胆囊鳞状细胞癌预后模型的构建与验证  

Construction and validation of the prognosis model for gallbladdersquamous cell carcinoma

作　　者：黄坤 黄正红 赵攀[1] 赵平武[1] 何运胜 白斗[3] HUANG Kun;HUANG Zhenghong;ZHAO Pan;ZHAO Pingwu;HE Yunsheng;BAI Dou(Department of General Surgery,Mian Yang Traditional Chinese Medicine Hospital,Mianyang,Sichuan 621000,China;College of Medical Technology,Chengdu University of Traditional Chinese Medicine,Chengdu 611137,China;Department of General Surgery,Division of Vascular Surgery,Mian Yang Central Hospital,Mianyang,Sichuan 621000,China)

机构地区：[1]四川省绵阳市中医医院普通外科,四川绵阳621000 [2]成都中医药大学医学技术学院,四川成都611137 [3]四川省绵阳市中心医院血管外科,四川绵阳621000

出　　处：《中国普通外科杂志》2023年第8期1187-1198,共12页China Journal of General Surgery

摘　　要：背景与目的:胆囊鳞状细胞癌(GSCC)是胆囊癌中一种罕见的病理学类型,占胆囊癌的1%~4%。该类型肿瘤预后差,目前关于GSCC的文献报道主要是个案报道和小样本系列病例报道,由于缺乏大样本高质量的临床研究证据,目前临床上尚无针对GSCC的治疗指南、共识和个体化的预后评价工具。因此,本研究通过SEER数据库中的大样本数据构建GSCC患者预后列线图,旨在精准化、个体化评价GSCC患者的预后,为临床决策制定提供参考。方法:提取SEER数据库中2000—2019年期间经病理确诊的GSCC患者的临床资料,按照7∶3的比例,将数据随机划分为训练集和验证集,在训练集中,分别采用多变量Cox比例风险模型和LASSO回归筛选影响GSCC患者预后的独立因素,利用这些因素,构建用于预测GSCC患者在3个月和6个月的肿瘤特异性生存期(CSS)和总生存期(OS)的列线图模型。随后,在训练集中,利用一致性指数(C指数)、ROC曲线和校准曲线,分别在训练集和验证集,对模型进行内部和外部验证,以评估模型的准确度和预测能力。结果:本研究共纳入257例患者,其中训练集179例,验证集78例。在训练集和验证集中,患者的中位随访时间分别为3(1~7)个月和4(2~8)个月。两组之间基线资料均衡可比。多变量Cox比例风险模型分析显示,年龄、SEER分期、手术和化疗是GSCC患者OS和CSS的独立影响因素(均P<0.05)。LASSO回归分析显示,年龄、SEER分期、放疗、手术和化疗与GSCC患者的OS相关;年龄、SEER分期、手术和化疗与GSCC患者的CSS相关。基于这些独立预后影响因素,构建了用于预测GSCC患者在3、6个月的OS和CSS的列线图。对模型的验证结果表明,训练集和验证集中,OS的C指数分别为0.739(95%CI=0.700~0.780)和0.729(95%CI=0.660~0.800);CSS的C指数分别为0.750(95%CI=0.710~0.790)和0.741(95%CI=0.670~0.810)。ROC曲线分析显示,曲线在训练集和验证集的AUC值均>0.8;校准曲�Background and Aims:Gallbladder squamous cell carcinoma(GSCC)is a rare histopathological subtype of gallbladder cancer,accounting for 1%to 4%of cases.This tumor type is associated with poor prognosis.Currently,the literature on GSCC mainly consists of case reports and small-sample case series.Due to the lack of large-sample high-quality clinical research evidence,there are no established treatment guidelines,consensus,or personalized prognostic assessment tools for GSCC.Therefore,this study aimed to construct prognostic nomograms for GSCC patients using large-scale real-world data from the SEER database to provide precise and individualized prognosis assessment for GSCC patients,offering valuable references for clinical decision-making.Methods:Clinical data of GSCC patients pathologically diagnosed between 2000 and 2019 were extracted from the SEER database.The data were randomly divided into training and validation sets in a 7∶3 ratio.In the training set,a multivariate Cox proportional hazards model and LASSO regression were used to identify independent prognostic factors for the survival of GSCC patients.These factors constructed nomogram models to predict tumor-specific survival(CSS)and overall survival(OS)at 3 and 6 months for GSCC patients.Subsequently,the models were internally and externally validated in training and validation sets using the concordance index(C-index),ROC,and calibration curves to assess their accuracy and predictive capacity.Results:A total of 257 patients were included in this study,179 in the training and 78 in the validation set.The median follow-up times were 3(1-7)months in the training set and 4(2-8)months in the validation set.Baseline characteristics were comparable between the two groups.The multivariate Cox proportional hazards model analysis revealed that age,SEER stage,surgery,and chemotherapy were independent factors for OS and CSS in GSCC patients(all P<0.05).LASSO regression analysis indicated that age,SEER stage,radiotherapy,surgery,and chemotherapy were associated with

关 键 词：胆囊肿瘤 癌 鳞状细胞 预后 SEER规划 列线图 

分 类 号：R735.8[医药卫生—肿瘤]