CT leads imaging field in plaque assessment

February 7, 2005

CT is moving beyond detection and quantification of coronary artery calcium to grading of coronary stenoses, identifying not only vulnerable plaques but, more important, vulnerable patients. Yet its ultimate role in predicting risk of cardiac events remains unclear.

CT is moving beyond detection and quantification of coronary artery calcium to grading of coronary stenoses, identifying not only vulnerable plaques but, more important, vulnerable patients. Yet its ultimate role in predicting risk of cardiac events remains unclear.

There's no question that coronary CT angiography will cut a swath through the estimated 3.5 million conventional coronary angiography exams performed annually. Only about one-third of these invasive, expensive procedures are undertaken as part of an intervention, while the remainder are performed to verify the presence and determine the degree of coronary artery disease.

A growing number of cardiac imagers agree that CT does a better job. Unlike coronary angiography, which assesses the lumen, CT exploits its cross-sectional capability to evaluate the vessel wall. It is here that early-stage stenosis, considered the culprit behind acute coronary syndromes, hides. Merely documenting lumen dimensions with conventional angiography fails to provide a complete portrait of plaque accumulation (Radiology 2004;232:7-17).

"Acute coronary events are triggered by rupture of rupture-prone plaques in the coronary arteries," said Dr. U. Joseph Schoepf, an associate professor of radiology and director of CT research and development at the Medical University of South Carolina in Charleston. "The importance of detecting and characterizing plaque is to prospectively identify patients who are prone to plaque rupture. It would be an incredible achievement to be able to actually prevent these events."

While radiologists have known for years that MSCT's greatest application is in the heart, only recently has the added value of the latest generation of multislice machines been documented in the peer-reviewed journals. Several studies presented at major meetings in the past year offered comparisons among CT and coronary angiography, intravascular ultrasound, and MR. They found that contrast-enhanced 16-slice CT accurately, and noninvasively, detects calcified and noncalcified coronary plaques.

A consensus is emerging that each modality has a role to play, depending on whether the goal is prevention, intervention, or therapy management.

PREVENTIVE ROLE

Prior to the introduction of 16-slice CT and ECG-gated coronary CT angiographic techniques, electron-beam CT devices were lauded for their ability to score coronary calcification. Many sites now performing coronary CTA to evaluate patients for atherosclerosis have discontinued calcium scoring in favor of the new, more comprehensive study.

Coronary calcium scoring leaves too much information on the table, according to Dr. David Dowe, medical director of Atlantic Imaging in Galloway, NJ.

"Coronary CTA is so much more than calcium scores," he said. "It shows the hard plaque, the soft plaque, and blockages. But calcium scoring has a big head of steam, and people confuse it with CTA, which is the reason I don't even offer calcium scoring."

CTA's negative predictive value is stellar, but its positive predictive value may need a little work. In a study presented at the 2004 American Heart Association meeting in New Orleans, Dr. Stephan Achenbach, an assistant professor of radiology at the University of Erlangen in Germany, used 16-slice CT to evaluate coronary artery segments in 50 patients scheduled for conventional angiography.

Sixty-five stenoses in 664 segments were present; in 5% of these segments, CT could not evaluate due to motion artifact or calcification. In the remaining segments, CT demonstrated 93% sensitivity, 95% specificity, negative predictive value of 99%, and positive predictive value of 68%.

Two additional studies based on 16-slice CT report 79% and 80% positive predictive values in evaluating stenoses of 50% or less (Circulation 2002;106[16]:2051-2054; Circulation 2003;107[5]:664-666).

Those results mesh with the recommendations of a research group headed by Dr. Axel Kuettner, a radiologist with Eberhard-Karls Uni-

versity in Tuebingen, Germany. Kuettner's study, published in the Journal of the American College of Cardiology, demonstrated CTA sensitivity and specificity for detecting coronary stenoses in the high 90% range. But it noted that 16-slice CT is at its best in evaluating low to moderate amounts of coronary artery calcium (JACC 2004;44[6]:1230-1237).

That's an opinion shared by several radiologists now routinely using coronary CTA for evaluating asymptomatic patients-the group expected to benefit most from the technique.

"Patients with advanced atherosclerotic disease, including heavily calcified vessels, have vessels that look like coral because of all the calcium," said Dr. Stephen Koch, medical director of Imaging Heart, a private practice in New York City. "This can lead to the blooming artifact phenomenon, where calcium looks bigger than it actually is in the anatomy."

The advent of 64-slice CT is expected to limit the artifact with thinner slices and improved resolution, but Koch considers asymptomatic patients the most appropriate for evaluation with coronary CTA. Dr. Melvin Clouse, vice chair of radiology and director of radiology research at Beth Israel Deaconess Medical Center in Boston, agrees.

"Coronary CTA will obviate the need for a lot of cardiac catheterizations, but I think the payoff is going to be in prevention," he said.

COMPARATIVE MODALITIES

Intravascular ultrasound has been called the gold standard for detecting soft plaque, which is the type most likely to rupture. But its invasiveness makes it an inappropriate tool for asymptomatic patient evaluation.

Because of IVUS, however, radiologists have a better understanding of vulnerable plaque characteristics: large lipid pool, thin fibrous cap, and an intense inflammatory response.

"The ability to detect plaque vulnerability and also assess plaque burden is important," said Dr. Paul Schoenhagen, a staff physician in the departments of radiology and cardiovascular medicine at the Cleveland Clinic. "We know from IVUS that these lesions develop early and progress over time. By the time a patient shows up in the cath lab with a coronary event, often there are large amounts of plaque in other arteries that might look normal on an angiogram. We can now quantify that plaque burden with IVUS."

