术前炎性指标构建的列线图模型预测结直肠癌患者术后生存分析

Establishment of a nomogram model for predicting postoperative survival of patients with colorectal cancer based on preoperative inflammatory indexes

目的探讨基于术前炎性指标构建的列线图模型预测结直肠癌患者术后生存的价值。方法采用队列研究设计,选取2011年1月至2014年6月空军第986医院行结直肠癌根治术的233例结直肠癌患者,根据5年随访结果,将患者分成生存组(99例)和死亡组(134例)。比较两组患者术前1 d炎性指标水平,单因素和Cox回归分析结直肠癌患者术后5年生存的影响因素,应用R软件建立列线图术后存活预测模型。结果两组患者术前淋巴细胞计数、中性粒细胞计数、血小板计数、C反应蛋白、血小板/淋巴细胞比值(PLR)、中性粒细胞/淋巴细胞(NLR)和C反应蛋白/白蛋白比值(CAR)等指标比较,差异有统计学意义(P<0.05),而白细胞计数和白蛋白比较,差异无统计学意义;肿瘤大小(OR=1.379,95%CI:1.094~1.737)、浸润深度(OR=2.020,95%CI:1.126~3.622)、NLR(OR=1.496,95%CI:1.009~2.219)、PLR(OR=1.927,95%CI:1.060~3.504)和CAR(OR=2.326,95%CI:1.479~3.657)是结直肠癌患者术后生存的独立影响因素(P<0.05)。列线图预测术后生存模型的C-index为0.831(95%CI:0.781~0.911),校准预测曲线和理想曲线拟合良好。结论术前NLR、PLR和CAR与结直肠癌术后生存呈负相关,且列线图具有预测结直肠癌患者术后生存情况的潜在价值。

Objective To establish a nomogram model for predicting postoperative survival of patients with colorectal cancer based on preoperative inflammatory indexes.Methods A cohort study design was used,233 patients with colorectal cancer who met the criteria were selected from January 2011 to June 2014 in Air Force 986th Hospital.All patients underwent surgical resection of colorectal cancer and were followed up for 5 years.They were divided into survival group(99 cases)and death group(134 cases).The levels of preoperative inflammatory indexes were compared between the two groups.Univariate and Cox regression were used to analyze the influencing factors of postoperative survival of patients,and R software was used to establish a prediction model of postoperative survival.Results There were significant differences in lymphocyte count,neutrophil count,platelet count,C-reactive protein,platelet/lymphocyte ratio(PLR),neutrophil/lymphocyte ratio(NLR)and C-reactive protein/albumin ratio(CAR)between the two groups(P<0.05),but no differences in white blood cell count and albumin.Tumor size(OR=1.379,95%CI:1.094-1.737),invasion depth(OR=2.020,95%CI:1.126-3.622),NLR(OR=1.496,95%CI:1.009-2.219),PLR(OR=1.927,95%CI:1.060-3.504)and CAR(OR=2.326,95%CI:1.479-3.657)were independent factors affecting postoperative survival.The C-index of the postoperative survival model predicted by the line chart was 0.831(95%CI:0.781-0.911),and the calibration prediction curve fitted well with the ideal curve.Conclusion Preoperative NLR,PLR,and CAR are negatively correlated with postoperative survival of colorectal cancer,and nomogram model can predict the postoperative survival of patients with colorectal cancer effectively.