Latest data from the international CONFIRM2 registry—an observational cohort study of patients undergoing clinically indicated coronary computed tomography angiography (CCTA)—have been presented at the Global CVCT Scientific Program 2025 (8–10 December, Washington, DC, USA), one of several recent reports from the research programme.
The study, assessing Cleerly’s artificial intelligence (AI)-enabled CCTA platform, looked at AI-guided quantification of atherosclerosis on coronary computed tomography (CT) for identification of high-risk individuals in non-obstructive coronary artery disease. Researchers analysed 6,550 patients (51.4% female, mean age 59 years) over 4.4 years who underwent CCTA with AI-based quantitative coronary CT analysis (AI-QCT) across international sites.
The findings indicated patients with non-obstructive coronary artery disease and high plaque burden (>750mm³) experienced a 21.9% major adverse cardiovascular events (MACE) rate, exceeding the 19.5% rate seen in obstructive coronary artery disease patients with similar plaque volumes, demonstrating that total plaque burden drives risk independent of stenosis severity.
AI-QCT plaque quantification identified a 12-fold risk gradient in non-obstructive disease, with MACE rates increasing from 1.7% with no measurable plaque to 21.9% with high plaque burden (>750mm³), revealing previously unrecognised risk stratification in patients traditionally considered lower risk.
While overall MACE rates differed between non-obstructive (3.3%) and obstructive coronary artery disease (14%), patients with non-obstructive disease and substantial plaque burden showed event rates approaching or exceeding those with flow-limiting stenoses, 18.4% vs 17% for Obstructive vs Non-Obstructive at TPV >750mm3.
After adjusting for baseline risk factors, MACE rates for patients with non-obstructive disease but high total and non-calcified plaque volumes were similar to those with obstructive disease and lower plaque volumes. This signifies a sufficient burden of plaque in non-obstructive disease may result in a critical mass of vulnerable disease that would warrant earlier intervention, Cleerly states in a press release.
“These findings fundamentally change how we should assess cardiac risk in patients by addressing the historically overlooked risk of non-obstructive coronary artery disease,” said Andrew D Choi (George Washington University School of Medicine, Washington, DC, USA). “By quantifying total plaque burden throughout the coronary tree, we’ve identified specific thresholds of patients with mild narrowing, but significant non-obstructive coronary artery disease that can carry equivalent or even potentially greater risk than those with traditionally identified severe narrowing. These findings reinforce the clinical value of whole heart plaque analysis in guiding clinical trials, enhancing preventive care decisions, and improving outcomes for the substantial population of patients with non-obstructive coronary disease.”
Other recent data releases from CONFIRM2 include a study looking at risk stratification according to the clinical likelihood of obstructive stenosis, presented at the American Heart Association (AHA) 2025 scientific sessions (7–10 November, New Orleans, USA).
Researchers reported that the addition of AI-QCT to European Society of Cardiology risk factor-weighted clinical likelihood (ESC RF-CL) scores improved prediction of MACE with area under the curve (AUC) increasing from 0.62 to 0.75 (p<0.001), and death/myocardial infarction (MI) prediction from AUC 0.60 to 0.71 (p<0.001).
While 78.4% of patients presenting with chest pain were classified as very-low (35.7%) or low (42.7%) likelihood for obstructive coronary artery disease (with true obstruction rates of only 5.8% and 14.2% respectively), AI-guided quantification of non-obstructive atherosclerosis successfully stratified these patients for future cardiovascular events, investigators reported.
“This is another step towards earlier identification and treatment for patients,” said Ibrahim Danad (VU University Medical Center, Amsterdam, the Netherlands), the study’s principal investigator. “In this case, it challenges what we know about obstructive disease and further validates the abilities of AI-QCT. This is another fascinating learning we’ve gained from the CONFIRM2 registry.”
An earlier study presented at the 2025 Transcatheter Cardiovascular Therapeutics (TCT) meeting (25–28 October, San Francisco, USA), assessing the use of the technology to identify patients at risk of future myocardial infarction or death, showed that both non-calcified plaque volume and diameter stenosis were independent predictors of these outcomes.
Patients who experienced events had nearly five times higher non-calcified plaque volumes compared to those without events (254 mm³ vs. 53 mm³), while those in the highest tertile of non-calcified plaque volume faced nearly double the risk of adverse outcomes when adjusted for maximal stenosis severity. Notably, many events occurred in patients who did not have obstructive coronary artery disease.
“These results challenge our traditional approach to cardiac risk assessment,” said Alexander Van Rosendael (Leiden University Medical Center, Leiden, the Netherlands), the study’s first author. “By identifying high-risk plaque features, particularly large non-calcified lesions, we can now detect vulnerable patients who would have been missed by conventional stenosis-focused evaluations.”
