What Your Heart Scan Really Shows: Predicting Dangerous Plaque in a Diverse Patient Population

Table of Contents

• Key Points
• Why This Research Matters
• Background: What We Already Knew
• How the Study Was Conducted
• Who Was in the Study
• Key Findings: High-Risk Plaque Features
• Important Differences Among Racial and Ethnic Groups
• Who Was More Likely to Have High-Risk Plaque?
• Patient Outcomes During Follow-Up
• What This Means for Patients
• Study Limitations
• Recommendations for Patients
• Frequently Asked Questions
• Source Information

Key Points

• In a study of 364 patients with stable chest pain, 31% had elevated low-attenuation plaque (LAP) even though only about 8% had obstructive blockages.
• Independent predictors of elevated LAP were older age, male sex, hypertension, and hyperlipidemia, after adjusting for race, diabetes, smoking, statin use, and Lp(a).
• Patients with elevated LAP were significantly more likely to have heart attack, revascularization, or cardiovascular death during a median 9.5-month follow-up.
• Lp(a) correlated with LAP burden specifically in Hispanic patients; Lp(a) is inherited and not significantly lowered by standard statin therapy.

Why This Research Matters

Heart disease remains the leading cause of death in the United States, but not all patients at risk are the same. Modern cardiology has shifted from simply looking at whether an artery is blocked to identifying "vulnerable" plaques — those that are more likely to rupture and cause a heart attack. This study, conducted at Montefiore Medical Center in the Bronx, New York, focused on a diverse patient population that reflects the real-world mix of racial and ethnic groups often underrepresented in cardiovascular research.

The research used advanced CT scan technology to look beyond blockages. It examined the composition of plaque inside coronary arteries, as well as inflammation around the arteries and fat surrounding the heart. These features can signal future heart attacks even before a blockage becomes severe.

What makes this study particularly valuable is its focus on a diverse group — 60% Hispanic, 21% non-Hispanic Black, 6% non-Hispanic White, and 4% non-Hispanic Asian — a population profile that is common in urban America but rare in many large clinical trials.

Background: What We Already Knew

Coronary computed tomography angiography (CCTA) is a specialized CT scan that produces detailed images of the heart and its blood vessels. Major guidelines have recently endorsed its use because total plaque burden is now recognized as the strongest predictor of future cardiovascular events. Recent advances in software have made it possible to measure plaque features automatically, moving from subjective visual assessment to rapid, semi-automated quantitative analysis.

Data from the landmark SCOT-HEART trial suggested that a high low-attenuation plaque (LAP) burden and elevated pericoronary adipose tissue (PCAT) attenuation could be the strongest predictors of subsequent myocardial infarction (heart attack), even better than the degree of vessel narrowing or traditional cardiovascular risk calculators. Based on statistical methods (the Youden index of ROC curves), researchers established two key cutoff points for identifying high-risk patients:

• LAP burden ≥4% — low-attenuation non-calcified plaque (the "soft," dangerous plaque prone to rupture) makes up at least 4% of the vessel volume
• RCA PCAT attenuation ≥-70.5 HU — increased inflammation in the fat surrounding the right coronary artery

However, most prior research on quantitative plaque phenotyping was conducted in predominantly White populations. Minority patients, who make up an increasing proportion of the US population, have been significantly understudied. This is a critical gap, because cardiovascular mortality is higher in non-Hispanic Black people, and Hispanic and non-Hispanic Black people tend to have poorer cardiometabolic risk profiles at every level of income and insurance status.

Additionally, risk factors like lipoprotein(a) — an inherited risk factor for heart disease — and epicardial adipose tissue (EAT, the fat surrounding the heart) may have different effects across racial and ethnic groups. This study was designed to fill that gap.

How the Study Was Conducted

Study Design and Patient Selection

Researchers conducted a retrospective analysis using the CT registry at Montefiore Healthcare Network. They identified all patients with stable chest pain or angina equivalent who underwent CCTA between June 2016 and March 2022 and had a complete cardiometabolic panel — including hemoglobin A1c (HbA1c, a measure of blood sugar control) and a lipid panel within 90 days before the scan, plus a lipoprotein(a) [Lp(a)] measurement at any point.

