Caristo’s coronary inflammation technology identified cardiovascular risk missed by conventional calcium scoring in a study published in JAMA Cardiology.

Researchers at the University of Oxford have demonstrated that measuring coronary inflammation using artificial intelligence (AI) can improve cardiovascular risk prediction beyond conventional coronary artery calcium (CAC) scoring.

The study, involving 19,110 patients, evaluated Caristo Diagnostics’ fat attenuation index (FAI) technology and found that it could identify people at increased risk of cardiovascular death, including those with no detectable coronary artery calcification.

The research, published in JAMA Cardiology, forms part of the Oxford Risk Factors and Non-Invasive Imaging (ORFAN) study.

During a median follow-up of 6.7 years, 536 participants died from cardiovascular causes and 1,727 experienced a major adverse cardiovascular event, defined as myocardial infarction, new-onset heart failure or cardiovascular death.

Importantly, 42.3% of cardiovascular deaths and 50.8% of major cardiovascular events occurred among patients with CAC scores below 100.

Coronary inflammation identifies risk missed by calcium scoring

CAC scoring measures calcium deposits in the coronary arteries and is commonly used to assess atherosclerotic burden and cardiovascular risk.

However, a low or zero calcium score does not necessarily mean that someone is free from cardiovascular risk. Inflammation can contribute to the development and rupture of non-calcified plaques, which may not be identified through conventional calcium scoring.

Caristo’s FAI-Score measures changes in the fat surrounding coronary arteries, known as perivascular adipose tissue (PVAT), which can indicate underlying inflammation.

The study found that adding FAI-Score measurements to CAC scoring improved cardiovascular risk discrimination. The area under the curve increased from 0.76 using calcium scoring alone to 0.80 with non-contrast FAI measurements and 0.81 with contrast-enhanced coronary CT angiography (CCTA) measurements.

Among patients with a CAC score of zero who subsequently died from cardiovascular causes, the non-contrast FAI score identified 44.9% as having high or very high cardiovascular risk.

Professor Charalambos Antoniades, British Heart Foundation chair of cardiovascular medicine at the University of Oxford and senior author of the study, said: “Unlike calcium and plaque, which reflect the result of atherosclerotic disease, coronary inflammation is the biology driving the disease. Our research shows that coronary inflammation is an independent predictor of cardiovascular death, adding vital prognostic information even in the absence of calcified or non-calcified plaque.”

AI extends inflammation assessment to non-contrast CT

The research also evaluated a new AI algorithm designed to measure coronary inflammation using non-contrast CT scans.

Previously, FAI-Score assessment required contrast-enhanced CCTA imaging. The new approach could allow coronary inflammation to be assessed using non-contrast scans already performed for calcium scoring.

The algorithm was developed using data from 6,818 participants and evaluated in the independent 19,110-patient validation cohort.

Researchers found that the non-contrast FAI score provided prognostic performance comparable to the established CCTA-based measurement. It improved risk discrimination and reclassification when added to calcium scoring.

The approach could potentially extend coronary inflammation assessment to a broader population, including people undergoing routine calcium scoring without the need for additional contrast administration, radiation exposure or scan time.

However, the study population consisted of people referred for clinically indicated cardiac imaging. Further research will be needed to establish the technology’s performance in asymptomatic populations undergoing primary prevention screening.

Caristo technology advances cardiovascular risk assessment

Caristo Diagnostics was founded as a spin-out from the University of Oxford and developed CaRi-Heart, an AI-enabled technology that analyses coronary CT angiography scans to quantify inflammation and assess cardiovascular risk.

FAI-Score is the underlying measurement used to assess coronary inflammation. The technology is authorised by the US Food and Drug Administration (FDA) for clinical use.

The latest findings could broaden its potential application by enabling inflammation assessment from non-contrast CT imaging.

The study authors said the approach could help identify patients who might benefit from more intensive preventive interventions and potentially support the selection of participants for future trials of targeted anti-inflammatory treatments.

Further research is needed to establish how the technology could be incorporated into wider cardiovascular prevention strategies.