The Physician and Sportsmedicine
Menubar Home Journal Personal Health Resource Center CME Advertiser Services About Us

Noninvasive Detection of Coronary Artery Disease

Can the New Imaging Techniques Help?

Robert Roos


In Brief: Researchers continually search for new ways to peer noninvasively into the heart for signs of disease. Techniques such as electron-beam and spiral computed tomography, contrast echocardiography, and high-speed magnetic resonance imaging seem to offer great promise for detecting and mapping cardiac lesions, but none is quite ready to become part of the standard tool kit.

Suppose a 42-year-old man sees you because he wants to get clearance to revive his running program after a 10-year interruption. He's a little overweight, and he smoked for 20 years, though he quit 4 years ago. Saying he wants to make sure his heart is OK, he asks if he can have one of those calcium scans he's read about on a health Web site.

Most likely, you first take his history and conduct a physical exam. Then, do you explain that you want him to take a treadmill test because it's a standard and reasonably reliable way to determine if any further investigation of his cardiovascular system is warranted? Or do you arrange for a calcium scan, on the grounds that it seems to be a quick, simple way to rule out any serious concern?

This is one example of the kinds of dilemmas that can arise in an era when the prevention and the diagnosis of coronary artery disease (CAD) are both advancing rapidly. Better understanding of risk factors and advances in drug treatment for lipid disorders and hypertension have improved prevention capabilities. At the same time, researchers continually announce new tests or refinements of existing tests to detect or map CAD noninvasively. Prominent on the growing list are computed tomographic scans for calcium, contrast and three-dimensional echocardiographic techniques, and high-speed magnetic resonance imaging (MRI) methods, the last of which may be a substitute for echocardiography in selected patients.

It's not yet clear what these developments mean for primary care physicians or for healthcare in general. No test has yet emerged as a proven, noninvasive screen that will take most of the guesswork out of determining who needs aggressive treatment to ward off heart disease. Nor has a noninvasive method appeared that can take the place of cardiac catheterization as a way to map coronary lesions. But several techniques look promising.

Electron-Beam Calcium Scanning

Calcium deposits are a constituent of atherosclerotic plaque in the coronary arteries and can be detected by electron-beam computed tomography (EBCT, also known as ultrafast CT), a technology that has existed for about 10 years. EBCT scanners are found in about 50 major medical centers in the United States, according to John D. Cantwell, MD, an Atlanta cardiologist and editorial board member of The Physician and Sportsmedicine.

In EBCT, an electron beam and tungsten target are used to generate x-rays while rotating around the patient. Images are obtained in 1/20 of a second, 20 times faster than a conventional CT scanner, according to reports in the literature. EBCT scans take only about 10 minutes.

EBCT has attracted much attention as a potential screen for CAD in asymptomatic patients. In a recent review in Clinical Cardiology, Budoff and Brundage (1) wrote that EBCT can "both detect and quantitate the presence of atherosclerosis." They found that a negative EBCT scan is highly accurate for excluding CAD and that a high calcium score signifies a 6-fold to 35-fold increase in the risk of having a cardiac event. However, they also reported that calcium in the arteries does not always correlate with findings of obstructive disease on angiography. And two other authors, writing recently in The Journal of the American Medical Association, (2) stated that "at best, there is only a moderate correlation between the quantity of calcification and the presence of stenosis and its severity."

Cantwell (3) has explained that calcified plaque may occur in arterial walls without obstructing the artery, since atherosclerosis triggers artery remodeling that may maintain the lumen. He also points out that false-negative EBCT scans (atherosclerosis without a calcium buildup) are possible, particularly in select patient populations such as smokers in their 40s and 50s.

Jeffrey Carr, MD, assistant professor of radiology and public health sciences at Wake Forest University School of Medicine in Winston-Salem, North Carolina, agrees that EBCT screening for coronary calcium is not ready for wide use. "We know that calcium is a marker of atherosclerosis and that your risk of a future heart attack goes up," he says. "But there's not enough information now to recommend screening of healthy people for calcium in the coronary arteries."

The American Heart Association (AHA), in a recently released report (4) on identifying patients who need treatment for primary prevention of heart disease, did not endorse routine calcium scanning for risk assessment in asymptomatic persons. An AHA writing group concluded that coronary calcification correlates strongly with atherosclerosis, but it is not yet clear how well calcium scores, independent of traditional risk factors, predict cardiac events. Nonetheless, the panel wrote that "selected use" of calcium scanning for patients with intermediate risk may be appropriate.