The flow of patient selection was as follows:

1. 380 patients were initially identified as having undergone CCTA with a complete cardiometabolic panel including Lp(a)
2. 13 were excluded due to prior myocardial infarction (heart attack), percutaneous coronary intervention (PCI, a stent procedure), or coronary artery bypass grafting (CABG)
3. 3 were excluded due to acute chest pain
4. 364 patients remained in the final study group

Race and ethnicity were self-identified at the time of registration. The researchers categorized patients as Hispanic, non-Hispanic Black (NHB), non-Hispanic White (NHW), non-Hispanic Asian (NHA), or declined/unknown. The investigators chose this approach because self-reported race among Hispanics is notoriously unreliable, and being Hispanic is technically an ethnicity rather than a race.

Imaging and Plaque Analysis

All patients underwent a non-contrast, ECG-gated CT scan for coronary artery calcium (CAC) scoring first, followed by the CCTA. Scans were performed on 64-slice Philips IQon and 64-slice GE VCT scanners, using 80–100 mL of Isovue-370 contrast material. Metoprolol was given to control heart rate, and sublingual nitroglycerin was administered before image acquisition.

Coronary artery calcium was scored using the standard Agatston method. Plaque analysis was performed with semiautomated software called AutoPlaque version 2.5 (Cedars-Sinai Medical Center, Los Angeles, CA) in all patients with a CAD-RADS score greater than 0. Two readers with level 3 certification equivalent experience, blinded to patient characteristics, performed the measurements. The software measured total plaque, non-calcified plaque (NCP), low-attenuation non-calcified plaque (LAP, defined as plaque with density below 30 Hounsfield Units — a marker of necrotic core), and calcified plaque (CP).

Plaque burdens were calculated by dividing plaque volumes by the coronary vessel volume and multiplying by 100. Pericoronary adipose tissue (PCAT) — a measure of vascular inflammation — was analyzed in the proximal right coronary artery (RCA, 10–50 mm from the ostium), left anterior descending artery (LAD, 0–40 mm), and left circumflex (LCx, 0–40 mm). PCAT attenuation was measured as the average density of all adipose tissue voxels (range -190 HU to -30 HU) within a 3 mm radial distance from the vessel wall.

Epicardial adipose tissue (EAT) — the fat surrounding the heart — was measured on the CAC scans using QFat software, a convolutional deep-learning tool for fully automated quantification. Both AutoPlaque and QFat have been previously validated for accuracy and reproducibility.

Laboratory Measurements

All blood tests were performed using standardized hospital laboratory procedures. LDL cholesterol was calculated using the Friedewald equation. Lipoprotein(a) was measured using an immunoturbidimetric assay from Quest Diagnostics, reported in molar concentration (nmol/L). The 10-year risk of atherosclerotic cardiovascular disease (ASCVD) was calculated using the pooled cohort equations risk calculator. When multiple values were available, the measurement closest to the CCTA date was used.

Clinical Outcome Tracking

The primary clinical outcome was a composite of type 1 myocardial infarction (heart attack), revascularization (stent or bypass surgery), or cardiovascular death. Researchers reviewed charts while blinded to CCTA findings, and linked records from other institutions were reviewed whenever available. Myocardial infarction was defined as elevation of troponin with ECG changes and symptoms. The median follow-up time was 288 days (interquartile range 169.75 to 403.00 days), with data censored on May 1st, 2022.

Statistical Analysis

The researchers used standard statistical tests appropriate for the data — Wilcoxon or Kruskal-Wallis tests for continuous data, and chi-square or Fisher's exact tests for categorical data. Kaplan-Meier curves were created based on LAP ≥4% versus <4%, RCA PCAT ≥-70.5 HU versus below that threshold, and CAC ≥100 versus <100, with log-rank tests to compare groups. The relationship between plaque characteristics and Lp(a) was estimated using simple linear regression, and independent predictors for LAP ≥4% and RCA PCAT ≥-70.5 HU were examined using multivariable logistic regression. The following variables were entered into the model: age, male sex, non-Hispanic Black race, diabetes, hypertension, hyperlipidemia, smoking history, statin use, and Lp(a). A p-value below 0.05 was considered statistically significant.

