Utility of CAD-RADS 2.0 Plaque Burden Grades and Stenosis Categories on Coronary CTA for Predicting Cardiac Events in Patients With Acute Chest Pain: A Multicenter Study

Background: Coronary Artery Disease-Reporting and Data System (CAD-RADS) 2.0 incorporated reporting of plaque burden grades in addition to stenosis categories. Objective: To assess the utilities of plaque burden grades and CAD-RADS categories, determined on coronary CTA using CAD-RADS 2.0, in predicting cardiac events in patients presenting to the emergency department (ED) with acute chest pain. Methods: This retrospective study included patients who underwent coronary CTA after presenting to the ED with acute chest pain at one of four centers from January 2018 to December 2021. A single radiologist reviewed examinations at each center to assign a plaque burden grade by calculating a coronary artery calcium score on noncontrast images and to assign a CAD-RADS category by assessing vessel stenosis on contrast-enhanced CTA images, both using CAD-RADS 2.0; each center's reviewing radiologist also assessed examinations for presence of high-risk plaque. The EMR was reviewed for cardiac events, reflecting a composite outcome of cardiac-related death, myocardial infarction, or hospitalization for unstable angina. Prognostic models were compared. Results: The study included 2032 patients (1085 male, 752 women; mean age, 58.4 years). During median follow-up of 15.2 months, 63 (3.1%) patients experienced cardiac events. In a multivariable Cox model adjusting for clinical variables, cardiac events showed significant independent associations with CAD-RADS 3 (HR=7.1), CADRADS 4 (HR=13.6), CAD-RADS 5 (HR=17.6), and high-risk plaque (HR=2.5), but not with plaque burden grades. For predicting cardiac events, C-statistic was 0.67 for a model including clinical variables; 0.74 for a model including clinical variables and plaque burden grades; 0.86 for a model including clinical variables, CAD-RADS categories, and high-risk plaque; and 0.87 for a model including clinical variables, plaque burden grades, CAD-RADS categories, and high-risk plaque. The model with clinical variables, CAD-RADS categories, and high-risk plaque, but without plaque burden grades, showed highest net clinical benefit across threshold probabilities from 20-100%. Conclusion: Addition of plaque burden grades did not provide further prognostic benefit in models using CADRADS categories. Clinical Impact: Inclusion of an additional noncontrast acquisition for CAC scoring within coronary CTA protocols may not improve risk stratification for patients presenting to the ED with chest pain.