An EBCT Alternative?

Another issue with regard to EBCT is the cost and availability of the equipment. Cantwell says the devices cost about $2 million; the reported cost of the test ranges from $420 to $500. However, electron-beam scanners may not be the only tool for the job. Recently, radiologists have reported the ability to detect coronary calcification with spiral, or helical, scanners, the type now found in most US hospitals. Using spiral CT could reduce the cost significantly. (In these scanners, the gantry holding the x-ray tube rotates around the patient in a spiral.)

Budoff and Brundage (2) reported that calcium scanning with conventional and spiral CT equipment is hampered by several problems that limit the sensitivity and reproducibility of tests. However, a technique developed by Carr and his colleagues may overcome these problems by linking a spiral scanner to an ECG machine. At the AHA's annual conference on epidemiology and prevention last March in Orlando, Carr et al (5) reported that electron-beam and spiral CT were equally effective in detecting coronary calcium in 36 patients. "We cardiac-gate the scanner, so we can get images of the heart in diastole, when it's relatively motion-free," he told The Physician and Sportsmedicine. The patient must hold his or her breath for about 10 seconds; the test takes about 15 minutes, he said. He added that most CT scanner manufacturers are developing systems for "cardiac-gating" their machines.

Another promising technology is contrast-enhanced EBCT, which provides better visualization of the coronary arteries than conventional EBCT does. The authors of a recent review (6) stated that, with special computer equipment, contrast-enhanced EBCT provides three-dimensional depictions of the coronary vessels. Currently, this method "is a reaonably robust technique for the visualization and assessment of the left main and left anterior descending coronary artery," but much of the right and circumflex coronary angiograms are not interpretable, the reviewers wrote.

Both EBCT and spiral CT scanners—among numerous other tests—will be evaluated in a long-term, multicenter study to determine which can best identify people at high risk for CAD (see "Mammoth Study Seeks Best Tests for Heart Disease Prevention, page 56).

New Dimensions in Echocardiography

Another frontier in imaging of heart disease is in echocardiography, where researchers are working on three-dimensional echo techniques and the use of contrast agents to make myocardial perfusion visible echocardiographically.

At Rush-Presbyterian-St. Luke's Medical Center in Chicago, physicians report they are using a new device to produce three-dimensional ultrasound scans of the whole heart in the time it normally takes to produce one two-dimensional image. The Volumetrics Model 1 (Volumetrics Medical Imaging, Inc, Durham, North Carolina) produces three-dimensional images in real time, according to Rush-Presbyterian officials.

Previously, the only way to create three-dimensional ultrasound images of the heart was to synthesize them from a series of two-dimensional images, which takes several hours, says James E. Macioch, DO, assistant director of echocardiography at Rush Medical College in Chicago. The Volumetrics machine produces instantaneous scans of the whole heart, without requiring ECG gating or breath-holding. The image is displayed in two dimensions on a monitor, but the machine's computer "sees and processes the image of the heart in three dimensions," allowing cardiologists to examine the heart from an infinite number of views, Rush-Presbyterian reported in a press release last July.

This technique is useful "whenever you need exact information on left ventricular function," says Macioch. The images reveal wall motion and the volume of each cardiac chamber (thus yielding left ventricular ejection fraction) and show valve anatomy, he explains. The information helps cardiologists determine the degree of heart failure and assess whether a faulty valve can be surgically repaired or must be replaced, according to the press release.

The Volumetrics machine is about the size and weight of two conventional echocardiography devices and costs about $250,000, according to Macioch. He estimates that the cost of a three-dimensional scan ranges anywhere from $200 to $1,000, depending on the setting. The press release noted that only 19 institutions worldwide had the device.

Other echocardiography specialists are working to make blood flow detectable by ultrasound. Conventional ultrasound techniques do not show myocardial perfusion, but the use of contrast agents may solve that problem. In January 1998 the FDA approved Optison, an ultrasound contrast agent invented by Steven B. Feinstein, MD, professor of medicine and director of echocardiography at Rush-Presbyterian-St. Luke's. Optison (Molecular Biosystems Inc, San Diego) has made it possible to detect myocardial perfusion abnormalities safely and consistently, according to Feinstein and colleagues.

Optison consists of microspheres, or bubbles, of human albumin that are half the size of red blood cells and contain fluoropropane gas, says Feinstein. "They are near-perfect reflectors of ultrasound energy," he says. For echocardiography, 0.5 mL of Optison is injected intravenously. "When you have severe narrowing [in a coronary artery], the bubbles don't get through that, and hence you don't get much signal," Feinstein says. "It's comparable to thallium scintigraphy, but you get much higher resolution, and you're imaging in real time."

