AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Agatston-2.0: AI-Derived Coronary Artery Calcium Burden and Density Profiling Improve Coronary Heart Disease Risk Stratification Within CAC Scores 1-99
By Naghavi M, Azimi A, Atlas K et al.
Background: Coronary artery calcium (CAC) scoring is recommended for cardiovascular risk stratification, with CAC scores of 1-99 reflecting low-to-intermediate risk. However, outcomes within this range are heterogeneous. The conventional Agatston scoring method (Agatston-1.0) applies a fixed attenuation threshold of ≥130 HU, potentially missing lower-attenuation calcification, and its peak-density weighting does not yield a continuous plaque density profile. Objectives: To validate Agatston-2.0, an AI-based framework integrating continuous calcium burden quantification and plaque density profiling, and to evaluate its incremental prognostic value beyond Agatston-1.0 among individuals with CAC scores of 1-99. Methods: We retrospectively analyzed 1,238 participants from a prospective multi-ethnic cohort with baseline CAC of 1-99, followed for incident CHD (median 17.4; maximum 19.4 years). Agatston-2.0 generates: (1) a continuous AI-derived CAC score (AI-CAC) via probabilistic voxel-wise quantification without fixed attenuation thresholds; and (2) a plaque density metric quantifying hyperdense voxels >10 SD above scanner calibration. Participants were stratified into four groups: AI-CAC ≤99 or >99 crossed with top-quartile (Dense) versus lower three quartiles (Non-dense) plaque density, with AI-CAC ≤99 Non-dense as reference. Cox models were adjusted for age, sex, race/ethnicity, blood pressure, antihypertensive use, LDL and HDL cholesterol, triglycerides, lipid-lowering therapy, diabetes, and smoking. Discrimination was assessed using C-index and net reclassification improvement (NRI). Results: Among 1,238 participants (153 CHD events; 12.4%; mean age 64.4±9.5 years; 53.2% male), 183 (14.8%) had AI-CAC>99. AI-CAC>99 was associated with higher CHD risk (unadjusted HR 1.54, 95% CI 1.04-2.30; adjusted HR 1.37, 95% CI 0.92-2.06). Higher plaque density was inversely associated with CHD risk (unadjusted HR 0.55, 95% CI 0.34-0.88; adjusted HR 0.58, 95% CI 0.35-0.95). Kaplan-Meier analysis demonstrated a risk gradient across the four groups: 19-year cumulative incidence was 26.7% for AI-CAC>99 Non-dense (p=0.0095), 13.7% for AI-CAC>99 Dense (p=0.8533), 15.3% for AI-CAC≤99 Non-dense (reference), and 9.4% for AI-CAC≤99 Dense (p=0.0420). Agatston-2.0 improved discrimination (C-index 0.619 vs. 0.558; p=0.006) and reclassification (NRI +0.247, 95% CI +0.107-+0.389; p<0.001) beyond Agatston-1.0. Conclusions: Agatston-2.0 improves risk discrimination within CAC scores of 1-99 by integrating calcium burden with plaque density profiling. AI-CAC>99 with lower plaque density identifies higher-risk individuals not captured by Agatston-1.0.