Biomarker Derived from Standard CT Could Predict Cardiac Death

Liam Davenport

August 30, 2018

MUNICH — Analysing the alteration in fat content in coronary arteries in standard coronary computed tomography angiography (CTA) images could predict the likelihood that an individual with apparently healthy arteries will go on to have a fatal heart attack, say UK researchers.

Studying coronary CTA images taken in nearly 4000 patients in Germany and the USA, researchers from the University of Oxford applied a novel biomarker, termed the Fat Attenuation Index (FAI), that measures changes in the composition of coronary artery walls to identify areas with inflamed plaques.

Presenting their findings at the ESC Congress, which were published simultaneously in The Lancet, they showed that a high FAI index in the right coronary artery was associated with an at least 5-fold increased risk of future cardiac mortality.

Residual Cardiovascular Risk

Lead researcher Charalambos Antoniades, PhD, Radcliffe Department of Medicine, University of Oxford, said in a news release: "This new technology may prove transformative for primary and secondary prevention.

"For the first time, we have a set of biomarkers, derived from a routine test that is already used in everyday clinical practice, that measures what we call the 'residual cardiovascular risk', currently missed by all risk scores and non-invasive tests."

He believes that knowing this information could allow clinicians to intervene early enough to prevent heart attacks.

"I expect these biomarkers to become an essential part of standard CT coronary angiography reporting in the coming years," he said.

Co-author Dr Stephan Achenbach, professor of cardiology at the University of Erlangen, added: "This study highlights an entirely new way in which patients at risk for myocardial infarction or heart attack could possibly be identified very early on.

"We very much need such novel approaches since a large part of all heart attacks occur without any prior notice. This study and its results may add to our armamentarium to prevent heart attacks, which makes the results so exciting from a scientific and clinical point of view."

Possible Game Changer

Metin Avkiran, PhD, associate medical director at the British Heart Foundation, which funded the research, noted: "Most heart scans are good at spotting blockages caused by large plaques, but not the smaller, high-risk plaques that are likely to rupture and cause a heart attack.

"This new technique could be a game changer – allowing doctors to spot those 'ticking time bomb' patients who are most at risk of a heart attack, and getting them on to intensive treatment. This would undoubtedly save lives."

Dr Luigi Badano, Department of Cardiac, Thoracic and Vascular Sciences, University of Padua, Italy, who was not involved in the study, told Medscape News UK that it was a "clever and innovative study".

He believes that the technique will make it into clinical practice, "particularly in patients presenting at the emergency room with chest pain, that are atypical patients with low-to-moderate cardiovascular probability of cardiovascular events".

He pointed out that "most of those patients have apparently normal coronary arteries but they have chest discomfort.

"Sometimes, they have some slight modification on the ECG, so having an additional marker of cardiovascular risk, despite the known markers being normal, it would definitely be beneficial in order to treat more aggressively those patients with preventive measures."

Badano believes, however, that this does not necessarily mean pharmacological intervention.

He said that it is also an opportunity to say to patients "look, despite your coronary arteries at this moment being normal you have an increased risk, so you have to change your lifestyle, exercise more, lower your weight, try to eat in a healthier way.

"I think this could be absolutely of benefit."


Prof Antoniades began his presentation by saying that the CANTOS trial, presented at the ESC Congress in 2017, demonstrated the key role that inflammation plays in cardiovascular events.

"The problem is we don't really know who has inflamed coronary arteries and we don't know who has inflamed plaques in order to target these particular subgroups of patients," he said, noting that treating patients based on circulating biomarkers "is probably too expensive" for healthcare systems.

Previous research by the team showed, however, that when coronary arteries become inflamed "they secrete a number of inflammatory molecules to the perivascular space and the vascular adipocytes then sense these signals and become smaller".

This, Prof Antoniades explained, is due to the inflammation inhibiting adipogenesis and inducing lipolysis, leading to oedema around the coronary arteries.

"Essentially, we have a shift in the composition of the vascular fat in the presence of inflammation inside the coronary arteries from fat content to more watery content, and this can be used as a thermometer to pick up inflammation inside the coronary arteries."

The team then used standard CT images to examine the perivascular space, using information currently suppressed to remove artefacts from normal images, to reveal the fat content in the perivascular space.

From this they developed the FAI, with a high score indicating an apparently healthy but inflamed artery and a low score a non-inflamed vessel.


