阴茎鳞状细胞癌预后列线图的开发与验证  

Development and validation of prognostic nomogram for patients with squamous cell carcinoma of the penis

作　　者：林东旭 张梦阳 罗长城 陈忠[1] LIN Dongxu;ZHANG Mengyang;LUO Changcheng;CHEN Zhong(Department of Urology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China)

机构地区：[1]华中科技大学同济医学院附属同济医院泌尿外科,武汉430030

出　　处：《现代泌尿生殖肿瘤杂志》2021年第2期85-92,102,共9页Journal of Contemporary Urologic and Reproductive Oncology

摘　　要：目的分析阴茎鳞状细胞癌(SCCP)死亡的独立预测因素,构建个性化的疾病预测模型。方法从美国国立癌症研究所监测、流行病学和结局(SEER)数据库中检索SCCP患者的临床数据,按照7∶3的比例随机分为试验集和验证集。对于试验集资料,构建单因素和多因素Cox回归模型分析多个变量(包括年龄、种族、肿瘤分化、分期、TNM分期、肿瘤直径、手术、放疗、化疗)对患者预后的影响,并绘制列线图预测疾病总生存率(OS)和肿瘤特异性生存率(CSS)。根据一致性指数和校准曲线评估模型的预测准确度,根据受试者工作特征(ROC)曲线、重分类改善指数和综合判别改善指数评价模型相对于TNM分期的预测能力是否有改善,采用决策曲线评估列线图模型在临床应用中的价值与获益,Kaplan-Meier法绘制生存曲线评价变量对预后的影响。结果患者中位随访时间为60(29~101)个月,在随访结束时,有928例(40.43%)死亡,其中512例(22.31%)因肿瘤而死亡。患者的年龄、肿瘤分化、分期、肿瘤直径、N分期与OS和CSS相关,是SCCP患者预后的独立影响因素。综合以上变量构建的列线图的一致性指数较高(试验集中OS为0.731、CSS为0.776;验证集中OS为0.722、CSS为0.756),且校准曲线显示模型与实际生存情况相吻合,提示该模型可以准确预测患者的预后情况。此外,ROC曲线显示模型相较于TNM分期系统有更好的敏感性和特异性,重分类改善指数和综合判别改善指数值进一步说明模型的预测能力较高,决策曲线显示列线图可以使得患者有较好的临床获益。生存分析显示高龄(>80岁)、分化差(低分化)、分期晚(远处转移)、直径大(>2.5 cm)及淋巴转移(N3)的患者预后更差。结论本研究构建的列线图模型能够较准确地预测SCCP患者的预后情况,并较经典TNM分期的预测能力有所改善,具有良好的临床应用价值。Objective According to the Surveillance,Epidemiology,and End Results(SEER)database,the independent prognostic factors of squamous cell carcinoma of the penile(SCCP)were identified to develop the prognostic models.Methods Patients information from SEER database were isolated for further analysis and randomly allocated to the training and validation cohorts at the ratio of 7 to 3.By univariable and multivariable cox analyses,nine variables,including age,race,grade,AJCC stage,TNM stage,tumor size,surgery,radiation,and chemotherapy contributed to the development of nomogram for predicting the overall survival(OS)and cancer-specific survival(CSS)of SCCP patients.C-indices and calibration curves were applied to determine the predictive accuracy in both training and validation cohorts.Receiver operating characteristic(ROC)curve,net reclassification improvement(NRI)and integrated discrimination improvement(IDI)were applied to compare the predicted performance between the new nomograms and the tumor,lymph node,and metastasis(TNM)system.Decision curve analysis(DCA)method were applied to evaluate the clinical benefit of nomogram relative to TNM system.Kaplan-Meier survival curves were plotted to evaluate the effect of these variable on patient prognosis.Results The median follow-up time was 60(29-101)months.At the end of the follow-up,928 cases(40.43%)died,while 512 cases(22.31%)died of tumors.After cox regression analysis,two nomograms for OS and CSS prediction were established by independent prognostic variables age,grade,AJCC stage,tumor size and N stage.The models showed great accuracy with preferable C-indices(0.731 and 0.776 for OS and CSS in the training cohort,and 0.722 and 0.756 in the validation cohort),and calibration curves showed great consistency between predicted survival and actual survival.Moreover,the ROC curve,NRI,IDI,and DCA showed a better predictive performance and clinical usefulness of the nomograms than the TNM system.Age>80,gradeⅢ(poor differentiated),distant metastasis,size>2.5 cm,and N stage we

关 键 词：阴茎鳞状细胞癌 列线图 总生存率 肿瘤特异性生存率 SEER数据库 

分 类 号：R73[医药卫生—肿瘤]