Who Was in the Study

The study included 364 consecutive patients with a median age of 56 years (interquartile range 48 to 63 years). Notably, 64% were female — a higher proportion of women than many cardiac imaging studies. The demographic and clinical profile of the group was:

• Race/ethnicity: Hispanic 60%, non-Hispanic Black 21%, non-Hispanic White 6%, non-Hispanic Asian 4%, and 9% declined or unknown
• Median BMI: 29.8 kg/m² (in the overweight range)
• Diabetes: 16.2% of patients
• Hypertension: 47.3% of patients
• Hyperlipidemia: 39.6% of patients
• Smoking history: 69.3% never smoked, 13.3% past smokers, 17.5% current smokers

Laboratory values showed a median total cholesterol of 183 mg/dL, LDL cholesterol of 105 mg/dL, HDL cholesterol of 49 mg/dL, and HbA1c of 5.6%. The median coronary artery calcium (CAC) score was 0 (interquartile range 0 to 32), and the median 10-year ASCVD risk was 5.0%. About 36% of patients were taking statins at baseline.

Importantly, the researchers confirmed that the included patients were representative — there was no significant difference in 10-year ASCVD risk between the included patients and the excluded patients who lacked a full cardiometabolic panel (5.00% vs. 5.30%, p=0.331).

Key Findings: High-Risk Plaque Features

Despite the fact that most patients had no or minimal narrowing of their arteries, a substantial number had high-risk plaque features on their CT scans:

• Elevated LAP burden (≥4%): present in 111 patients (30.5%)
• Elevated RCA PCAT attenuation (≥-70.5 HU): present in 65 patients (17.9%)
• Both high-risk features together: present in only 14 patients (3.8%)

In terms of actual blockages, 51.1% of patients had no stenosis, 41.8% had non-obstructive stenosis, 4.9% had borderline stenosis, and only 2.2% had obstructive stenosis (roughly 8% total when borderline and obstructive are combined). This is a striking finding: many patients with minimal blockages still harbored dangerous plaque features — which is exactly why looking beyond stenosis alone matters.

The median LAP burden for the entire group was 0.00% (interquartile range 0.00 to 5.13%). Total plaque burden was 0.00% (0.00 to 41.24%), and total plaque volume was 0.00 mm³ (0.00 to 294.50 mm³). The median EAT volume was 88 mL (63 to 117 mL).

Important Differences Among Racial and Ethnic Groups

One of the study's most important contributions is its detailed breakdown of plaque characteristics by race and ethnicity. There were significant differences in plaque characteristics among racial/ethnic groups (p<0.001).

Low-Attentuation Plaque (LAP) by Group

The prevalence of elevated LAP (≥4%) differed dramatically across groups (p=0.003):

• Non-Hispanic White: 52.4% had elevated LAP
• Non-Hispanic Asian: 50.0% had elevated LAP
• Hispanic: 32.3% had elevated LAP
• Non-Hispanic Black: only 17.1% had elevated LAP

Median LAP burden also differed significantly (p=0.002), with the highest in non-Hispanic White patients (5.38%) and non-Hispanic Asian patients (2.71%), followed by Hispanic patients (0.00%) and non-Hispanic Black patients (0.00%). Total plaque burden was similarly highest in non-Hispanic White patients (48.83%) and non-Hispanic Asian patients (17.88%) compared to Hispanic (0.00%) and non-Hispanic Black patients (0.00%).

Lipoprotein(a) — A Key Inherited Risk Factor

Lipoprotein(a), or Lp(a), is a genetically determined risk factor for heart disease that is not routinely measured in most patients. The study found major differences in Lp(a) levels across racial groups (p<0.001):

• Non-Hispanic Black: median 120.50 nmol/L — by far the highest
• Hispanic: median 52.00 nmol/L
• Non-Hispanic White: median 31.00 nmol/L
• Non-Hispanic Asian: median 22.50 nmol/L

Interestingly, even though non-Hispanic Black patients had the highest Lp(a) levels, they had the lowest prevalence of elevated LAP. However, the study found that Lp(a) correlated with LAP burden specifically in Hispanic patients — suggesting that the relationship between this inherited risk factor and dangerous plaque may vary by ethnic background.

Epicardial Adipose Tissue (EAT)

EAT volume also differed significantly across groups (p<0.001). Non-Hispanic White patients had the highest median EAT volume (124 mL), followed by Hispanic patients (94 mL), non-Hispanic Black patients (65.5 mL), and non-Hispanic Asian patients (59.5 mL).

Other Notable Differences

Non-Hispanic White patients had significantly higher CAC scores (median 56 vs. 0 in other groups) and higher CAD-RADS scores (median 2.0 vs. 0–0.5). Hypertension was most common in non-Hispanic White patients (80%) and least common in non-Hispanic Asian patients (25%). Non-Hispanic White patients also had the highest 10-year ASCVD risk (median 11.0%) and the highest systolic blood pressure (median 130.5 mmHg).

