【作者】 李倩；

【导师】 黄鹤；

【作者基本信息】 四川大学 ， 内科学（心内）（专业学位）， 2023， 硕士

【摘要】 目的:探究预测非瓣膜性房颤患者致密自发显影/左房血栓的独立危险因素,建立致密自发显影/左房血栓的预测模型,并绘制列线图、推出网页计算器。材料和方法:回顾2011年1月1日至2020年12月31日,我院接受经食管超声心动图检查的拟行射频消融手术(radiofrequency ablation:RFCA)非瓣膜性房颤患者,最终筛选得到致密自发显影(dense spontaneous echo contrast,dense SEC)/左房血栓(left atrial thrombus,LAT)的患者170例(阳性组),同时随机选取经食道超声检查排除致密自发显影/左房血栓的非瓣膜性房颤患者510例(阴性组)。收集两组以下材料:(1)一般人口学资料,体重指数,心血管疾病史,抗凝药物使用情况,CHADS2及CHA2DS2-VASc总分;(2)实验室查血资料:红细胞计数、血红蛋白、红细胞压积、红细胞分布宽度-标准差、红细胞分布跨度-变异系数、白细胞计数、血小板计数、总蛋白、白蛋白、血肌酐、估计肾小球滤过率、血尿酸、甘油三脂、总胆固醇、高密度脂蛋白、低密度脂蛋白、血纤维蛋白原;(3)心脏超声指标:常规超声心动图测量左房直径、左室射血分数;经食道超声心动图测量左心耳排空速度。以7:3比例将数据集划分为训练集及验证集。训练集用于模型构建,验证集用于模型验证。单因素及多因素逻辑回归分析识别dense SEC/LAT的独立危险因素,建立dense SEC/LAT的预测模型。ROC曲线评价模型的预测能力,校准曲线评估模型预测概率与实际概率的一致性,依据模型绘制列线图并推出网页计算器。结果:1.致密自发显影/左房血栓组与阴性组:基线资料:致密自发显影/左房血栓组平均年龄更高,充血性心衰(congestive heart failure,C)、高血压、卒中/TIA(stroke:previous stroke/TIA/thromboembolism,S)及血管性疾病占比更大,CHADS2及CHA2DS2-VASc评分更高(P均<0.05),且致密显影/左房血栓占比随着CHADS2/CHA2DS2-VASc评分增加而增加,而性别分布、体重指数、口服抗凝药物及糖尿病比例两组间无差异。血液学指标中:致密显影/左房血栓组,红细胞分布宽度-标准差及红细胞分布宽度-变异系数、白细胞计数、肌酐、尿酸及纤维蛋白原值(fibrinogen,FIB)更高,白蛋白水平及估计肌酐清除率值更低(P均<0.05)。超声指标中:致密显影/左房血栓组左心耳排空速度(left atrial appendage-emptying velocity,LAA-EV)明显降低,左房直径(left atrial diameter,LAD)明显增大,左室射血分数(left ventricular ejection fraction,LVEF)较阴性组降低,两组间差异均具有统计学意义(P均<0.05)。2.训练集总数476例(其中阳性占比24.2%)、验证集总数204例(其中阳性占比27%),两组在基线资料、血液学指标及超声指标组间均无差异(P均>0.05)。3.训练集二元逻辑回归结果,以OR值(95%可信区间)表示:红细胞分布宽度标准差(RDW-SD)1.07(1.02-1.12),纤维蛋白原(FIB)2.33(1.61-3.38),左房直径(LAD)1.15(1.10-1.20),左室射血分数(LVEF)0.94(0.91-0.97),充血性心衰(C)4.89(1.89-16.00),卒中/TIA(S)3.31(1.51-7.23),将这六个变量用于建立Nomogram。4.训练集及验证集Nomogram的AUC值分别为:0.86,0.88,均较CHADS2及CHA2DS2-VASc模型高,表明新模型能更好地预测致密自发显影/左房血栓。校准曲线表明无论训练集、验证集,Nomogram的预测概率与实际概率均有较高的一致性(Hosmer–Lemeshow检验:训练集及验证集分别为:0.185,0.460,推出了网页计算器,决策者仅需要选择是/否或输入变量具体的值即可得到致密自发显影/左房血栓的预测概率。5.在CHA2DS2-VASc评估为低风险(男性0分,女性1分)的患者中,LAD及FIB是致密自发显影/左房血栓的独立影响因子,以OR值及95%可信区间表示:1.28(1.16-1.41),3.25(1.14-9.22)。6.CHA2DS2-VASc评分、左房直径、左室射血分数是规律3周抗凝后仍存在致密自发显影/左房血栓的独立影响因子,其OR值与95%可信区间分别为:1.33(1.08-1.65),1.10(1.03-1.17),0.93(0.89-0.97)。结论:RDW-SD,FIB,LAD,LVEF,C,S是非瓣膜性房颤患者致密自发显影/左房血栓的独立危险因素。结合独立危险因素建立了预测非瓣膜性房颤患者致密自发显影/左房血栓的Nomogram,并推出了快捷使用的网页计算器。对于CHA2DS2-VASc评估为低风险(男性得分为0,女性得分为1)的NVAF患者,LAD,FIB是致密自发显影/左房血栓的独立危险因子。CHA2DS2-VASc评分、LAD、LVEF是NVAF患者抗凝效果欠佳的独立危险因子。更多还原

