非小细胞肺癌患者手术预后影响因素及预测模型构建  

Influencing factors and prediction model construction of prognosis in patients with non-small cell lung cancer undergoing surgery

作　　者：杨善岚 王楠 李洋洋 戴西望 陈新泰 刘勇[1] 吴磊[1] 江涵[2] YANG Shan-lan;WANG Nan;LI Yang-yang;DAI Xi-wang;CHEN Xin-tai;LIU Yong;WU Lei;JIANG Han(School of Public Health of Nanchang University,Jiangxi Province Key Laboratory of Preventive Medicine,Nanchang330006,China;Second Affiliated Hospital of Nanchang University,Nanchang330006,China)

机构地区：[1]南昌大学公共卫生学院,江西省预防医学重点实验室,江西南昌330006 [2]南昌大学第二附属医院胸外科,江西南昌330006

出　　处：《中华肿瘤防治杂志》2022年第23期1676-1684,共9页Chinese Journal of Cancer Prevention and Treatment

基　　金：国家自然科学基金(81960661,81560550);江西省自然科学基金(20202ACBL206016);南昌大学大学生创新创业训练计划(S202110403010)。

摘　　要：目的分析手术治疗的非小细胞肺癌(NSCLC)患者预后情况及多元影响因素,并有针对性构建预测模型,为临床工作提供参考。方法回顾性收集2015-01-01-2018-12-31南昌大学第二附属医院首次进行手术切除治疗的674例NSCLC患者信息。其中,男475例,女199例;年龄37~87岁,平均年龄(61.60±8.51)岁。患者均随访至术后3年。采用Cox模型对术后总生存期(OS)和无进展生存期(PFS)进行分析,通过列线图实现结果可视化,并通过受试者工作特征曲线下面积(AUC)、C指数等指标判断模型性能。结果术后1、2和3年的OS率分别为95.08%、87.23%和80.67%,PFS率分别为84.32%、72.43%和64.71%。单因素分析筛选出17个影响OS和PFS的指标,术后OS受Ki-67、用力肺活量(FVC)、系统免疫炎症指数(SII)等7个指标影响,术后PFS受CA199和一秒用力呼气容积(FEV1)/FVC影响,均P<0.05。基于Cox的OS预测模型纳入了确诊年龄、T分期、N分期、术前预后营养指数(PNI)水平等4个变量,PFS预测模型还纳入术中清除淋巴结数。OS和PFS预测模型的C指数分别为0.715和0.711,AUC分别为0.715和0.685;绘制的2和3年校准曲线显示了列线图的良好预测性能。结论NSCLC手术患者预后受年龄、T分期、N分期、术前PNI水平以及术中清除淋巴结数的影响,构建的术后OS和PFS预测模型综合考虑了以上多元因素,有较高的预测价值,可帮助临床医生制定NSCLC患者治疗和随访计划。Objective To explore the prognosis and multiple influencing factors of patients with non-small cell lung cancer(NSCLC)undergoing surgical treatment,and construct a prediction model in a targeted manner to provide a scientific basis for clinical work.Methods The information of 674patients with NSCLC who underwent surgical resection for the first time on January 1,2015to December 31,2018at the Second Affiliated Hospital of Nanchang University was retrospectively collected.Among them,475were males and 199were females;age 37-87years old,average age(61.60±8.51)years.All patients were followed up for 3years after surgery.The Cox model was used to analyze the total postoperative survival(OS)and progression-free survival(PFS),visualize the results through the nomograms,and judge the model performance by area under receiver operating characteristic curve(AUC),concordance-index(C-index)and other indicators.Results A total of 674people were followed up in this study,and the overall survival rates at 1,2and 3years after surgery were 95.08%,87.23%and 80.67%,respectively,and PFS rates were 84.32%,72.43%,and 64.71%,respectively.Univariate analysis screened out of 17indicators both affecting OS and PFS.Postoperative OS was affected by 7indicators including Ki-67,forced vital capacity(FVC),systemic immune-inflammation index(SII),and postoperative PFS which was affected by CA199and forced expiratory volume in one second(FEV1)/FVC(all P<0.05).The Cox-based OS prediction model included four variables,including age at diagnosis,T stage,N stage,and preoperative prognostic nutritional index(PNI)level,and the PFS prediction model also included the LNs dissected during surgery.The C-index of OS and PFS prediction models was 0.715and 0.711,respectively,and the area under curve(AUC)was 0.715and 0.685,respectively;the 2-year and 3-year calibration curves plotted showed the good prediction performance of the nomogram.Conclusion The prognosis of patients with NSCLC surgery is affected by age at diagnosis,T stage,N stage,preoperative PNI leve

关 键 词：非小细胞肺癌 预后 COX模型 列线图 总生存期 无进展生存期 

分 类 号：R734.2[医药卫生—肿瘤]