The indications for using contrast echo are the same as for conventional echo, except the test cannot be used with patients who have drug-related or blood-product allergies, Feinstein says. "We use it on all of our stress tests and 40% of our portable echoes." Contrast echo uses a conventional echo machine with upgraded computer software and costs about $100 more than the conventional test, Feinstein reports.

Feinstein describes contrast echo as a complement to, rather than a replacement for, coronary angiography. "Angiography shows the anatomy of the blood vessels—the branches of the tree," he says. "We show the leaves on the branches—the living blood in the tissue."

In a press release, Rush-Presbyterian officials said contrast echo costs less than thallium scintigraphy and positron emission tomography and does not involve radiation. Contrast echocardiograms currently account for about 1% of the echocardiograms produced annually in the United States, but their use is growing rapidly, officials said.

A group at Emory University in Atlanta (7) recently compared the overall diagnostic costs associated with contrast and noncontrast echo in 203 patients who were involved in a phase 3 clinical trial of Optison. Inadequate echocardiograms rendered additional tests necessary for 42% of the patients in the noncontrast group, compared with 12% of those in the contrast group. The use of the contrast agent increased the initial diagnostic cost by $125 per patient, but because fewer additional tests were necessary, overall costs were 17% lower with contrast echo.

Despite such promising reports, contrast echo is not mentioned in the AHA's recent "Prevention Conference V" statement (4) on identifying asymptomatic patients who need treatment to prevent heart disease. The statement notes that conventional stress echocardiography may be useful for assessing women and elderly patients in this category. Regarding myocardial perfusion imaging for asymptomatic patients, the report summary says only that stress thallium scintigraphy may be useful in assessing postmenopausal women and elderly men.

Cardiac Magnetic Resonance Imaging

In a 1998 review (8) cardiovascular radiologist George Hartnell wrote, "Ischaemic heart disease remains the one area of cardiac investigation where MRI currently has relatively little to offer." But Hartnell predicted that this will change as MR imaging gets faster and sharper, and recent reports suggest that certain MRI techniques are already useful for diagnosing CAD, especially in patients who are poor candidates for echocardiography.

For example, cardiologists at Wake Forest University in recent months have reported using MRI to detect coronary artery obstructions and to show wall-motion abnormalities at high heart rates. When used with special software, MRI was just as effective as catheterization in detecting significant coronary blockages in 30 patients, the group reported in an article (9) published last June in Circulation. Patients who underwent the test were given dobutamine to induce tachycardia and were required to hold their breath for 8 to 15 seconds.

"MRI allows us not only to locate a blockage, but to determine whether it limits blood flow enough to warrant treatment," said lead author W. Gregory Hundley, MD, in a Wake Forest press release. "There is no other noninvasive test that can do this." Hundley is an assistant professor of cardiology and radiology at Wake Forest University Baptist Medical Center.

Tested against quantitative coronary angiography, MRI had a sensitivity of 100% and a specificity of 83% for identifying stenoses of more than 70% in the left main and left anterior descending coronaries, according to Hundley and colleagues. Their report states that MRI assessment of coronary flow reserve may prove useful for identifying candidates for coronary artery bypass surgery. However, the Wake Forest press release noted that MRI cannot currently replace catheterization as a guide for performing angioplasty or bypass surgery.

Hundley et al (10) also recently reported the successful use of high-speed MRI to show heart wall motion in patients for whom stress echocardiography is contraindicated or ineffective, such as those who are obese or have lung disease. A technique called "fast cine MRI" was used to study the rapidly beating heart in patients who were given dobutamine. The technique shows the heart's movement in close to real time, according to a news release from the AHA (publisher of the report in Circulation).

"MRI gives you a noninvasive alternative for people who cannot have an ultrasound, who are often our sickest patients," Hundley said in a Wake Forest news release. More than 350 Wake Forest patients have had the fast cine MRI scan, which takes about 35 minutes, according to the release.

Elsewhere, a group at Stanford University in Stanford, California (11), has reported the successful use of high-speed MRI as a complement to echocardiography for assessing left ventricular function. The Stanford group used an MRI scanner with computer-based modifications that enable it to display any scan plane at 16 images per second, without the need for cardiac gating or breath-holding, according to their 1998 published report.

