Development and validation of a diagnostic nomogram to predict significant stenosis of the left anterior descending branch of the coronary artery by stress echocardiography.

Background: Noninvasive detection of coronary artery disease (CAD) with significant coronary stenosis by echocardiography remains challenging. Myocardial work (MW) is a noninvasive method for the quantitative assessment of left ventricular function, which, in combination with stress imaging, enables the detection of myocardial ischemia during myocardial stimulation. We aimed to preliminarily explore the diagnostic value of regional stress MW combined with coronary flow reserve (CFR) in identifying significant stenosis of the left anterior descending artery (LAD). Methods: A retrospective collection of 120 patients suspected of CAD with coronary angiography was performed, including 63 with nonsignificant stenosis in the LAD and 57 with significant stenosis in the LAD. In addition to conventional echocardiographic parameters, all individuals underwent stress echocardiography (SE) with pharmacological stress. We statistically compared longitudinal strain (LSLAD), peak strain dispersion (PSD), work index (WILAD), work efficiency (WELAD), and CFR before and after drug stress. The area under the curve (AUC) of the receiver operating characteristic (ROC) curve was used to evaluate the diagnostic value of the univariate logistic regression model (ULRM) and multivariate logistic regression model (MLRM) in detecting significant stenosis in the LAD and myocardial ischemia. Decision curve analysis (DCA) was applied to evaluate the clinical net benefit of the best model and validate its robustness using an independent cohort. Results: Conventional echocardiographic parameters showed no significant difference between the nonsignificant and significant stenosis groups (P>0.05). Analysis of MW-related parameters showed that the group with nonsignificant stenosis differed from the group with significant stenosis. WILAD and WELAD were reduced in the significant stenosis group at rest (P<0.05). LSLAD, WILAD, and WELAD were reduced, and PSD was elevated in the significant stenosis group at peak stress (all P<0.05). Patients in the significant stenosis group had significantly lower CFR than those in the nonsignificant stenosis group (P<0.001). In univariate logistic regression, CFR had the largest AUC (0.837), with sensitivity and specificity of 0.634 and 0.791, respectively; among MW parameters, Peak WELAD had the largest AUC (0.825), with sensitivity and specificity of 0.902 and 0.581, respectively. Multivariate logistic regression showed that the best MLRM1 (AUC =0.889, sensitivity of 0.902, specificity of 0.721) and the most concise MLRM2 (AUC =0.857, sensitivity of 0.756, specificity of 0.884) demonstrated superior performance in predicting significant stenosis of the LAD and impairment of left ventricular function. DCA showed that the best MLRM1 provides higher net gains within a reasonable threshold. Furthermore, in the independent validation cohort, the best MLRM1 achieved an AUC of 0.888, suggesting excellent generalizability of the model. Conclusions: Stress MW parameters and CFR demonstrated potential discriminatory ability for early myocardial ischemia caused by significant stenosis of LAD in this cohort. These results support the feasibility of exploring MLRM models that incorporate stress MW and CFR as noninvasive screening tools in future prospective studies.