News|Videos|November 20, 2026

AI-Enabled CPA With CCTA Associated With Improved LDL-C, Better Goal Achievement

Fact checked by: Ryan Livingston

Leslee Shaw, PhD, discusses her analysis of the DECIDE registry, indicating the possible connection between AI-CPA and superior lipid management.

Supplementing coronary computed tomographic angiography (CCTA) with Artificial Intelligence-enabled Coronary Plaque Analysis (AI-CPA) was associated with greater LDL-C reductions, according to new data from the DECIDE registry.1

These data were presented at the American Heart Association (AHA) Scientific Sessions 2026 in Chicago, Illinois, by Leslee Shaw, PhD, director of the Blavatnik Family Women’s Health Research Institute and professor of medicine, population health science and policy, and obstetrics at the Icahn School of Medicine at Mount Sinai.1

“With CCTA, you get stenosis measurements, and you can do visual plaque analysis,” Shaw told HCPLive in an exclusive interview. “But AI-enabled coronary plaque analysis provides you with a quantitative assessment, which is a huge advancement in the field. It provides you not only with total plaque volume across all the major epicardial coronary arteries, but also compositional changes, lipid-rich plaque, calcified plaque, and even the very low-density, very high-risk plaque.”

The DECIDE registry is an ongoing prospective, multicenter, pre-post interventional study assessing the influence of AI-CPA on changes in clinical management for symptomatic patients undergoing CCTA with CCTA-defined atherosclerotic plaque. The study enrolled patients into 6 main groups, all of which were required to have CCTA and plaque detected. Additionally, group 3 patients were required to have a stress test and no CCTA within 90 days before site activation, while group 6 was required to be clinically stable without symptoms of CAD at CCTA.2,3

Group 1 patients were given no AI-CPA, while those in group 2 were administered AI-CPA 90 days after CCTA, and group 3 was administered only a stress test. Group 4 was comprised of newly scanned, symptomatic patients with plaque detected who had not undergone prior revascularization; group 5 included newly scanned symptomatic patients with plaque detected and prior revascularization; and group 6 included newly scanned asymptomatic patients with plaque detected and no prior revascularization. Groups 4-6 all received AI-CPA.2

A total of 972 symptomatic patients with atherosclerotic plaque were enrolled in the study, with a median age of 64 years (interquartile range [IQR], 56-72 years). Changes in management after AI-CPA occurred in 51.3% of patients (95% CI, 48.2-54.5%) and were more frequent among those with more extensive plaque (≤67.8% in the highest stage; P <.001). Intensifying medical therapy was the most common change, effecting 35.6% of patients.3

In the present analysis, Shaw and colleagues divided patients into 2 main groups: those with CCTA alone, and those with both CCTA and AI-CPA. The primary endpoints included changes in lipid-lowering therapy (LLT), changes in LDL-C, achievement of LDL-C targets of <70 mg/dL, and 1-year incidence of cardiovascular death, myocardial infarction (MI), or cardiovascular hospitalization.1

A total of 12,850 patients were evaluated; Shaw and colleagues divided these patients into a retrospective cohort of 1049 patients managed with CCTA alone and 11,801 receiving both CCTA and AI-CPA. Baseline characteristics were well-matched between the 2 groups, with baseline LDL-C averaging 104.6 and 101.3 mg/dL, respectively.1

In paired analysis with serial lipid panels, Shaw and colleagues found that patients managed with AI-CPA demonstrated substantially lower LDL-C levels (78.2 mg/dL vs 85.3 mg/dL; P = .041) and more frequent achievement of LDL-C goals (49.3% vs 35.4%; P = .007) at 90 days post-scan. At 1 year of follow-up, the patients receiving AI-CPA also displayed lower rates of MI (0.5% vs 1%; P = .04), while maintaining similar rates of cardiovascular death (0.03% vs 0%; P = 1) and cardiovascular hospitalization (7% vs 5.7%; P = .12).1

Ultimately, Shaw and colleagues determined that AI-CPA on top of standard CCTA led to greater LDL-C management and reductions compared to CCTA alone. This, coupled with its reductions in MI risk over 1 year, indicated its potential for optimizing preventive cardiovascular care. However, Shaw also acknowledges the need for further investigation before any practical applications of AI-CPA can be implemented.1

“We need randomized clinical trials, and in some cases, we may need to know much more information it gives over a calcium score, for example, in an asymptomatic population,” Shaw said. “We also need to know if we need to do a second CTA after treating patients to look for changes. All of these are important pieces of the equation, which we will need in order to really drive guideline change and have plaque analysis be a core element. Right now, we’re in the exploratory phase.”

Editors’ Note: Shaw reports a relevant disclosure with HeartFlow.

References
  1. Shaw L, Blankstein R, McCarthy C, et al. Artificial Intelligence-Enabled Coronary Plaque Analysis Improves Medical Management and Clinical Outcomes Compared to Coronary Computed Tomographic Angiography: Insights from the DECIDE Registry. Presented at the American Heart Association (AHA) Scientific Sessions 2026, Chicago, IL. November 6-9, 2026.
  2. HeartFlow. AI DErived Plaque Quantification: CCTA and AI-QCPA for Determining Effective CAD Management (DECIDE). ClinicalTrials.gov Identifier: NCT06376851. Updated November 19, 2026. Accessed November 20, 2026. https://clinicaltrials.gov/study/NCT06376851
  3. Rinehart S, Blankstein R, McCarthy CP, et al. Guiding preventive care strategies for patients with atherosclerotic plaque on Coronary CTA. JACC: Cardiovascular Imaging. 2026;19(6):706-716. doi:10.1016/j.jcmg.2026.01.014

Related to this article