AI-guided Quantitative Coronary Computed Tomography Angiography (AI-QCT) may help guide tailoring of anti-atherosclerotic therapies or coronary interventional procedures to reduce cardiovascular events in patients with suspected coronary artery disease (CAD), based on findings from the CONFIRM-2 trial presented at TCT 2024.

"AI-QCT allows standardized, reproducible, quantitative automatic whole heart plaque quantification with high degree of correlation to invasive gold standards," according to Alexander van Rosendael, MD, PhD, who presented the findings on behalf of the trial investigators.

The trial assessed 3,551 patients from 18 sites in 13 countries, approximately 50% of whom were female and 79% were Caucasian. The median follow-up period was 4.8 years. None of the participants had known CAD and had been referred to coronary computed tomography angiography (CTA) for clinical suspicion of CAD.

The primary composite endpoint was all-cause mortality, myocardial infarction, stroke, congestive heart failure, late revascularizations (>90 days) and hospitalizations for unstable angina. The composite secondary endpoint was all-cause mortality and myocardial infarction. Overall results found patients with a low AI-QCT had fewer primary and secondary endpoint events than those with low Diamond Forrester Pre-Test Probability (PTP). Similarly, patients with high AI-QCT had more events than those with high Diamond Forrester PTP.

"The AI-QCT quantification of coronary CTA lumen diameter stenosis and the noncalcified plaque volume were most predictive of MACE and added discriminative accuracy to Diamond and Forrester PTP and risk factors," said van Rosendael.

"The next question is whether AI-generated risk estimation leads to better outcomes," says ACC.org Editor-in-Chief Kim A. Eagle, MD, MACC. "This would require a randomized study of using the data to adjust treatment."