Stress Tests vs. Plaque Scans: What 523 Patients’ Prescriptions Showed

Stress Tests vs. Plaque Scans: What 523 Patients' Prescriptions Showed
https://www.youtube.com/watch?v=yBeTfBfnhMw
AI Coronary Plaque Scan: How Plaque Detection Shifts Cardiac Care

The Scan Changes the Prescription: How AI Plaque Detection Shifts Cardiac Care from Stress to Substrate

In a randomized trial of 523 patients, AI-assisted coronary plaque scan analysis prompted more certain diagnoses and preventive treatment—but stopped short of proving it saves lives

From Functional Testing to Plaque Measurement: A Paradigm Shift in Chest Pain Diagnosis

For decades, stress testing has helped clinicians evaluating chest pain answer a functional question: Does blood flow keep up when the heart works harder? This approach looks for inducible ischemia—the downstream effect of coronary disease under load—rather than directly measuring the amount of plaque in the artery wall.

The PARAMOUNT trial compared that functional approach with a different starting point. Rather than asking only whether coronary disease limits blood flow under stress, clinicians in the plaque-guided arm also received a quantitative assessment of atherosclerotic plaque in the coronary arteries. All participants were already symptomatic, so the trial does not show detection before symptoms or support population screening.

Plaque-guided care uses AI-assisted coronary CT analysis to quantify plaque and support visualization of coronary anatomy. In PARAMOUNT, that information was paired with plaque-based ischemia assessment and used to guide care. The preliminary results show changes in diagnostic certainty, LDL cholesterol, treatment use, and the pattern among patients referred for invasive angiography; they do not establish prevention of future cardiac events.

The comparison therefore concerns two diagnostic pathways: one centered on functional stress testing and one centered on direct plaque assessment plus an ischemia estimate.

PARAMOUNT by the Numbers: Trial Design and Population

The PARAMOUNT trial represents a carefully structured investigation into how artificial intelligence can reshape cardiovascular diagnosis. Conducted across 21 medical centers, the prospective randomized trial enrolled 530 symptomatic outpatients between January 2025 and February 2026. All participants met specific criteria: suspected coronary artery disease, LDL cholesterol levels of 100 mg/dL or higher, and no prior history of coronary disease. This focused approach ensured researchers studied a defined population rather than casting a wide net across all chest-pain presentations.

The locked preliminary analysis included 523 participants, revealing a striking demographic detail: 57% were women—a historically rare majority in cardiovascular research. The typical participant was a median age of 57 years with a baseline LDL cholesterol of 127 mg/dL, representing stable symptomatic patients rather than emergency cases or asymptomatic screening populations.

Participants were randomly assigned to one of two pathways: the experimental arm received plaque-guided care using AI quantitative analysis to guide treatment, while the control arm received usual care based on traditional functional stress testing. Researchers assessed the hierarchical primary endpoint at 180 days, a timeframe that captured diagnostic certainty, LDL cholesterol, and treatment-pathway differences rather than long-term clinical outcomes.

The Results: Better Diagnosis, More Preventive Therapy, No Proof of Life Saved

The PARAMOUNT trial met its prespecified hierarchical composite endpoint, with a reported win ratio of 2.0 (95% confidence interval 1.37–2.56). This metric compares ordered outcomes across patient pairs; it is not a percentage reduction in heart attacks or deaths and does not establish universal superiority over other diagnostic approaches.

The largest contributor centered on diagnostic certainty. The reported win difference for certainty about whether coronary disease caused a patient’s symptoms was 16.6 percentage points in favor of plaque-guided care. The supplied preliminary report supports a clinician-certainty difference; it does not report patient reassurance as an outcome.

Treatment use also differed. In the plaque-guided group, lipid-lowering therapy use rose from 31% before testing to 75% afterward. In the usual-care group, it increased from 32% to 47%. The preliminary report therefore supports a difference in prescribing after the two testing pathways.

Among participants who underwent invasive coronary angiography, 38% in the plaque-guided arm did not go on to revascularization, compared with 60% in the usual-care arm. Investigators also reported no increase in overall catheterization or revascularization rates. This is a difference within the angiography subgroup; it does not by itself establish who truly needed intervention or whether avoiding a procedure was safe.

The preliminary report did not establish or report a reduction in heart attacks or mortality. More diagnostic certainty, lower LDL cholesterol, and greater use of lipid-lowering therapy are pathway results; whether they translate into longer life or fewer cardiovascular events remains unknown.

