To construct a perioperative risk prediction model for patients with coronary artery disease and heart failure with reduced ejection fraction undergoing surgical treatment based on cardiac magnetic resonance imaging
10.3760/cma.j.cn112434-20240303-00056
- VernacularTitle:基于心脏核磁构建冠心病合并射血分数降低型心衰外科治疗围手术期风险评估模型
- Author:
Kui ZHANG
1
;
Kaiwen LIU
;
Wei FU
;
Hongkai ZHANG
;
Jubing ZHENG
;
Yiping SUN
;
Lisong WU
;
Taoshuai LIU
;
Ran DONG
Author Information
1. 首都医科大学附属北京安贞医院心脏外科，北京 100029
- Publication Type:Journal Article
- Keywords:
Coronary heart disease;
Heart failure;
Cardiac magnetic resonance imaging;
Coronary artery bypass grafting
- From:
Chinese Journal of Thoracic and Cardiovascular Surgery
2025;41(1):22-29
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To investigate the application value of cardiac magnetic resonance imaging(CMR) in evaluating the perioperative risk of surgical treatment in patients of coronary artery disease and heart failure with reduced ejection fraction.Methods:A total of 78 patients diagnosed with CAD and HFrEF who underwent CABG in Beijing Anzhen Hospital from January 2018 to October 2021 were retrospectively enrolled. All patients underwent CMR examination before CABG. The perioperative data of the two groups were compared, and the risk factors that may lead to perioperative MACCE were analyzed by LASSO regression. Then, logistic regression was used to establish a prediction model and internal validation was performed to evaluate the diagnostic efficiency and accuracy of the model by Bootstrap method. Results:There were 24 patients(30.8%) with perioperative MACCE and 54 patients(69.2%) without perioperative MACCE. LASSO regression was used to screen out three factors related to the outcome. Multivariate logistic regression analysis showed that LGE in the third and eighth segment of left ventricle and diastolic radial strain rate were independent risk factors for perioperative MACCE. The area under the curve of the prediction model constructed with CMR was 0.799(95% CI: 0.696-0.901), so the discrimination was good. The calibration curve showed that the prediction curve was basically fit to the standard curve, and the Hosmer- Lemeshow test P=0.797, indicating high prediction accuracy. Conclusion:CMR is a valuable test for evaluating perioperative risk in patients with coronary heart disease complicated with HFrEF. To establish the risk prediction model combined with CMR can provide some reference for the assessment of perioperative risk in these patients undergoing surgical treatment.
