食管癌根治性放化疗患者血小板减少的预测模型构建  

Construction of a predictive model for thrombocytopenia in patients with esophageal cancer undergoing radical radiotherapy and chemotherapy

作　　者：程欣宇 周海文 汪晨宇 张耀文[1] 纪媛媛[1] Cheng Xinyu;Zhou Haiwen;Wang Chenyu;Zhang Yaowen;Ji Yuanyuan(Department of Radiotherapy,Anyang Tumor Hospital,Anyang,Henan 455000,China)

机构地区：[1]安阳市肿瘤医院放疗科,河南安阳455000

出　　处：《齐齐哈尔医学院学报》2024年第14期1305-1310,共6页Journal of Qiqihar Medical University

基　　金：2020年度河南省医学科技攻关计划(联合共建)项目(LHGJ20200805);2019年度河南省医学科技攻关计划(联合共建)项目(LHGJ20191288)。

摘　　要：目的探讨食管癌根治性放化疗患者血小板减少的相关因素,并构建预测模型。方法回顾性分析2021年7月—2023年9月本院收治的根治性放化疗的123例食管癌患者的临床病历资料,将43例(34.96%)发生血小板减少患者的病历资料归入发生组,其余80例(65.04%)未发生血小板减少患者病历资料归入未发生组。比较两组患者基线资料、实验室检查资料等,并采用Logistic回归分析明确食管癌根治性放化疗患者血小板减少的影响因素,根据Logistic回归系数构建预测模型,通过Bootstrap内部验证法、标准曲线、受试者工作特征曲线(ROC)验证预测模型的预测效能。结果发生组年龄、肿瘤直径大于未发生组;血红蛋白(HGB)、白蛋白(ALB)水平低于未发生组,糖尿病占比高于未发生组,差异有统计学意义(P<0.05);经Logistic回归分析显示,年龄、糖尿病、肿瘤直径、HGB、ALB是食管癌根治性放化疗患者血小板减少发生的影响因素(P<0.05)。根据Logistic回归分析系数构建食管癌根治性放化疗患者血小板减少发生的预测模型,通过Bootstrap内部验证法验证模型区分度,显示C-index=0.875,具有良好区分度;绘制标准曲线,发现校准曲线和Y-X直线相近,提示预测模型准确度良好。绘制ROC曲线对预测模型进行内部验证,预测模型评估食管癌根治性放化疗患者血小板减少的AUC为0.875,AUC的95%CI为0.812~0.937、P<0.001、特异度为0.838,敏感度为0.791,约登指数为0.629,具有良好预测效能。结论年龄、糖尿病、肿瘤直径、HGB、ALB是食管癌根治性放化疗患者血小板减少发生的影响因素,根据上述因素构建的预测模型具有较好预测效能。Objective To explore the related factors of thrombocytopenia in patients with esophageal cancer undergoing radical radiotherapy and chemotherapy,and to construct a predictive model.Methods A retrospective analysis on the clinical records of 123 patients suffering from esophageal cancer who underwent radical radiotherapy and chemotherapy at the hospital from July,2021 to September,2023 was conducted.The medical records of 43 patients(34.96%)who experienced thrombocytopenia were classified as the occurrence group,while the medical records of the remaining 80 patients(65.04%)who did not experience thrombocytopenia were classified as the non occurrence group.The baseline data and laboratory examination data of the two groups were compared.Logistic regression analysis was used to identify the influencing factors of thrombocytopenia in patients with radical radiotherapy and chemotherapy for esophageal cancer.The prediction model was constructed according to the Logistic regression coefficient,and the prediction efficiency of the prediction model was verified by Bootstrap internal validation method,standard curve and receiver operating characteristic curve(ROC).Results The age and tumor diameter of the occurrence group were larger than those of the non occurrence group;The levels of hemoglobin(HGB)and albumin(ALB)were lower than those in the non occurrence group,and the proportion of diabetes was higher than that in the non occurrence group(P<0.05).Logistic regression analysis showed that age,diabetes,tumor diameter,HGB and ALB were the influencing factors of thrombocytopenia in patients with esophageal cancer undergoing radical radiotherapy and chemotherapy(P<0.05).A predictive model for thrombocytopenia in esophageal cancer patients undergoing radical radiotherapy and chemotherapy was constructed based on Logistic regression analysis coefficients.The model discrimination was validated using Bootstrap internal validation method,and C-index=0.875,indicating good discrimination.Standard curve was draw and found that the

关 键 词：食管癌 根治性放化疗 血小板减少症 影响因素 预测模型 

分 类 号：R735.1[医药卫生—肿瘤]