摘要
目的:基于超声检查、肿瘤细胞因子构建儿童肾母细胞瘤的Nomogram诊断模型,以提高对肾母细胞瘤的临床诊断效能。方法:选取2019年1月至2023年1月我院收治的82例肾母细胞瘤患儿为研究组,另选取同期82例囊肿性肾病患儿为对照组,均经手术病理证实。术前均行超声检查,比较两组超声特征、临床资料。通过Logistic分析肾母细胞瘤发生的影响因素,根据影响因素构建肾母细胞瘤的Nomogram诊断模型。采用受试者工作特征(ROC)曲线、校准曲线进行Nomogram诊断模型验证。结果:两组肿块性质、内部回声、邻近结构侵犯及转移、伴下腔静脉瘤栓、伴同侧肾静脉血栓、动脉期峰值流速、血流阻力指数、肾体积局部增大、基质金属蛋白酶-9(MMP-9)、血管内皮生长因子(VEGF)比较,差异有统计学意义(P<0.05);LASSO回归分析显示,可使模型性能优良且影响因素最少的最佳惩罚系数λ下诊断肾母细胞瘤的变量数为6个:即肿块性质、内部回声、邻近结构侵犯及转移、伴同侧肾静脉血栓、VEGF、MMP-9;Logistic回归分析显示,肿块性质、内部回声、邻近结构侵犯及转移、伴同侧肾静脉血栓、VEGF、MMP-9是肾母细胞瘤的相关危险因素(P<0.05);运用R语言绘制肾母细胞瘤诊断模型显示,其C-index为0.939,诊断肾母细胞瘤的AUC为0.939。结论:肿块性质、内部回声、邻近结构侵犯及转移、伴同侧肾静脉血栓、VEGF、MMP-9与肾母细胞瘤发生有关,基于超声检查和肿瘤细胞因子所构建的诊断模型具有较高区分度、校准度以及诊断效能,可为肾母细胞瘤临床诊断提供参考。
Objective:To construct a Nomogram diagnostic model for nephroblastoma in children based on ultrasonography and tumor cytokines,so as to improve the clinical diagnostic efficiency of nephroblastoma.Methods:A total of 82 children with nephroblastoma admitted to our hospital from January 2019 to January 2023 were selected as the study group,and another 82 children with cystic nephropathy during the same period were selected as the control group,which were confirmed by surgery pathological results.The ultrasonography was performed before surgery.The ultrasonic characteristics and clinical data of the two groups were compared,and the influencing factors of the occurrence of nephroblastoma were analyzed by Logistic analysis.A Nomogram diagnostic model for nephroblastoma based on the influencing factors was constructed,and the Nomogram diagnostic model was verified by receiver operating characteristic(ROC) curve and calibration curve.Results:There were significant differences in mass nature,internal echo,adjacent structures invasion and metastasis,inferior vena cava thrombosis,accompanying ipsilateral renal vein thrombosis,arterial peak velocity,blood flow resistance index,local renal volume increase,vascular endothelial growth factor(VEGF) and matrix metalloproteinase-9(MMP-9) between the two groups(P<0.05).LASSO regression analysis showed that there were 6 variables in the diagnosis of nephroblastoma under the optimal penalty coefficient λ,which could make the model perform well and had the least influence factors:mass nature,internal echo,adjacent structures invasion and metastasis,accompanying ipsilateral renal vein thrombosis,VEGF and MMP-9.Logistic regression analysis showed that mass nature,internal echo,adjacent structures invasion and metastasis,accompanying ipsilateral renal vein thrombosis,VEGF and MMP-9 were the risk factors for nephroblastoma(P<0.05).R-language was used to draw the diagnostic model of nephroblastoma,which showed that the C-index was 0.939 and the AUC for diagnosis of nephroblastoma was 0.9
作者
张扬
梁勇
聂智睿
ZHANG Yang;LIANG Yong;NIE Zhirui(Department of Ultrasound,Linfen People's Hospital,Linfen 041000,China)
出处
《东南大学学报（医学版）》
CAS
2024年第3期387-394,共8页
Journal of Southeast University(Medical Science Edition)
基金
山西省应用基础研究计划(201811D110241)。
