Comparative evaluation of PET quantitative methods for myocardial viability assessment in CAD patients and its...

To compare different methods for myocardium quantification and exploratively evaluate the prognostic relationship between PET-measured viable myocardium and clinical outcomes in coronary artery disease (CAD) patients with ischemic cardiomyopathy (ICM). 183 CAD patients who underwent one-day ¹³NH₃ and ¹⁸F-FDG PET/CT myocardial viability assessment were retrospectively enrolled. Using Corridor4DM software, different types of myocardium (viable, normal, scar) were quantified via visual analysis, semi-quantitative scoring, and software automatic quantification. As an exploratory secondary analysis, patients were grouped and method-dependent grouping differences were analyzed. The correlation between viable myocardium, treatment decisions, and prognosis was retrospectively evaluated. Software identified the most normal myocardium and the least viable myocardium, while semi-quantitative scoring was the opposite. High inter-method consistency (all ICC>0.75) was observed, in which visual and software analyses showed optimal agreement, contrasting with scoring analysis. No significant differences were found among when viable myocardium<20%, but scoring analysis underestimated viable myocardium as its extent increased (P<0.001). Notably, all three methods exhibited concordant and comparable prognostic predictive value (AUC: 0.595-0.669 for OS and 0.529-0.533 for MACE). And exploratory prognostic analysis (n = 113) revealed no statistically significant interaction between treatment strategy and viability for either OS or MACE. However, a potential trend was observed where revascularization was associated with improved outcomes specifically in patients with a higher extent of viable myocardium (≥ 37%). Three analyses demonstrated high consistency in myocardium quantification, with visual and software analysis showing optimal agreement. Software analysis enhances clinical efficiency, offering improved sensitivity and objectivity. Furthermore, the three methods showed concordant prognostic predictive value; While exploratory analyses suggested potential prognostic trends, no statistically significant interaction was found between treatment strategy and viability. Accurately assessing viable myocardium via software automatic quantification offers an objective tool for assessment, though its role in guiding treatment decisions requires further validation.