基于冠状动脉CT-血流储备分数构建冠状动脉中度狭窄进行性加重的动态评分模型

Construction of a dynamic scoring model for progressive worsening of moderate coronary stenosis based on coronary CT-fractional flow reserve

摘要:

目的]探讨冠状动脉中度狭窄进行性加重的危险因素,并基于冠状动脉CT-血流储备分数(FFR)构建冠状动脉中度狭窄进行性加重的列线图模型。 [方法]选取2020年4月—2023年4月医院收治的293例冠状动脉中度狭窄患者作为研究对象,对所选患者进行冠状动脉CT血管造影(CCTA)检查,使用CT-FFR软件分析CCTA图像并得到FFR。根据冠状动脉中度狭窄进行性加重情况将其分为加重组和稳定组,并收集两组患者的性别、年龄、高血脂、脑梗死史、多支血管病变、肥胖、糖尿病、营养不良、缺乏体力活动、高血压、饮酒、吸烟及居住地等资料。采用LASSO分析中的十折交叉验证筛选冠状动脉中度狭窄进行性加重的预测因素,采用Logistic回归筛选冠状动脉中度狭窄进行性加重的危险因素,采用R(4.2.3)建立冠状动脉中度狭窄进行性加重的列线图模型,并对此列线图模型进行验证。 [结果]在293例冠状动脉中度狭窄患者中,进行性加重患者有61例,冠状动脉中度狭窄进行性加重的发生率为20.82%(61/293)。加重组与稳定组的性别、年龄、高血脂、脑梗死史、多支血管病变、营养不良、高血压、饮酒及居住地差异均无统计学意义(P＞0.05),而加重组的肥胖、糖尿病、缺乏体力活动及吸烟占比显著高于稳定组,FFR显著低于稳定组(P＜0.05)。LASSO分析显示,脑梗死史、肥胖、糖尿病、缺乏体力活动、吸烟及FFR是系数不为零的预测因素,经过十折交叉验证后,保留脑梗死史、肥胖、糖尿病、缺乏体力活动、吸烟及FFR 6个变量进入模型。Logistic回归分析显示,肥胖(OR＝2.411,95%CI:1.151～5.053)、糖尿病(OR＝3.401,95%CI:1.671～6.923)、缺乏体力活动(OR＝2.818,95%CI:1.427～5.564)、吸烟(OR＝3.577,95%CI:1.526～8.387)均是冠状动脉中度狭窄进行性加重的危险因素,FFR(OR＝0.001,95%CI:0.000～0.036)是冠状动脉中度狭窄进行性加重的拮抗因素(P＜0.05)。基于危险因素建立了冠状动脉中度狭窄进行性加重的动态列线图模型,该列线图模型ROC曲线下面积是0.777(95%CI:0.711～0.842)；校正曲线的预测值和实际值基本吻合；决策曲线显示阈值概率是2%～64%时,列线图对冠状动脉中度狭窄进行性加重的预测具有良好的获益值。 [结论]本研究构建的列线图模型对冠状动脉中度狭窄的进展预测具有较高准确性,临床实用性良好。

Abstract:

Aim To explore the risk factors for the progressive worsening of moderate coronary stenosis and construct a nomogram model for predicting the progression of moderate coronary stenosis based on coronary CT-fractional flow reserve (FFR). Methods 293 patients with moderate coronary stenosis admitted to the hospital from April 2020 to April 2023 were selected as the research subjects. Coronary CT angiography (CCTA) was performed on the selected patients, and CT-FFR software was used to analyze the CCTA images and obtain the FFR values. According to the progressive worsening of moderate coronary stenosis, the patients were divided into two groups:the progression group and the stable group, and the data of gender, age, hyperlipidemia, history of cerebral infarction, multivessel disease, obesity, diabetes, malnutrition, lack of physical activity, hypertension, drinking, smoking and place of residence of the patients in the two groups were collected. The ten-fold cross-validation in LASSO analysis was used to screen for predictive factors of progressive worsening of moderate coronary stenosis, Logistic regression was used to screen for risk factors of progressive worsening of moderate coronary stenosis,Rü (4.2.3) was used to establish a nomogram model of progressive worsening of moderate coronary stenosis, and this nomogram model was validated. Results Among 293 patients with moderate coronary stenosis, there were 61 cases of progressive worsening, and the incidence of progressive worsening of moderate coronary stenosis was 20.82% (61/293). The sex, age, hyperlipidemia, history of cerebral infarction, multivessel disease, malnutrition, hypertension, alcohol consumption and place of residence of the progression group and the stable group had no statistical significance (P＞0.05), while the proportion of obesity, diabetes, lack of physical activity and smoking in the progression group was significantly higher than those in the stable group, and the FFR was significantly lower than that in the stable group (P＜0.05). LASSO analysis showed that the history of cerebral infarction, obesity, diabetes, lack of physical activity, smoking and FFR were predictive factors with non-zero coefficients. After ten-fold cross-validation, six variables including the history of cerebral infarction, obesity, diabetes, lack of physical activity, smoking and FFR were retained into the model. Logistic regression analysis showed that obesity (OR＝2.1,5%CI:1.151～5.053), diabetes (OR＝3.1,5%CI:1.671～6.923), lack of physical activity (OR＝2.8,5%CI:1.427～5.564), smoking (OR＝3.7,5%CI:1.526～8.387) were all risk factors for progressive worsening of moderate coronary stenosis, and FFR (OR＝0.1,5%CI:0.000～0.036) was the antagonistic factor for progressive worsening of moderate coronary stenosis (P＜0.05). A dynamic nomogram model for the progressive worsening of moderate coronary stenosis was established based on risk factors, and the area under the ROC curve of the nomogram model was 0.777 (95%CI:0.711～0.842); The predicted values of the calibration curve are basically consistent with the actual values; When the decision curve showed a threshold probability of 2% to 64%, the nomogram model had a good benefit value for predicting the progressive worsening of moderate coronary stenosis. Conclusion The nomogram model constructed in this study has high accuracy in predicting the progression of moderate coronary stenosis and good clinical practicality.