developed a nomogram for early identification of sepsis risk in newly diagnosed AML patients during their initial induction chemotherapy. This tool may assist clinicians in identifying patients requiring close monitoring and intensive supportive care.
The records of 192 newly diagnosed AML patients from February 2017 to December 2023 at our institution were retrospectively reviewed. Kaplan-Meier survival analysis was performed to explore the prognosis difference between groups with or without sepsis. Mann-Whitney U test, Chi-square test, and multivariable logistic regression analysis were used to identify independent risk factors and establish a nomogram for early risk stratification of sepsis. The predictive accuracy of the nomogram was evaluated using the area under the curve (AUC), calibration curves, and concordance index (C-index). The nomogram is intended for application on days 3-5 of induction chemotherapy.
Multivariate logistic regression identified TP53 mutation, elevated aspartate aminotransferase (AST) and C-reactive protein (CRP) at diagnosis, increased procalcitonin (PCT), and higher urea levels post-induction as independent risk factors for sepsis progression. The bootstrap-corrected AUC was 0.892 (C-index 0.892). The bias-corrected calibration curve was close to the ideal line, demonstrating strong consistency between actual observations and predictions. Patients with sepsis experienced a poorer prognosis (p < 0.001) both at 1-year and 5-year survival rates. The incidence rates of pneumonia and invasive fungal infection were higher in the sepsis group.
We developed a nomogram for early identification of sepsis risk in newly diagnosed AML patients during their initial induction chemotherapy. This tool may assist clinicians in identifying patients requiring close monitoring and intensive supportive care.