Construction of a nomogram model for predicting MACE in patients with moderate coronary stenosis based on coronary CT fractional flow reserve and clinical characteristics

Objective  To construct a nomogram model for predicting major adverse cardiovascular events (MACE) in patients with moderate coronary stenosis on the basis of coronary CT fractional flow reserve (FFR) and clinical characteristics.

Methods  A retrospective cohort study was performed with 290 patients who had moderate coronary stenosis (comprising 158 males and 132 females with an age of (56.9±6.7) years) and were admitted to Qianjiang Central Hospital Affiliated to Yangtze University from February 2020 to August 2023. The patients were divided into MACE and non-MACE groups according to the incidence of MACE. Predictive factors for MACE were then analyzed using least absolute shrinkage and selection operator (LASSO) regression analysis, and influencing factors for MACE were identified through Logistic regression analysis. A nomogram model was developed using R software (Version 4.2.3) to predict MACE. The model′s predictive performance was then assessed using the receiver operating characteristic (ROC) curve, and its overall clinical value was assessed using calibration curves, the Hosmer-Lemeshow test, and decision curve analysis.

Results  Of the 290 patients, 57 were assigned to the MACE group, and 233 were assigned to the non-MACE group. LASSO regression analysis results indicated that multivessel disease, sleep disturbance, history of smoking, sarcopenia, history of diabetes, and FFR are predictors of MACE in patients with moderate coronary stenosis. Logistic regression analysis indicated that multivessel disease (OR=2.779, 95%CI: 1.317–5.861, P=0.007), history of smoking (OR=4.878, 95%CI: 2.019–11.787, P<0.001), sarcopenia (OR=4.382, 95%CI: 1.825–10.523, P=0.001), and history of diabetes (OR=2.205, 95%CI: 1.025–4.745, P=0.043) are risk factors for MACE, whereas FFR served as a protective factor (OR=0.161, 95%CI: 0.132–0.582, P<0.001). A nomogram model was developed on the basis of the influencing factors to predict the incidence of MACE in the patients. The area under the ROC curve for the model was 0.850 (95%CI: 0.797–0.903, P<0.001). Moreover, the calibration curve results was basically consistent with the ideal curve, and the Hosmer-Lemeshow test confirmed a satisfactory model fit (R²=0.409, P=0.875). Decision curve analysis revealed that the nomogram model yielded a higher net benefit rate within a threshold probability range of 1%–92% for predicting MACE.

Conclusion  The nomogram model developed for predicting MACE in patients with moderate coronary stenosis demonstrated strong predictive performance and can provide theoretical support for the clinical prevention of MACE in this patient population.