【Abstract】 Objective:To explore risk factors for predicting dense spontaneous echo contrast(dense SEC)/left atrial thrombus(LAT),then develop a nomogram based on the available parameters and validate its efficacy in prediction of dense SEC/LAT.Materials and Methods:We reviewed the patients who had undergone a transesophageal echocardiography over the past 10 years.A total of 170 patients with non-valvular atrial fibrillation was diagnosed dense SEC/LAT(Dense SEC/LAT group).We randomly selected 510 patients without dense SEC/LAT(Negative group).The following data were collected:(1)General demographic data: body mass index,cardiovascular disease history,oral anticoagulant history,CHADS2,and CHA2DS2-VASc scores.(2)Laboratory data: red blood cell count,hemoglobin,hematocrit,red blood cell distribution width(RDW)-standard deviation,RDWcoefficient of variation(CV),white blood cell count,platelet count,total protein,albumin,serum creatinine,blood uric acid,triglyceride,total cholesterol,high density lipoprotein,low density lipoprotein,and fibrinogen.(3)Echocardiographic parameters: conventional echocardiography was used to evaluate the diameter of left atrium(LAD),and left ventricular ejection fraction(LVEF);transesophageal echocardiography was used to measure the emptying velocity of left atrial appendage(LAA-EV).The patients were divided into the training and validation cohorts with a ratio of 7:3.The training cohort was used for nomogram development,and the validation cohort was used to test the model’s performance.Logistic regression was used to build the model with the screened variables.The discriminative ability was measured using the area under the ROC curve(AUC).Calibration of the model was assessed by comparison of the predicted and observed probability of dense SEC/LAT,accompanied by the Hosmer-Lemeshow goodness of fit test.Then a web calculator was launched according to the Nomogram.Results:1.Comparison of the dense SEC/LAT group and the negative group: In baseline data,dense SEC/LAT group has a higher average age,and more congestive heart failure(C),hypertension,and stroke /TIA(stroke: Previous stroke/TIA/thromboembolism,S)and vascular disease,higher CHADS2 and CHA2DS2-VASc score(P < 0.05),the higher ratio of dense SEC/LAT associated with increasing CHADS2 and CHA2DS2-VASc scores.There were no differences in sex,body mass index,oral anticoagulant history and diabetes mellitus between the two groups.Laboratory indexes: In the dense SEC/LAT group,the white blood cell count,RDW-SD,RDW-CV,creatinine,uric acid and fibrinogen(FIB)were higher,and the albumin level were lower(P < 0.05).In terms of echocardiographic factors,the emptying velocity of the left atrial appendage(LAA-EV)and left ventricular ejection fraction(LVEF)was significantly decreased,while left atrial diameter(LAD)was significantly increased in the dense SEC/LAT group.(all P < 0.05).2.Dense SEC/LAT was found in 115(24.2%)in the training cohorts,and 55(27%)in the validation cohorts.There was no significant difference in baseline data,hematology indexes and echo factors(all P > 0.05).3.Based on the univariate and multivariate Logistic regression analysis of the training cohorts,results were reported as odds ratio(95%CI),RDW-SD 1.07(1.02-1.12),FIB2.33(1.61-3.38),LAD 1.15(1.10-1.20),LVEF 0.94(0.91-0.97),C 4.89(1.89-16.00),S3.31(1.51-7.23)were independently associated with dense SEC/LAT and were combined to construct a Nomogram.4.The discrimination performance of the model was good in both the training(AUC0.86 [95% CI,0.83–0.90])and validation(AUC 0.88 [95% CI,0.83–0.93])datasets.The P value for the Hosmer–Lemeshow test was more than 0.05 in the training and validation set(P =0.185,0.460 respectively).The calibration curves for training and validation sets showed that our Nomogram’s predicting probability was highly consistent with the actual probability.A web calculator was then developed,in which decision-makers can obtain the predicted probability of dense SEC/LAT by selecting Yes/No or entering specific values of variables.5.In patients assessed as low risk(0 in men and 1 in women)by CHA2DS2-VASc,LAD and FIB were independent impact factors,expressed as ORs and 95% confidence intervals: 1.28(1.61 to 1.41),3.25(1.14 to 9.22).6.CHA2DS2-VASc,LAD,LVEF were independent influencing factors of dense SEC/LAT despite anticoagulation,with OR and 95% confidence intervals of 1.33(1.08-1.65),1.10(1.03-1.17),and 0.93(0.89-0.97),respectively.Conclusions:RDW-SD,FIB,LAD,EF,C,and S were independent risk factors for dense SEC/LAT in patients with nonvalvular atrial fibrillation.Combined with those variables,a model for predicting dense SEC/LAT was established.Based on the model,a nomogram was drawn and a web calculator was launched.Besides,in patients assessed as low risk(0 in men and 1 in women)by CHA2DS2-VASc,LAD and FIB were independent impact factors.Furthermore,CHA2DS2-VASc score,LAD,LVEF were independent influencing factors of dense SEC/LAT despite anticoagulation.更多还原