Who Was More Likely to Have High-Risk Plaque?

The researchers compared patients with elevated LAP (≥4%) to those without it. The differences were striking across nearly every major risk factor (all comparisons in Table 2):

• Age: 59 years in the elevated LAP group vs. 55 years in the lower LAP group (p<0.001)
• Male sex: 46.8% vs. 32.0% (p=0.01)
• Diabetes: 25.0% vs. 12.3% (p=0.005)
• Hypertension: 69.4% vs. 37.4% (p<0.001)
• Hyperlipidemia: 58.3% vs. 31.3% (p<0.001)
• Smoking: 24.3% current smokers in the high-LAP group vs. 14.4% in the low-LAP group (p=0.023)
• CAC score: 40 vs. 0 (p<0.001)
• CAD-RADS score: 2.0 vs. 0 (p<0.001)

Statin use at baseline trended higher in the elevated LAP group (44.2% vs. 32.9%), though this did not quite reach statistical significance (p=0.059).

After multivariable logistic regression accounting for all these factors simultaneously, the independent predictors of elevated LAP were age, male sex, hypertension, and hyperlipidemia. These findings held even after adjusting for race, diabetes, smoking, statin use, and Lp(a) levels.

Patient Outcomes During Follow-Up

During the median follow-up of approximately 9.5 months (288 days), the following clinical events occurred:

• Invasive coronary angiography: performed in 15 patients (4.1%)
• Percutaneous coronary intervention (stent): 11 patients (3.0%)
• Coronary artery bypass grafting (CABG): 2 patients (0.6%)
• Myocardial infarction (heart attack): 2 patients (0.6%)
• Cardiovascular death: 0 patients
• All-cause death: 1 patient (0.3%)

There was a significant difference in revascularization rates across racial/ethnic groups (p=0.04), with non-Hispanic White patients having the highest rates of PCI (12.5%) and CABG (4.8%).

The key outcome finding was that patients with elevated LAP were significantly more likely to develop the primary clinical outcome (composite of heart attack, revascularization, or cardiovascular death) with p<0.001. In contrast, patients with elevated PCAT alone were not (p=0.797). This suggests that the presence of dangerous plaque itself — more than the surrounding vascular inflammation — was the stronger driver of clinical events in this patient population.

What This Means for Patients

This study reinforces a paradigm shift in cardiology: the degree of arterial blockage is only part of the story. Even patients with no obstructive stenosis can harbor high-risk plaque features that predispose them to future heart attacks. For the 31% of patients in this study who had elevated LAP, a standard stress test or even a traditional CCTA reading that focused only on blockages might have missed their true risk.

The findings also highlight important racial and ethnic considerations. Non-Hispanic Black patients had the highest Lp(a) levels yet the lowest prevalence of dangerous plaque, while non-Hispanic White and non-Hispanic Asian patients had the highest LAP prevalence. This suggests that risk factors may operate differently across populations, and that a one-size-fits-all approach to cardiovascular risk assessment may not be appropriate.

For Hispanic patients specifically, the finding that Lp(a) correlated with LAP burden is clinically relevant. Lp(a) is largely genetically determined and not significantly lowered by standard statin therapy. Measuring Lp(a) in Hispanic patients may help identify those with a higher likelihood of dangerous plaque who might benefit from more aggressive preventive therapy or closer imaging follow-up.

Finally, the study confirms that elevated LAP has prognostic significance — it predicted real clinical events (heart attacks, stents, bypass surgery) in this diverse population. This supports the use of quantitative plaque analysis in routine clinical practice for patients undergoing CCTA.

Study Limitations

Like all research, this study has important limitations that patients and clinicians should understand:

• Retrospective design: The study analyzed existing medical records, which can introduce selection bias and limits the ability to establish cause-and-effect relationships.
• Single center: All patients were from Montefiore Healthcare Network in the Bronx, New York, so the results may not generalize to other geographic regions or healthcare settings.
• Small numbers in some subgroups: Non-Hispanic White (21 patients) and non-Hispanic Asian (14 patients) groups were relatively small, which limits the statistical power for subgroup analyses.
• Short follow-up: The median follow-up was only 288 days (about 9.5 months), and the number of hard clinical events (heart attacks, deaths) was small — 2 heart attacks and 0 cardiovascular deaths. The composite outcome was driven largely by revascularization procedures.
• Referral population: All patients were referred for CCTA for stable chest pain, so these findings may not apply to asymptomatic individuals or those with acute symptoms.
• Lp(a) timing: Lp(a) was measured "at any point" rather than at a standardized time relative to the CCTA, which could introduce variability.