These preliminary results were presented November 7, 2026, and remain unpublished and subject to peer review. Full publication is expected in early 2027.

The Regulatory Story: FDA Clearance Is Not Clinical Proof

In July 2026, the FDA cleared Cleerly Plaque through the Class II 510(k) pathway as an image-processing tool for measurement and visualization of coronary arteries following CT angiography. The substantial-equivalence submission included a validation dataset of 120 de-identified coronary CT studies representing multiple scanner manufacturers.

However, here is a critical distinction: FDA clearance does not prove that using this tool improves patient outcomes. The 510(k) decision established substantial equivalence for the intended image-processing use; the clearance summary says no clinical studies were necessary to support that determination.

Think of it this way: clearing a thermometer for its intended use does not prove that using it prevents fever. Similarly, a substantial-equivalence determination for an AI coronary plaque image-processing tool does not demonstrate that incorporating its output into clinical decision-making saves lives or prevents heart attacks.

PARAMOUNT tested a nearer-term clinical question: whether a care pathway guided by the software’s plaque analysis changes a 180-day hierarchical endpoint, diagnostic certainty, LDL cholesterol, treatment use, and invasive-referral patterns. The software remains a decision-support tool rather than a replacement for clinician judgment. The trial did not test whether the pathway saves lives.

What PARAMOUNT Proves—and What It Does Not

PARAMOUNT represents a meaningful step forward in cardiovascular research methodology. Unlike traditional software benchmarking studies, it randomized real patients to different diagnostic strategies and measured actual clinical decisions—demonstrating that detailed plaque imaging data changes how doctors prescribe preventive medications. This is a concrete, observable outcome that moves beyond laboratory performance metrics.

However, there is an important distinction between showing that a test influences medical decisions and proving it saves lives. More certain diagnoses and higher rates of preventive medication initiation are promising early signals along the pathway to better health outcomes. They suggest the strategy might prevent heart attacks, reduce hospitalizations, or improve survival—but the trial does not yet demonstrate these benefits. That remains a crucial unknown.

Several design features warrant consideration. The study used an open-label approach, meaning clinicians and patients knew their group assignments. This transparency may have influenced behavior in ways beyond simply providing better information. Additionally, full peer-reviewed publication will be essential for evaluating missing data, subgroup findings, adverse events, and whether patients adhered to prescribed protocols as intended.

The population studied was also quite specific: symptomatic outpatients with suspected coronary artery disease, elevated cholesterol, and no prior heart disease at 21 experienced medical centers. These results may not extend to population-wide screening efforts, patients presenting with acute chest pain, or healthcare systems with fewer specialized resources.

Real-world implementation raises practical questions as well. Advanced CT imaging requires specialized equipment, trained technicians, contrast agents, and radiation exposure. Whether this approach integrates smoothly into diverse clinical workflows and proves cost-effective across different healthcare settings remains unstudied. These gaps do not diminish PARAMOUNT’s contributions—they simply define what questions remain to be answered.

The Path Forward: Outcomes Still Matter

The PARAMOUNT trial adds randomized preliminary evidence about how a plaque-guided pathway can change coronary-disease diagnosis and management. It reported greater diagnostic certainty, different lipid-lowering treatment use, and lower LDL cholesterol at 180 days. These are intermediate pathway results, not proof of improved long-term health.

However, a critical question remains unanswered: Will the changed pathway improve patient-centered and cardiovascular outcomes over longer follow-up? PARAMOUNT reported treatment use and LDL differences, not sustained medication adherence, symptom relief, strokes, heart attacks, or deaths. Those outcomes require dedicated follow-up rather than inference from the preliminary presentation.

The next phase of investigation must be comprehensive and rigorous. Researchers need to track multiple outcomes over extended periods: adherence to preventive medications, symptom resolution, major adverse cardiovascular events, hospitalizations, healthcare costs, quality of life, and medication side effects. These real-world metrics will paint the complete picture of whether plaque-guided care truly transforms patient health.

Before broader adoption can be justified from these results, researchers and health systems will need to ask how the pathway performs in diverse populations and settings, which patients benefit, and whether the added imaging and software costs are warranted. The path forward demands both innovation and accountability.

Stay ahead of the curve! Subscribe for more insights on the latest breakthroughs and innovations.