The Stanford investigators compared echo and MRI results in 85 patients who were grouped according to the quality of their echocardiographic images. In patients whose echo images were good, the MRI results were of equal clinical utility. In two other groups, MRI findings were clinically adequate nearly all the time, whereas echo images were adequate only 38% and 58% of the time.

Like contrast echocardiography, cardiac MRI was not endorsed by the AHA in its recent report (4) on assessment for primary prevention of heart disease. The AHA panel called cardiac MRI "a promising research tool" but concluded that it "is not ready for application in the identification of patients at high risk for CAD."

Maximizing Available Tests

New radiologic and ultrasound techniques sound promising, but which, if any, will become a practical, widely accessible tool for detecting CAD is anyone's guess. For now, some suggest that physicians can make better use of standard tools for diagnosing heart disease and thereby decrease their reliance on high-tech tests (see "Scoring System May Improve Utility of Exercise Testing," page 62). In any case, it's clear that physicians will continue to use a broad and changing array of tools to detect the disease that kills more people than any other.


  1. Budoff MJ, Brundage BH: Electron beam computed tomography: screening for coronary artery disease. Clin Cardiol 1999;22(Sept):554-558
  2. Siegel MJ, Evens RG: Advances in the use of computed tomography. JAMA 1999;281(14):1252-1254
  3. Cantwell JD: A new cardiac test for an older doctor. Phys Sportsmed 1998;26(6):87-90
  4. Smith SC, Greenland P, Grundy SM: Prevention Conference V: Beyond secondary prevention: identifying the high-risk patient for primary prevention, executive summary. Presented at the American Heart Association 72nd Annual Scientific Sessions, Atlanta, November 7-10, 1999
  5. Carr JJ, Burke GL, Goff DC Jr, et al: Coronary artery calcium scores correlate strongly between fast gated helical and electron beam computed tomography. Presented at the American Heart Association's 39th Annual Conference on Cardiovascular Disease, Epidemiology, and Prevention, Orlando, March 26, 1999
  6. Rensing BJ, Bongaerts AH, van Geuns RJ, et al: Intravenous coronary angiography using electron beam computed tomography. Prog Cardiovasc Dis 1999;42(2):139-148
  7. Shaw LJ, Gillam L, Feinstein S, et al: Use of an intravenous contrast agent (Optison) to enhance echocardiography: efficacy and cost implications. Optison Multicenter Study Group. Am J Manag Care 1998;25(4):SP169-176
  8. Hartnell GG: New developments in cardiac magnetic resonance imaging. Hosp Med 1998;59(7):567-573
  9. 9. Hundley WG, Hamilton CA, Clarke GD, et al: Visualization and functional assessment of proximal and middle left anterior descending coronary stenoses in humans with magnetic resonance imaging. Circulation 1999;99(25):3248-3254
  10. Hundley WG, Hamilton CA, Thomas MS, et al: Utility of fast cine magnetic resonance imaging and display for the detection of myocardial ischemia in patients not well suited for second harmonic stress echocardiography. Circulation 1999;100(16):1697-1702
  11. Yang PC, Kerr AB, Liu AC, et al: New real-time cardiac magnetic resonance imaging system complements echocardiography. J Am Coll Cardiol 1998;32(7):2049-2056

Mammoth Study Seeks Best Tests for Heart Disease Prevention

If a test or characteristic has shown any promise at all as a tool for detecting heart disease early, it's probably included in the Multi-Ethnic Study of Atherosclerosis. In MESA, as it's called, 20,000 middle-aged and older people (ages 45 to 84) will be examined and followed for 8 1/2 years to study the relationship between coronary artery disease and a wide variety of signs and risk factors.

Protocol development for the study began in 1999, and examination of patients at six sites around the country is scheduled to start in July, according to Diane Bild, MD, MPH, project officer and assistant director for field studies and clinical epidemiology at the National Heart, Lung, and Blood Institute (NHLBI) in Bethesda, Maryland. The NHLBI is the study's sponsor.

"We want to study the pathophysiology—what characteristics of what types of subclinical [heart] disease, along with what risk factors, lead to clinical events?" Bild says.

Among the tests to be used, she says, are electron-beam and helical (spiral) computed tomography scans for coronary calcification, magnetic resonance imaging scans, carotid artery ultrasound, ultrasound testing of endothelial function in the brachial artery, resting electrocardiography, and wave forms in the radial artery. Researchers also will assess a number of blood factors, and subjects will be questioned about diet, exercise, and psychosocial stress, she adds.