Recommendations for Patients

Based on this study's findings, here are key takeaways that patients can discuss with their healthcare providers:

1. Ask about Lp(a) testing. Lipoprotein(a) is an inherited risk factor that was measured in all patients in this study. It's a simple blood test and can provide important information about your genetic risk for heart disease, especially for Hispanic patients where it correlated with dangerous plaque.
2. Know that "no blockage" doesn't mean "no risk." This study found that 31% of patients had high-risk plaque despite most having minimal or no narrowing of their arteries. If you've had a heart CT scan, ask whether plaque composition — not just blockages — was assessed.
3. Control blood pressure and cholesterol aggressively. Hypertension and hyperlipidemia were two of the four independent predictors of dangerous plaque in this study. Keeping these well-controlled is critical.
4. Don't ignore your age and sex. While you can't change being older or male, these factors were independent predictors of high-risk plaque. If you're older — particularly a man — and have additional risk factors, you may need more vigilant screening.
5. Quit smoking. Current smoking was significantly more common in patients with elevated LAP (24.3% vs. 14.4%).
6. Discuss EAT measurement with your doctor. Epicardial adipose tissue surrounding the heart was higher in patients with elevated LAP. This measurement, available from a standard calcium-score CT scan, is another marker of cardiometabolic risk that some clinicians use to guide treatment intensity.

If you have already had a CCTA, ask your doctor whether quantitative plaque analysis was performed. If it wasn't, and you have significant risk factors, it may be worth asking whether re-analysis of your existing scan with semiautomated software could provide additional risk information — no new imaging needed.

Frequently Asked Questions

What are the limitations of this study?

This was a retrospective, single-center study of 364 patients referred for stable chest pain, so findings may not apply to other settings or to asymptomatic people. Follow-up was short (median 288 days), hard events were few, and some racial/ethnic subgroups were small. Lp(a) was measured at any point rather than at a standardized time.

What can I discuss with my doctor based on this research?

You can ask about Lp(a) testing, whether plaque composition was assessed on any heart CT scan you have had, and whether re-analysis of an existing scan with semiautomated software could provide extra risk information without new imaging. Also discuss controlling blood pressure and cholesterol, quitting smoking, and whether measuring epicardial adipose tissue is useful for you.

My heart CT scan showed no blockages but my doctor didn't mention plaque composition — when should I get a second opinion on my heart scan results?

A second opinion is worth considering when a heart CT report addresses only blockages. In this research, 31% of patients with stable chest pain had elevated low-attenuation plaque even though only about 8% had obstructive stenosis, and elevated low-attenuation plaque predicted heart attacks, revascularization, or cardiovascular death. If your scan was not analyzed for plaque composition, ask whether quantitative plaque analysis of the existing scan could add risk information without new imaging. Diagnostic Detectives Network provides independent expert second opinions.

Source Information

Original article title: Cardiometabolic predictors of high-risk CCTA phenotype in a diverse patient population

Authors: Toshiki Kuno, Javier Arce, Michael Fattouh, Sharmila Sarkar, John P Skendelas, Jonathan Daich, Aldo L Schenone, Lili Zhang, Carlos J Rodriguez, Salim S Virani, Piotr J Slomka, Leslee J Shaw, Eric E Williamson, Daniel S Berman, Mario J Garcia, Damini Dey, Leandro Slipczuk

Journal: American Journal of Preventive Cardiology, Volume 15, 2023, Article 100578

Publication date: Received June 4, 2023; revised August 6, 2023; accepted August 13, 2023; available online August 22, 2023

Affiliations: Montefiore Medical Center/Albert Einstein College of Medicine, Bronx, NY; The Aga Khan University, Karachi, Pakistan; The Texas Heart Institute/Baylor College of Medicine, Houston, TX; Cedars-Sinai Medical Center, Los Angeles, CA; Icahn School of Medicine at Mount Sinai, New York, NY; Mayo Clinic, Rochester, MN

DOI: https://doi.org/10.1016/j.ajpc.2023.100578

Funding: This is an open access article under the CC BY-NC-ND license.

Note: This patient-friendly article is based on peer-reviewed research. It is intended for educational purposes and should not replace individualized medical advice from your healthcare provider.