肺腺癌纵隔淋巴结转移风险分层的预测模型  

Risk stratification of mediastinal lymph node metastasis in patients with lung adenocarcinoma

作　　者：刘天阳 闫思颖 李强明 叶飞 赵松[1] 李向楠[1] LIU Tianyang;YAN Siying;LI Qiangming;YE Fei;ZHAO Song;LI Xiangnan(Department of Thoracic Surgery,the First Affiliated Hospital of Zhengzhou University,Zhengzhou 450052,China)

机构地区：[1]郑州大学第一附属医院胸外科,郑州450052

出　　处：《中华实用诊断与治疗杂志》2020年第12期1259-1262,共4页Journal of Chinese Practical Diagnosis and Therapy

基　　金：郑州大学学科建设重点专项(前言交叉专项XKZDQY202006);河南省科技厅科技攻关项目(172102310047)。

摘　　要：目的探讨肺腺癌纵隔淋巴结转移的影响因素,并建立肺腺癌纵隔淋巴结转移的预测模型。方法245例肺腺癌患者,根据是否发生纵隔淋巴结转移分为转移组66例和未转移组179例,比较2组临床资料,采用多因素logistic回归分析肺腺癌患者发生纵隔淋巴结转移的影响因素。以肺腺癌患者发生纵隔淋巴结转移的影响因素构建列线图预测模型,绘制其校正曲线和决策曲线,预测模型的预测性能采用c-指数评估。根据列线图模型对肺腺癌纵隔淋巴结转移预测的总分进一步采用"递归分割分析方法"建立决策树模型。结果转移组年龄<60岁、肿瘤直径>3 cm、肿瘤组织学亚型乳头型和实体型、肿瘤低分化比率高于未转移组(P<0.05);年龄、肿瘤直径和肿瘤分化程度是肺腺癌患者发生纵隔淋巴结转移的影响因素(OR=0.366,95%CI:0.189~0.706,P=0.003;OR=1.416,95%CI:1.059~1.893,P=0.019;OR=5.151,95%CI:2.720~9.755,P<0.001)。以年龄、肿瘤直径和肿瘤分化程度构建肺腺癌纵隔淋巴结转移列线图预测模型,c-指数为0.819。校正曲线显示列线图预测模型预测肺腺癌纵隔淋巴结转移概率与实际淋巴结转移概率吻合度较高。决策曲线显示肺腺癌纵隔淋巴结转移概率阈值为0.001~0.713时,应用该列线图预测模型有较好收益。递归分割分析依据转移概率将纵隔淋巴结转移概率分为4个风险亚组,低危组为肿瘤高/中分化,中低危组为肿瘤低分化+肿瘤直径≤4.0 cm+年龄≥60岁,中高危组为肿瘤低分化+肿瘤直径≤4.0 cm+年龄<60岁,高危组为肿瘤低分化+肿瘤直径>4.0 cm;4个风险亚组纵隔淋巴结转移概率分别为15%、39%、58%和77%。结论年龄、肿瘤直径和肿瘤分化程度是肺腺癌纵隔淋巴结转移的的影响因素,建立肺腺癌纵隔淋巴结转移的预测模型可为临床行纵隔淋巴结清扫提供理论依据。Objective To investigate the influencing factors of mediastinal lymph node metastasis in patients with lung adenocarcinoma,and prepare a prediction model of mediastinal lymph node metastasis.Methods A total of 245 patients with lung adenocarcinoma were divided into metastasis group(n=66)and non-metastasis group(n=179)according to the occurrence of mediastinal lymph node metastasis.The clinical data of two groups were compared,and the influencing factors of mediastinal lymph node metastasis in patients with lung adenocarcinoma were analyzed by multivariate logistic regression.The nomogram prediction model was constructed based on the influencing factors of mediastinal lymph node metastasis in patients with lung adenocarcinoma,and calibration curve and decision curve were drawn to predict the predictable performance of the model and to evaluate by concordance index.According to the total score of the prediction of lung adenocarcinoma mediastinal lymph node metastasis based on the nomogram model,the decision-making tree model was further established by using the recursive partition analysis method.Results The percentages of patients aged<60 years,with tumor diameter>3 cm,papillary and solid histological subtypes,and poor differentiation were higher in metastasis group than those in non-metastasis(P<0.05).The age(OR=0.366,95%CI:0.189-0.706,P=0.003),the diameter of tumor(OR=1.416,95%CI:1.059-1.893,P=0.019)and differentiation degree(OR=5.151,95%CI:2.720-9.755,P=0.001)were the influencing risk factors of mediastinal lymph node metastasis in patients with lung adenocarcinoma.Based on the age,tumor diameter and degree of tumor differentiation,a prediction model of mediastinal lymph node metastasis of lung adenocarcinoma was established,and the concordance index of the nomogram was 0.819.The calibration curve showed that the predicted probability of mediastinal lymph node metastasis of lung adenocarcinoma was in good agreement with the actual probability of lymph node metastasis.When the decision curve showed that the prob

关 键 词：肺腺癌 纵隔淋巴结转移 影响因素 递归分割 

分 类 号：R734.2[医药卫生—肿瘤]