AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Radiomics-Derived Liver Fat from Baseline CAC CT Identifies CAC-Zero Individuals at Risk for Annualized Coronary Calcium Progression
By Modanwal G, Glidden MD, Sirasapalli S et al.
Background: A coronary artery calcium (CAC) score of 0 indicates low near-term cardiovascular risk, yet CAC-zero individuals remain heterogeneous, with some showing accelerated annualized CAC progression (aCAC-P). Hepatic steatosis, a marker of systemic metabolic dysfunction, has been linked to accelerated atherosclerosis and may partly explain this heterogeneity. Opportunistic hepatic fat phenotyping from routine CAC CT may therefore help identify individuals within this otherwise low-risk group who are prone to greater aCAC-P. QuLF-CT is a validated radiomics-derived tool that estimates liver fat fraction from CAC CT, enabling this phenotyping without additional cost or radiation. Research Question: We hypothesized that higher baseline QuLF-CT liver fat estimates would identify CAC-zero individuals at increased risk for future aCAC-P, independent of obesity and traditional cardiometabolic risk factors. Methods: We analyzed 1,487 participants from the CLARIFY registry (ClinicalTrials.gov: NCT04075162) with baseline CAC = 0 and available follow-up CAC assessment. QuLF-CT was applied to baseline non-contrast CAC CT scans to estimate liver fat fraction. aCAC-P was calculated as change in CAC score divided by interscan interval; median interscan interval was 4.93 years (IQR 3.29-5.55). aCAC-P was evaluated using thresholds of >1, >5, and >10 Agatston units/year. Multivariable logistic regression assessed associations between QuLF-CT and aCAC-P after adjustment for age, sex, hypertension, smoking, diabetes, obesity, and dyslipidemia. Odds ratios were reported per 5-unit increase in QuLF-CT. Results: aCAC-P >1 Agatston unit/year occurred in 233 participants (15.7%). Mean aCAC-P increased across QuLF-CT tertiles, demonstrating a graded dose-response relationship between liver fat and aCAC-P (p-trend=0.024). In adjusted models, each 5-unit increase in QuLF-CT was associated with higher odds of aCAC-P >1 unit/year (OR 1.36, 95% CI 1.09-1.70; p=0.006). Associations strengthened with more substantial progression thresholds, including aCAC-P >5 units/year (OR 1.48, 95% CI 1.08-2.02; p=0.015) and aCAC-P >10 units/year (OR 1.87, 95% CI 1.29-2.72; p=0.001). Conclusion(s): Among individuals with baseline CAC = 0, higher liver fat estimates derived opportunistically from routine CAC CT were independently associated with future aCAC-P. These findings suggest that liver fat captures residual metabolic-vascular risk not fully reflected by CAC-zero status or obesity alone, and may help identify a subgroup of CAC-zero individuals more likely to demonstrate future calcium progression, potentially informing earlier repeat scanning or closer preventive assessment. This opportunistic approach requires no additional imaging, radiation exposure, or patient testing, supporting its potential integration into routine CAC-based prevention workflows.