Dahdal · European heart journal. Cardiovascular Imaging 2025 · observational cohort study · n=2404

CCTA-Derived coronary plaque burden offers enhanced prognostic value over CAC scoring in suspected CAD patients.

Cited 21 times in the scientific literature.

Level 3 - non-randomized controlled study

Observational cohort study with follow-up evaluating prognostic utility

PubMed 40131307 · doi:10.1093/ehjci/jeaf093 · record verified 2026-08-27

What was done

An observational cohort study evaluated 2,404 patients with suspected coronary artery disease (CAD) and no prior CAD history who underwent coronary artery calcium (CAC) scoring and coronary computed tomography angiography (CCTA). Plaque metrics—percent atheroma volume (PAV) and non-calcified plaque volume percentage (NCPV%)—were quantified using a commercial AI tool (Cleerly, Inc). The primary endpoint was a composite of all-cause mortality and non-fatal myocardial infarction (MI), analyzed using Cox proportional hazards models adjusted for clinical risk factors and early revascularization over a median follow-up of 7.0 years.

What was found

Over 7.0 years of median follow-up, 208 patients (8.7%) experienced the primary endpoint, including 73 non-fatal MIs (3.0%). For the composite outcome, PAV had significantly greater discriminatory power than CAC scoring (AUC = 0.729 vs. 0.706, P = 0.016). For MI prediction, PAV significantly outperformed CAC (AUC = 0.791 vs. 0.699, P < 0.001), with NCPV% providing the highest accuracy (AUC = 0.814, P < 0.001).

Why it matters

Quantitative AI-based CCTA assessment of plaque burden, especially non-calcified plaque volume, provides superior risk stratification for myocardial infarction and mortality compared with standard CAC scoring alone.

Limits

The abstract does not specify whether the cohort was prospectively or retrospectively enrolled, nor does it report baseline demographics, hazard ratios with confidence intervals, or calibration metrics. Findings rely on a specific proprietary AI platform, which may not generalize to other quantification algorithms.

Cited by