Software helps CTA identify lesions, but fails to ID highest risk plaque
Using plaque analysis software, researchers found that CT angiography (CTA) reliably identified and quantified coronary atherosclerotic plaques when compared with intravascular ultrasound. The software, however, had limited reliability to differentiate the composition of noncalcified plaques.
Once plaques are detected, cardiologists may want to differentiate them between calcified (stable) and noncalcified (unstable and more prone to rupture). Intravascular ultrasound (IVUS) is the gold standard to visualize coronary artery walls, but it is expensive and not suited for routine risk stratification.
Junyan Sun, MD, and colleagues from radiology and cardiology at Beijing An Zhen Hospital in China sought to determine how reliably 64-slice CTA could identify, classify, and quantify coronary plaques compared with IVUS. They published the study in the March issue of the American Journal of Roentgenology.
The researchers evaluated 26 patients with CT (Aquilion 64, Toshiba America Medical Systems) and IVUS (Eagle Eye, Volcano). Eleven patients had acute coronary syndrome and 15 patients had stable angina. All patients had abnormal ECG results.
The plaque analysis software (Sure Plaque, Toshiba) displayed different CT density ranges (Hounsfield units)—hypothesized to represent different components of plaque—with various colors.
Of 40 coronary arteries, 247 of 263 segments were imaged and analyzed by both CTA and IVUS. Sixteen segments were ruled out because of poor CT image quality.
Compared with IVUS, CTA enabled correct detection in 86 of 89 segments containing noncalcified plaques, 25 of 27 segments containing calcified plaques, and 118 of 131 segments without atherosclerotic plaques. Results for CTA to detect plaques were:
He attributed the better results to the higher spatial and temporal resolution of 64-slice CT and to the plaque analysis software.
Lipid-rich, fibrous, and calcified plaques displayed different CT densities and can be reliably differentiated, according to the study. However, because of the overlap in the range of Hounsfield units among the various plaque components, it remains difficult to determine the definite composition of individual noncalcified plaques, researchers concluded.
“Granted, improving CT detection of plaque with automated software is progress
toward understanding plaque composition and its natural history,” Sun said. “However, it is many steps away from identifying the high-risk vulnerable plaque.”
David Dowe, MD, chief operating officer and medical director of Atlantic Medical Imaging in Galloway, N.J., has used color-coding software from GE Healthcare for several years. He said it nicely differentiates between calcified and noncalcified plaque, but it also makes the calcium deposits look smaller when compared to the black and white multiplanar reconstructed (MPR) images and the maximum intensity projection (MIP) images.
“What makes a plaque vulnerable is the degree of inflammation, which can only be surmised to be more present the lower in HU density the plaque is,” Dowe said. “The inflammation itself is never imaged.”
Rather than rely on CT or IVUS to identify vulnerable plaque, Dowe said that molecular/PET imaging has a better chance of leading this effort, especially with the use of radiotracers directly targeting proteins expressed by inflammatory cells such as macrophages.
However, CT still has a role to play regarding lesion identification.
“When I can see low density ‘fat’ globules in plaques, I will more strongly urge the patient and their physician to use statins regardless of the patient’s serum cholesterol profile,” Dowe said.
Once plaques are detected, cardiologists may want to differentiate them between calcified (stable) and noncalcified (unstable and more prone to rupture). Intravascular ultrasound (IVUS) is the gold standard to visualize coronary artery walls, but it is expensive and not suited for routine risk stratification.
Junyan Sun, MD, and colleagues from radiology and cardiology at Beijing An Zhen Hospital in China sought to determine how reliably 64-slice CTA could identify, classify, and quantify coronary plaques compared with IVUS. They published the study in the March issue of the American Journal of Roentgenology.
The researchers evaluated 26 patients with CT (Aquilion 64, Toshiba America Medical Systems) and IVUS (Eagle Eye, Volcano). Eleven patients had acute coronary syndrome and 15 patients had stable angina. All patients had abnormal ECG results.
The plaque analysis software (Sure Plaque, Toshiba) displayed different CT density ranges (Hounsfield units)—hypothesized to represent different components of plaque—with various colors.
- Red represented lipid composition (range, –100 to 29 HU)
- Blue represented fibrous composition (range, 30-189 HU)
- Yellow represented calcification (range, 350-1,000 HU)
- Green represented lumen (range, 190-349 HU), and
- Luteous represented vessel wall (range, 30-189 HU), same range as for fibrous composition.
 | |
| Coronary artery cross section of 48-year-old man obtained by intravascular ultrasound indicates fibrous area (hyperechoic, thick arrow) and lipid-rich area (hypoechoic, thin arrow). (Source: All images from American Roentgen Ray Society) | |
 | |
| Cross-sectional view of same noncalcified plaque as above obtained by CT and analyzed with plaque software. Red area indicates lipid-rich area (hypodensity, thin arrow) and blue area indicates fibrous area (hyperdensity, thick straight arrow). Green indicates contrast-enhanced vessel lumen (curved arrow). | |
Compared with IVUS, CTA enabled correct detection in 86 of 89 segments containing noncalcified plaques, 25 of 27 segments containing calcified plaques, and 118 of 131 segments without atherosclerotic plaques. Results for CTA to detect plaques were:
- Sensitivity—97.4%
- Specificity—90.1%
- Positive predictive value—89.7%, and
- Negative predictive value—97.5%.
He attributed the better results to the higher spatial and temporal resolution of 64-slice CT and to the plaque analysis software.
Lipid-rich, fibrous, and calcified plaques displayed different CT densities and can be reliably differentiated, according to the study. However, because of the overlap in the range of Hounsfield units among the various plaque components, it remains difficult to determine the definite composition of individual noncalcified plaques, researchers concluded.
“Granted, improving CT detection of plaque with automated software is progress
toward understanding plaque composition and its natural history,” Sun said. “However, it is many steps away from identifying the high-risk vulnerable plaque.”
 | |
| Cross-sectional view of calcified plaque in 69-year-old man obtained by CT and analyzed with plaque software. Red area indicates lipid-rich or lipid pool area (hypodensity, thin arrow) and blue area indicates fibrous area (hyperdensity). Yellow indicates calcification (thick straight arrow) in this plaque. Green indicates contrast-enhanced vessel lumen (curved arrow) | |
 | |
| Intravascular ultrasound cross section of same patient and plaque as above indicates lipid-rich or lipid pool (hyperechoic and echolucent, thin arrow) and spotty calcification (hyperechoic with shadow, thick arrow). |
“What makes a plaque vulnerable is the degree of inflammation, which can only be surmised to be more present the lower in HU density the plaque is,” Dowe said. “The inflammation itself is never imaged.”
Rather than rely on CT or IVUS to identify vulnerable plaque, Dowe said that molecular/PET imaging has a better chance of leading this effort, especially with the use of radiotracers directly targeting proteins expressed by inflammatory cells such as macrophages.
However, CT still has a role to play regarding lesion identification.
“When I can see low density ‘fat’ globules in plaques, I will more strongly urge the patient and their physician to use statins regardless of the patient’s serum cholesterol profile,” Dowe said.