摘要
目的探讨前列腺特异抗原(PSA)灰区前列腺癌(PCa)患者预后的独立预测因素,并为其建立个体化预测的列线图模型。方法回顾性检索2010—2016年PCa患者的临床资料,随机将其分为训练集(2/3)和验证集(1/3)。在多因素Cox回归分析的基础上,建立了独立危险因素的列线图。分别采用C指数(C-index)、ROC曲线及校准曲线来验证列线图预测准确性。结果共收集PCa患者82537例,其中训练集患者57777例,验证集24760例。多因素Cox分析显示,年龄、种族、婚姻状况、手术方式、是否放疗、是否化疗、病理分级及T分期是总生存期(OS)的独立危险因素。训练集OS的C指数(C-index)为0.741(95%CI:0.731~0.751),验证集为0.743(95%CI:0.725~0.761)。表明预测OS的列线图在训练集和验证集上都显示出较好的识别力。此外,ROC曲线及校准曲线验证了预测的生存概率与实际生存概率之间良好的一致性,并且其区分度优于传统病理分级系统。结论本研究基于SEER数据库建立了国内外首个可以个体化预测PSA灰区PCa患者预后的列线图模型;且经内部及外部验证,其预测性能良好。列线图模型的建立将有助于辅助临床医生更加精准地预测PCa患者预后结局,为PCa患者个体化管理提供依据。
Objective To explore the independent predictors of the prognosis of prostate-specific antigen(PSA)gray zone prostate cancer(PCa)patients,and to establish a nomogram model for individualized prediction.Methods The clinical data of PCa patients from 2010 to 2016 were retrospectively searched,and they were randomly divided into a training set(2/3)and a validation set(1/3).On the basis of multivariate Cox regression analysis,a nomogram of independent risk factors was established.The C-index,ROC curve,and calibration curve were used to verify the accuracy of the nomogram prediction.Results A total of 82537 PCa patients were collected,including 57777 patients in the training set and 24760 patients in the validation set.Multivariate Cox analysis showed that age,race,marital status,surgical method,radiotherapy,chemotherapy,pathological grade,and T stage were independent risk factors for overall survival(OS).The C-index of the training set OS was 0.741(95%CI:0.731-0.751),and the validation set was 0.743(95%CI:0.725-0.761).The nomogram of predicting OS showed good recognition ability on both the training set and the validation set.In addition,the ROC curve and calibration curve verified the good agreement between the predicted survival probability and the actual survival probability,and its discrimination was better than traditional pathological grading systems.Conclusion Based on the SEER database,this study established the first domestic and foreign nomogram model that could individually predict the prognosis of PCa patients in the PSA gray zone.Through internal and external verification,its prediction performance is good.The establishment of the nomogram model will help clinicians to more accurately predict the prognosis of PCa patients and provide a basis for individualized management of PCa patients.
作者
杭天昆
李通义
石明凯
陈建舟
马志方
HANG Tiankun;LI Tongyi;SHI Mingkai;CHEN Jianzhou;MA Zhifang(First Clinical Medical College,Shanxi Medical University,Taiyuan,030001,China;Department of Urology,First Hospital of Shanxi Medical University)
出处
《临床泌尿外科杂志》
CAS
2022年第8期606-614,共9页
Journal of Clinical Urology
