Meaningful nomograms based on systemic immune inflammation index predicted survival in metastatic pancreatic cancer patients receiving chemotherapy

作者：Yanan Sun, Jiahe Hu, Rongfang Wang, Rongfang Wang, Xinlian Du, Xiaoling Zhang, E Jiaoting, Shaoyue Zheng, Yuxin Zhou, Ruishu Mou, Xuedong Li, Han‐Bo Zhang, Ying Xu, Yuan Liao, Wenjie Jiang, Lijia Liu, Ruitao Wang, Ruitao Wang, Jiuxin Zhu, Rui Xie · 发表于：Cancer Medicine · 年份：2024 · DOI：10.1002/cam4.7453 · 被引用次数：7 · 研究领域：Inflammatory Biomarkers in Disease Prognosis、Cancer Immunotherapy and Biomarkers、Ferroptosis and cancer prognosis

OBJECTIVE: The purpose of the study is to construct meaningful nomogram models according to the independent prognostic factor for metastatic pancreatic cancer receiving chemotherapy. METHODS: This study is retrospective and consecutively included 143 patients from January 2013 to June 2021. The receiver operating characteristic (ROC) curve with the area under the curve (AUC) is utilized to determine the optimal cut-off value. The Kaplan-Meier survival analysis, univariate and multivariable Cox regression analysis are exploited to identify the correlation of inflammatory biomarkers and clinicopathological features with survival. R software are run to construct nomograms based on independent risk factors to visualize survival. Nomogram model is examined using calibration curve and decision curve analysis (DCA). RESULTS: The best cut-off values of 966.71, 0.257, and 2.54 for the systemic immunological inflammation index (SII), monocyte-to-lymphocyte ratio (MLR), and neutrophil-to-lymphocyte ratio (NLR) were obtained by ROC analysis. Cox proportional-hazards model revealed that baseline SII, history of drinking and metastasis sites were independent prognostic indices for survival. We established prognostic nomograms for primary endpoints of this study. The nomograms' predictive potential and clinical efficacy have been evaluated by calibration curves and DCA. CONCLUSION: We constructed nomograms based on independent prognostic factors, these models have promising applications in ...