The ultimate goal is to learn which tests "add to our ability to select people at higher risk so that we can intervene," says Jeffrey Carr, MD, a radiologist and public health scientist at Wake Forest University School of Medicine in Winston-Salem, North Carolina, one of the participating centers.

As its name suggests, the study will include significant numbers of minorities and women, according to Carr. "For a lot of these tests we don't have enough information on women and minorities," he says. "Our hope is to target the groups that up till now are difficult to diagnose."

Scoring System May Improve Utility of Exercise Testing

Tests to diagnose heart disease and determine its severity grow ever more sophisticated, but a standard exercise test yields data that, when plugged into certain equations, can reduce the need for more costly tests, according to researchers from California and Michigan.

A mathematical scoring system based on exercise test results makes it possible to assess a patient's risk level more accurately than physicians can estimate it from clinical findings and the raw exercise test results, according to the report by Victor F. Froelicher, MD, and several colleagues (1). The scoring system uses three equations that are detailed in the exercise testing guidelines (2) of the American College of Cardiology and American Heart Association.

In the study, the scoring system was 70% accurate in predicting the presence or absence of significant coronary artery stenosis—a result that Froelicher calls "quite comparable" with the yield from echocardiography or thallium scintigraphy. Froelicher is a professor of medicine and director of the division of cardiovascular medicine at Palo Alto (California) VA Medical Center and a senior associate editor of The Physician and Sportsmedicine.

The retrospective study, conducted at Stanford University in California and William Beaumont Hospital in Birmingham, Michigan, involved 599 male patients who had not had a myocardial infarction. The men underwent an exercise test and angiography. The clinical data and ECG test results, but not the angiography results, were reviewed by three groups of physicians: "expert" cardiologists, other cardiologists, and internists (a total of 142 physicians). The physicians classified the patients as having a high, intermediate, or low risk of CAD and assigned a numeric probability for each.

The probability of CAD was also assessed by weighing the three treadmill scores. These values and the physicians' estimates were then compared with the angiographic findings, which showed that 58% of the men had CAD, defined as 50% or greater stenosis in a coronary artery.

Formula Wins Out
The scoring technique turned out to be more accurate than the physicians. When a probability cut point of 70% was used (ie, a patient was predicted to have CAD if the estimated probability exceeded 70%), the scores had a predictive accuracy of 70%, vs 64% for both groups of cardiologists and 62.5% for the internists. When the two methods were compared over a wide range of cut points, the scores were 77% accurate overall, vs 69% for the expert cardiologists, 65% for the other cardiologists, and 66% for the internists.

The treadmill scores can determine the next diagnostic or treatment step for a patient, says Froelicher. "Low risk gets nothing, high risk proceeds to catheterization, and intermediate risk to other tests," he says. For patients identified as having a low or high risk, he suggests, the treadmill score renders echocardiography or thallium scintigraphy unnecessary. He also says treadmill scoring may reduce the need for calcium scanning by electron-beam computed tomography.

The treadmill scoring equations use age, gender, symptoms, risk factors, maximal metabolic equivalents (METs), maximal heart rate, and ST-segment depression, according to Froelicher. Though the equations are not new, they are not much used, he says, adding, "We're trying to get them to be used more. People lean to the more complex tests, but we hope that when people see that this [method] does as well as physicians [in predicting risk], they'll be more apt to use it."

Froelicher notes that the equations are "a bit complex" but are not difficult to use when programmed into a computer.

John D. Cantwell, MD, an Atlanta cardiologist and editorial board member of The Physician and Sportsmedicine, says that a number of treadmill scores have been devised but that most are "a little too complex to use for the guy out there on the firing line." He prefers the Duke treadmill score, a simple formula that uses only time on the treadmill, maximal ST-segment change, and the degree of angina to estimate a patient's risk of cardiovascular death within 4 or 5 years.


  1. Lipinski MJ, West J, Osterberg L, et al: Can physicians diagnose coronary disease better than scores? Presented at the American Heart Association Annual Scientific Sessions, Atlanta, November 9, 1999
  2. American College of Cardiology and American Heart Association: ACC/ AHA guidelines for exercise testing. J Am Coll Card 1997;30(1):260-315

Mr Roos is senior editor of The Physician and Sportsmedicine.



The McGraw-Hill Companies Gradient

Copyright (C) 2000. The McGraw-Hill Companies. All Rights Reserved
Privacy Policy.   Privacy Notice.