Schoenhagen and colleagues scored CT against IVUS in a 2003 study, finding that the tomographic technique can accurately identify mildly stenotic coronary atherosclerosis and also provide assessment of morphology and arterial remodeling. The team stopped short of a full endorsement, however, noting that CT's ability to identify calcified and noncalcified plaque components is not sufficient to predict clinical events (Coronary Artery Disease 2003;14(6):459-462).

If the goal is simply to identify the best noninvasive screening tool, IVUS is out of the running. That leaves CT and MR, both of which have advantages. CT has higher resolution and can image the coronary arteries better than MR, Schoenhagen said. But MR has better tissue characterization, as has been shown in the carotid arteries.

Factor in 64-detector functionality to CT's existing speed and spatial resolution, and it becomes possible to investigate the entire coronary artery tree.

"With 64-slice CT, using an ECG-gated scan, we are in the range where it takes about 10 seconds to scan the entire heart," Schoepf said.

CT's weakness lies in its limited contrast resolution for differentiating types of tissue, which affects its ability to characterize plaque. It's an inherent limitation, one that even the improved spatial resolution of 64-slice CT will only partly address.

"CT does a very good job of differentiating between calcified and noncalcified lesions," he said. "But it cannot reliably differentiate between lipid-rich lesions and fibrous tissue within a coronary artery lesion. That's a problem."

By reducing volume averaging artifacts, it may be possible to improve measurement of different tissue attenuations, according to Schoepf.

FORMING A STRATEGY

Rather than pitting one modality against another for plaque detection, Schoepf favors a coordinated approach.

"With MR, CT, and IVUS, you are talking about three completely different entities with their own sets of advantages and disadvantages," he said. "We should be thinking of strategies for using them in a complementary way."

IVUS is best for evaluating specific lesions identified by invasive catheter angiography or by noninvasive tests. Use of MR as a noninvasive screening tool means accepting a host of trade-offs: speed for spatial resolution, or vice versa. Yet MR's superior contrast resolution makes it an important tool for evaluating plaque composition, Schoepf said.

"We can use CT for detecting lesions and then look into different modalities, such as MR, to differentiate between tissue types and

look at lesion composition," he said. "So a noninvasive approach would begin with CT for detection and continue to MR for lesion differentiation. If the lesion is significant enough to warrant therapy, you can perform IVUS prior to stenting the lesions or performing another intervention."

Schoenhagen agrees that a multipronged approach that is guided by clinical indications is appropriate. The method should be different, however, depending on whether the goal is to prevent disease or to determine an appropriate intervention strategy for specific lesion types, he added.

"Although there are groups working on CT and MR, CT is clearly leading the field," he said. "Most people do not anticipate that MR will contribute significantly in the next few years, while CT is expected to do so."

Combining the capabilities of 32-slice CT with a stress test (most likely pharmacologically induced) might result in a tool that provides all physicians need to know about a patient's coronary artery disease.

Coronary CTA may also be able to detect relevant coronary artery obstructions in addition to necrosis and hibernation, according to Dr. Myron Weisfeldt, chair of the department of medicine at Johns Hopkins University. And it may offer quantitative measures of flow in the arteries and heart muscle.

THERAPY MANAGEMENT

CT may also play a role in monitoring response to statin therapy, particularly in light of the potential risks that lipid-shrinking drugs pose for some patients (Radiology 2004;232:18-37).

"We want to make sure that patients who receive statin therapy are actually responding to the treatment, so that we're neither wasting money nor increasing the risk of side effects in patients who do not benefit from it," Schoepf said. "In theory, CT is a very elegant, noninvasive way of doing that."

One study, based on coronary calcium scores acquired with EBCT to evaluate response to statin therapy, offered promising results. Researchers were able to track stability and, in some cases, regression of plaque burden (NEJM 1998; 329(27):1972-1978). But reproducibility of such results with EBCT is a problem, Schoepf said.

Variability between two tests even only five minutes apart can range as high as 30%. Use of retrospective ECG-gating combined with the thinner sections of 16-slice CT can reduce such variability and provide a more reliable measurement of calcified plaque burden, Schoepf said.

Calcified plaque is only the tip of the iceberg, however. CTA is emerging as a tool to evaluate the entire plaque burden: fibrous tissue, lipid tissue, and the pathophysiology of vascular remodeling that occurs in noncalcified regions of the vessels. It's this tissue that responds more readily to statin therapy.

"What we hope is that by looking at the noncalcified disease component with CTA, we'll get a faster and more reliable response rate assessment," Schoepf said.

VULNERABLE PATIENTS

While CT may indeed prove important as an endpoint application to evaluate treatment response, its use as a noninvasive screening tool is stirring up much more interest. Its promise has opened the door to a new approach in patient workup, Clouse said.

"We can see amazing things with 16-slice CT," he said. "It's important that we continue to address early diagnosis of atherosclerosis by looking for soft plaque. We spend more than one-third of our trillion-dollar healthcare economy treating endstage cardiovascular disease. It's a figure that boggles the mind."

Research continues on identifying and quantifying plaque, as well as documenting arterial remodeling, plaque progression, and regression. But the discovery of one magic bullet is unlikely.

"It's a misconception that any imaging test can look at the coronary tree of a patient, see a lesion that looks rupture-prone, and predict that this lesion is going to rupture," Schoepf said. "It's part of the hype around this modality."

To know CT's ultimate role in plaque characterization is to understand the complexity of atherosclerotic disease. Patients with noncalcified plaques in coronary arteries are also likely to harbor disease in the carotids or ascending aorta.

"Atherosclerosis happens everywhere there is a vessel in the body," he said. "Simply looking for vulnerable plaque isn't leading us anywhere. Better to identify vulnerable patients who have lesions prone to rupture, using all of the tools we have, and determine who is at risk and who may benefit from specific therapies."