Mental stress-induced myocardial ischemia detected by global longitudinal strain and quantitative myocardial contrast echocardiography in women with nonobstructive coronary artery disease

J Am Soc Echocardiogr. 2024 May 16:S0894-7317(24)00231-1. doi: 10.1016/j.echo.2024.05.008. Online ahead of print.

Abstract

Background: The utility of radionuclide myocardial perfusion imaging including positron emission tomography (PET) for diagnosing mental stress-induced myocardial ischemia (MSIMI) is clinically restricted. This study aims to assess the diagnostic performance of novel echocardiographic techniques, including automated strain and quantitative myocardial contrast echocardiography (MCE) with dedicated software and deep neural network (DNN) model, for MSIMI detection. The secondary objective was to explore the correlation between changes in myocardial blood flow (MBF) and MSIMI.

Methods: 72 female patients aged 18 to 75 with angina and nonobstructive coronary artery disease (ANOCA), and 23 healthy controls were prospectively recruited. Both echocardiography with contrast agent and PET imaging were performed during structured mental stress testing. MSIMI was defined as a summed difference score ≥3 on PET. Echocardiographic parameters including left ventricular global longitudinal strain (LVGLS), β and A×β were obtained, and their trends during mental stress testing were observed. ΔGLS, β reserve and A×β reserve were respectively calculated.

Results: 32 ANOCA patients (44%) and 1 control (4%) were diagnosed with MSIMI (P<0.01). For ANOCA patients with MSIMI, LVGLS, β and A×β declined to varied extent during mental stress testing compared to those without MSIMI and the controls (P<0.05). Bland-Altman plots demonstrated good consistency between β reserve and A×β reserve output by the DNN model and iMCE software. Receiver operating characteristic (ROC) curve analyses showed that ΔGLS, β reserve and A×β reserve demonstrated favorable ability to predict MSIMI, especially the combination of A×β reserve using iMCE analysis and ΔGLS (area under the curve [AUC] 0.94, sensitivity 83%, specificity 97%).

Conclusions: Novel technologies in echocardiography exhibit the potential to be a clinical alternative to cardiac PET for effectively detecting MSIMI. Attenuated MBF response during structured mental stress testing was correlated with MSIMI, providing a reasonable explanation for the chest discomfort persisting in ANOCA women.

Keywords: Deep neural network; Global longitudinal strain; Mental stress-induced myocardial ischemia; Myocardial blood flow; Myocardial contrast echocardiography.