CT影像组学列线图术前预测结直肠癌BRAF突变和预后  

Preoperative Prediction of BRAF Mutation and Prognosis in Colorectal Cancer Using CT Radiomics Nomogram

作　　者：李曼曼 徐国栋 徐高峰 周慧 冯峰[2] LI Manman;XU Guodong;XU Gaofeng;ZHOU Hui;FENG Feng(Department of Radiology,the Affiliated Tumor Hospital of Nantong University,Nantong 226001,China)

机构地区：[1]盐城市第一人民医院影像科,江苏盐城224000 [2]南通大学附属肿瘤医院影像科,江苏南通226001

出　　处：《中国医学影像学杂志》2024年第8期821-827,共7页Chinese Journal of Medical Imaging

基　　金：盐城市卫健委医学科研立项项目(YK2023056,YK2023058);南通市卫健委科研立项项目(MS2023052)。

摘　　要：目的利用CT影像组学列线图预测结直肠癌患者术前BRAF突变,并进行预后分层。资料与方法回顾性分析2017年1月—2020年6月盐城市第一人民医院经病理证实的333例结直肠癌,其中训练集234例,验证集99例。分割整个肿瘤的感兴趣区并提取影像组学特征。采用最大相关最小冗余算法及最小绝对收缩和选择算子算法筛选与BRAF突变密切相关的影像组学特征。结合影像组学特征和显著的临床参数构建列线图。采用受试者工作特征曲线下面积量化诊断效能。使用Kaplan-Meier分析绘制高风险组和低风险组的生存曲线。结果保留7个影像组学特征以构建影像组学模型。结合影像组学特征和3个显著的临床参数(年龄、肿瘤部位、癌胚抗原)构建列线图,其在训练集、验证集的曲线下面积分别为0.850、0.829。列线图预测高风险组(BRAF突变组)和低风险组(BRAF野生组)患者的总生存期存在显著差异(P<0.001),且高风险组的总生存期比低风险组短。结论CT影像组学列线图可术前准确预测结直肠癌的BRAF突变,并对预后进行分层。Purpose To predict preoperative BRAF mutation and perform prognostic stratification in colorectal cancer patients using CT radiomics nomogram.Materials and Methods A retrospective analysis was conducted on 333 pathologically confirmed colorectal cancer in the First People's Hosptial of Yancheng from January 2017 to June 2020,comprising a training set(234 cases)and a validation set(99 cases).The region of interest for the entire tumor was segmented,and radiomics features were extracted.The maximum-relevance minimumredundancy algorithm and the least absolute shrinkage and selection operator algorithm were used to screen for radiomics features closely associated with BRAF mutation.A nomogram was constructed by combining these radiomics features with clinically significant parameters.The diagnostic performance was quantified using the area under the receiver operating characteristic curve.Kaplan-Meier analysis was utilized to depict survival curves for the high-risk and low-risk groups.Results Seven radiomics features were retained to construct the radiomics model.The radiomics nomogram,integrating these radiomics features with three clinically significant parameters(age,tumor location,carcinoembryonic antigen),achieved area under the curve values of 0.850 in the training set and 0.829 in the validation set.Furthermore,there was a significant disparity in overall survival between the high-risk group(BRAF mutation)and low-risk group(wild-type BRAF)predicted by the nomogram(P<0.001),with the high-risk group exhibiting a markedly shorter overall survival compared to the low-risk group.Conclusion CT radiomics nomogram demonstrates the capability to accurately predict preoperative BRAF mutation in colorectal cancer patients and effectively stratify their prognosis.

关 键 词：结直肠肿瘤 体层摄影术 X线计算机 影像组学 BRAF突变 预后 

分 类 号：R445.3[医药卫生—影像医学与核医学] R735.37[医药卫生—诊断学]