To determine whether the FAI predicts cardiovascular risk, they conducted the CRISP-CT study, involving 1872 patients from Erlangen, Germany, who acted as the derivation cohort, and 2040 individuals from the Cleveland Clinic, in the US, who served as the validation cohort.

All participants had undergone diagnostic coronary CTA. The median age in the derivation cohort was 62 years, while that in the validation cohort was 53 years. The median follow-up was 72 months and 54 months, respectively.

During follow-up, there were 114 deaths in the derivation cohort, of which 26 were confirmed to be of cardiac causes. In the validation cohort, there were 85 deaths, of which 48 were confirmed cardiac deaths.

The researchers found that, in both cohorts, high perivascular FAI levels around the proximal right coronary artery and left anterior descending artery, although not around the left circumflex artery, were predictive of all-cause and cardiac mortality, and were highly correlated.

Using perivascular FAI at the right coronary artery as a representative marker of coronary inflammation, the team determined that a high score predicted cardiac mortality at a hazard ratio of 2.15 (p=0.017) in the derivation cohort and 2.06 (p<0.0001) in the validation cohort.

They calculated that the optimum cut-off for perivascular FAI was -70.1 Houndfield units or higher, yielding a hazard ratio for cardiac mortality of 9.04 (p<0.0001) and for all-cause mortality of 2.55 (p<0.0001) in the derivation cohort.

Applying this cut off to the validation cohort, the hazard ratio for cardiac mortality was 5.62 (p<0.0001), while that for all-cause mortality was 3.69 (p<0.0001).

Taking into account age, sex, hypertension, hypercholesterolaemia, diabetes, smoker status, epicardial fat volume, modified Duke CAD index and high-risk plaque features on coronary CTA, FAI predicted cardiac death in an area under the curve of 0.049 (p=0.0054) in the derivation cohort and 0.075 (p=0.0069) in the validation cohort.

This was seen across all subgroups analysed, including patients with and without obstructive coronary artery disease.

Interestingly, the team found that the risk of cardiac mortality in the derivation cohort was modified by initiation of aspirin and/or statin therapy after the coronary CTA, with the adjusted hazard ratio no longer being significant.

'Striking Prognostic Value'

Prof Antoniades concluded that FAI "has a striking prognostic value for cardiac death...over and above current scores and state-of-the-art interpretation of coronary CTA", that is potentially modifiable by treatment.

While he said that the technique "is applicable to any standard coronary CTA from any scanner", he warned that it needs appropriate correction for obesity, scanner type, scan settings and other technical and biological factors, and so validated image analysis tools would be needed.

In the post-presentation discussion, Prof Antoniades was asked whether the test will be lengthy and costly, thus limiting its applicability to the general population.

He replied that, rather than performing CT scans in the whole population, they propose to incorporate the analysis into the standard cardiac CT scans, "so this should be a part, in the future, of the standard reporting of coronary CT angiography".

This, he said, will "make better use of the information already there".

Consequently, any extra costs involved in identifying the 10% of patients who have apparently no significant disease but highly inflamed plaques and then treating then with aspirin and/or statin would be offset with the savings to the healthcare system down the line.

In an accompanying Comment, Dr Amir Mahabadi and Dr Tienush Rassaf, West German Heart and Vascular Centre Essen, University Hospital Essen, Germany, say that, currently, the technique is not ready to be rolled out.

They say that, "measurement of the pericoronary FAI is relatively time-consuming and currently restricted to centres with dedicated expertise in this area.

"This complexity necessitates automated software programs, enabling quick and reliable quantification of the perivascular FAI to allow its implementation in everyday clinical routine."

Drs Mahabadi and Rassaf add: "If these obstacles are resolved, characterisation of coronary inflammation in addition to detection of atherosclerosis could have a role in preventive cardiology."

To those ends, the University of Oxford has formed a spinout company to develop the FAI as a software-as-a-service offering.

The study was funded by the British Heart Foundation, and the National Institute of Health Research Oxford Biomedical Research Centre.

Antoniades and other authors are founders and shareholders of Caristo Diagnostics, a CT image analysis company, report grants from the British Heart Foundation during the conduct of the study and research grants from the National Insitute of Health Research Oxford Biomedical Research Centre. Antoniades also reports a consultancy fee from Mitsubishi Tanabe. Stefan Neubauer reports grants outside the submitted work from Boehringer Ingelheim and the US National Institutes of Health, and personal fees from Perspectum Diagnostics.

ESC Congress 2018. Abstract 5886. Presented August 28.

The Lancet 2018. doi: 10.1016/ S0140-6736(18)31114-0. Full text


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.