基于二维经胸超声心动图与三维斑点追踪成像相关参数特征构建家族肥厚型心肌病突变基因携带者的列线图预测模型  

Early identification of familial hypertrophic cardiomyopathy mutation gene carriers by constructing Nomogram prediction model based on parametric characteristics of two-dimensional transthoracic echocardiography and three-dimensional speckle tracking imaging

作　　者：段奕全 梁青青 徐艳萍 叶晶晶 王芳 黄璇 王栎茗 纳丽莎 Duan Yiquan;Liang Qingqing;Xu Yanping;Ye Jingjing;Wang Fang;Huang Xuan;Wang Liming;Na Lisha(School of Clinical Medicine,Ningxia Medical University,Yinchuan 750004,China;Department of Functional Diagnosis,the Affiliated Hospital of Shaanxi University of Chinese Medicine,Xianyang 712000,China;Department of Cardiac Function Examination of Heart Centre,General Hospital of Ningxia Medical University,Yinchuan 750004,China)

机构地区：[1]宁夏医科大学临床医学院,银川750004 [2]陕西中医药大学附属医院功能诊断科,咸阳712000 [3]宁夏医科大学总医院心脏中心功能检查部超声心动图室,银川750004

出　　处：《中华超声影像学杂志》2023年第9期773-781,共9页Chinese Journal of Ultrasonography

基　　金：宁夏回族自治区重点研发计划项目(2021BEG03063)。

摘　　要：目的探索二维经胸超声心动图(2D-TTE)、三维斑点追踪成像(3D-STI)众多参数中可用于早期识别家族肥厚型心肌病(FHCM)突变基因携带者(G+P-)的超声心动图参数特征,并构建Nomogram(列线图)预测模型,以期为临床提供一种早期识别G+P-者的诊断方法。方法共纳入2017年11月至2022年8月就诊于宁夏医科大学总医院的FHCM家系15个,应用全外显子测序与Sanger测序技术进行基因检测,筛选出G+P-(54例)与G-P-(75例)共129例,应用分层随机抽样将受检者按照7∶3比例区分为训练集(90例)与测试集(39例)。应用Philips iE33超声诊断仪及TomTec脱机软件获取相关超声参数。使用Lasso回归及Logistic回归筛选超声心动图参数并获得可早期预测G+P-的独立危险因素,依此建立Nomogram预测模型。结果①Lasso-Logistic回归显示:整体纵向应变(GLS)(OR=1.739,95%CI=1.305~2.316)、左室流出道速度时间积分(LVOT-VTI)(OR=1.358,95%CI=1.072~1.722)可作为早期预测G+P-的独立危险因素。②将以上指标建立Nomogram预测模型,经过1000次Bootstrap自采样内部验证后,在训练集和测试集的C-指数分别为0.885(95%CI=0.816~0.954)、0.878(95%CI=0.764~0.992),具有良好的内部一致性。③校准曲线结果显示,Nomogram模型预测G+P-的风险与实际风险基本一致(训练集P=0.990、测试集P=0.961);临床决策曲线表明在不同阈概率下,使用该预测模型为患者提供临床决策,可以为患者带来获益。结论超声心动图参数GLS、LVOT-VTI可作为预测FHCM突变基因携带者的独立危险因素。Nomogram预测模型在识别FHCM患者家属是否携带突变基因中具有良好的区分度、拟合度、患者临床获益,可为FHCM突变基因携带者的超声心动图早期识别提供一种新的思路和评价方法。Objective To explore the characteristics of echocardiographic parameters among the many parameters of two-dimensional transthoracic echocardiography(2D-TTE)and three-dimensional speckle tracking imaging(3D-STI)that can be used for early identification of familial hypertrophic cardiomyopathy(FHCM)mutation gene carriers,and construct a Nomogram prediction model,in order to provide a diagnostic method for early identification of G+P-patients for clinical practice.Methods A total of 15 FHCM families admitted to the General Hospital of Ningxia Medical University from November 2017 to August 2022 were enrolled.Whole exome sequencing and Sanger sequencing technology were used for gene detection,among which 54 were G+P-and 75 were G-P-.Stratified random sampling was used to divide the subjects into training set(n=90)and test set(n=39)according to the ratio of 7∶3.Philips iE33 ultrasonic diagnostic instrument and TomTec offline software were used to obtain relevant ultrasonic parameters.Lasso regression and Logistic regression were used to screen echocardiographic parameters and obtain independent risk factors for early prediction of G+P-,based on which a Nomogram prediction model was established.Results①Lasso-Logistic regression showed that global longitudinal strain(GLS)(OR=1.739,95%CI=1.305-2.316)and left ventricular outflow trac velocity time integral(LVOT-VTI)(OR=1.358,95%CI=1.072-1.722)could be used as independent risk factors for early prediction of G+P-.②The Nomogram prediction model was established based on the above indicators.After 1000 internal verifications of Bootstrap self-sampling,the C-indices of the training set and the test set were 0.885(95%CI=0.816-0.954),0.878(95%CI=0.764-0.992),which had good internal consistency.③The results of the calibration curve showed that the risk of G+P-predicted by the Nomogram model was basically consistent with the actual risk(training set P=0.990,test set P=0.961);the clinical decision curve shows that under different threshold probabilities,using this prediction

关 键 词：超声心动描记术 心肌病 肥厚型 三维斑点追踪成像 列线图 

分 类 号：R540.45[医药卫生—心血管疾病] R542.2[医药卫生—内科学]