基于MRI的Node-RADS评分系统在膀胱癌淋巴结转移中的诊断价值  

Performance of MRI-based Node-RADS scoring system for the assessment of lymph node metastasis in bladder cancer patients

作　　者：柏荣杰 杨潇[1] 蔡令凯 王成浩 吴启开 庄俊涛 曹强[1] 李鹏超[1] 孙雪莹 吕强[1] BAI Rongjie;YANG Xiao;CAI Lingkai;WANG Chenghao;WU Qikai;ZHUANG Juntao;CAO Qiang;LI Pengchao;SUN Xueying;LYU Qiang(Department of Urology,First Affiliated Hospital of Nanjing Medical University,Jiangsu Province Hospital,Nanjing,210029,China;Department of Radiology,First Affiliated Hospitalof Nanjing Medical University,Jiangsu Province Hospital)

机构地区：[1]南京医科大学第一附属医院(江苏省人民医院)泌尿外科,南京210029 [2]南京医科大学第一附属医院(江苏省人民医院)放射科,南京210029

出　　处：《临床泌尿外科杂志》2024年第12期1050-1054,共5页Journal of Clinical Urology

基　　金：江苏省人民医院临床能力提升工程(No:JSPH-MA-2022-5);江苏省科教能力提升工程(No:ZDXK202219)。

摘　　要：目的:探讨基于MRI的Node-RADS评分系统在评估膀胱癌淋巴结转移(lymph node metastasis,LNM)中的适用性。方法:回顾性分析2019年1月—2024年2月于南京医科大学第一附属医院术前行MRI检查并行根治性膀胱切除术联合盆腔淋巴结清扫的223例患者的围术期临床病理及影像学资料,接受新辅助治疗的患者被排除在外,由2位医生对患者图像按照Node-RADS的评分标准进行独立评分。采用受试者工作特征(receiver operating characteristic,ROC)曲线评估Node-RADS评分系统对膀胱癌LNM的诊断效能并计算不同临界值的灵敏度、特异度、阳性预测值和阴性预测值(≥1,≥2,≥3,≥4,≥5)。结果:2位评分者的Node-RADS评分一致性良好(Kappa值=0.842,P<0.01)。Node-RADS评分评估LNM的ROC曲线下面积为0.841(95%置信区间:0.786~0.886)。根据ROC曲线分析,以Node-RADS≥4分为阈值,其约登指数为0.552,预测淋巴结转移的灵敏度为62.75%,特异度为92.44%,阳性预测值71.11%,阴性预测值89.33%。结论:以核磁共振为基础的Node-RADS在不同评分者之间具有较高的一致性,并且能有效地预测膀胱癌的LNM。Objective:To explore the applicability of MRI-based Node-RADS scoring system in the evaluation of lymph node metastasis(LNM)of bladder cancer(BCa).Methods:We retrospectively reviewed and collected clinical and MRI data of the BCa patients treated with radical cystectomy and pelvic lymph node dissection in First Affiliated Hospital of Nanjing Medical University from January 2019to February 2024.Patients receiving neoadjuvant chemotherapy were excluded.Two doctors reviewed and scored each patient separately according to Node-RADS.The sensitivity,specificity,positive predictive value and negative predictive value(≥1,≥2,≥3,≥4,≥5)of the Node-RADS scoring system for LNM were evaluated by receiver operating characteristic(ROC)curve.Results:Interobserver agreement of Node-RADS score between the two doctors was good(Kappa=0.842,P<0.01).The area under the ROC curve for the Node-RADS score assessing LNM was 0.841(95%confidence interval:0.786-0.886).According to the ROC analysis,the best cutoff-point was 4for the detection of LNM.The Youden index was 0.552,and the sensitivity,specificity,positive predictive value,and negative predictive value were 62.75%,92.44%,71.11%,and 89.33%,respectively.Conclusion:MRI-based Node-RADS exhibited a high agreement between different doctors,and can effectively predict the LNM of bladder cancer.

关 键 词：膀胱癌 淋巴结转移 Node-RADS评分 

分 类 号：R737.14[医药卫生—肿瘤]