| John A Rumberger, PhD, MD |
Apr 12, 2000 |
Electron beam computed tomography (EBCT) is the current name applied to a unique x-ray imaging device formerly known as "ultrafast" CT or cine CT. EBCT uses a unique electron beam configuration with no moving parts to the imaging chain and inherently defines the heart via three-dimensional acquisition of multiple, high spatial and/or temporal resolution, parallel tomograms (show figure 1). True "snap shot" images (50 msec or 100 msec) are acquired by timing scan acquisition to the subject's cardiac cycle.
EBCT functions as a body scanner and is therefore able to perform high resolution imaging of the thorax and abdomen in much the same fashion as standard and spiral CT devices. This includes imaging of the thoracic and abdominal aorta, high resolution lung scanning, and CT of the abdomen and pelvis. EBCT also has a number of specific cardiovascular applications (eg, coronary heart disease) because of the unique ability to scan with the ECG trigger. The current status of the use of EBCT for these purposes will be reviewed here.
RIGHT AND LEFT VENTRICULAR CHAMBER VOLUMES, FUNCTION, AND MUSCLE MASS ！ Providing quantitative data about the size and function of the heart has always been one of the goals of cardiac imaging. EBCT provides a quick, efficient, and minimally invasive means to obtain high spatial, density, and temporal resolution images of the beating heart at multiple tomographic levels in a known three-dimensional registration. Quantitative right and left ventricular chamber volumes, left ventricular muscle mass, and regional function are determined to a very high degree of accuracy (within 5 percent of the actual values) by acquiring polytomographic images (50 msec/image at a rate of 17 frames/second) from the left ventricular apex through the right ventricular outflow tract [1,2].
Up to 12 contiguous tomographic levels (approximately 1.0 cm center-to-center) with 10 to 20 images through the cardiac cycle can be performed during a single intravenous injection of nonionic contrast media (show radiograph 1). Composite calculations of ventricular chamber volumes and muscle mass are determined using a modification of Simpson's rule (show figure 2).
Since images are also acquired throughout the cardiac cycle, EBCT can quantify regional ventricular function and wall thickening [3,4], ejection fraction [5], contractility [6], the peak rate of systolic emptying, and early diastolic function [6,7]. Additional applications include measurement of right ventricular mass (in single slice mode) [8], quantitation of univalvular regurgitation [9], assessment of infarct size [10], and serial studies of cardiac remodeling after infarction [6,11,12].
CORONARY ARTERY DISEASE ！ There are several methods by which EBCT can be used to evaluate the degree of coronary atherosclerosis: detection of coronary artery calcification; noninvasive coronary angiography; and assessment of myocardial perfusion. This technique can also assess coronary artery bypass graft patency.
Coronary artery calcification ！ Coronary artery calcification is regulated in a manner similar to bone formation. Variable amounts of calcium are found with atherosclerosis and can connote active plaque development [13]. Although calcification is seen more frequently in advanced lesions, it can also occur in small amounts in earlier stages of atherosclerosis, particularly in high risk patients (show radiograph 1). Among patients with familial hypercholesterolemia, for example, EBCT can detect significant coronary calcium as early as 11 to 24 years of age [14].
EBCT can noninvasively detect and quantify coronary calcification, even when small and of low density, using serial and contiguous ECG-triggered 100 msec, thin section (3 mm) tomograms from the aorta through the apex of the heart (show figure 3) [13]. Direct relationships have been established between coronary calcium as measured by EBCT and histologic [15,16], ultrasonic [17], and angiographic [18,19,20] measures of coronary disease on a heart-by-heart, vessel-by-vessel, and segment-by-segment basis (show figure 4).
A number of studies have demonstrated that coronary calcification detected by EBCT is found in individuals who have significant angiographic stenosis, with a sensitivity ranging from 90 to 100 percent, a specificity of 45 to 76 percent, a positive predictive accuracy of 55 to 84 percent, and a negative predictive accuracy of 84 to 100 percent [15,18,19,21,22,23]. One series of 290 men and women undergoing coronary angiography found that a coronary calcium score (a weighted sum of x-ray density and total calcium area)
80 was associated with an increased likelihood of any coronary artery disease
regardless of the number of clinical risk factors for coronary disease that
were present; a score 170
was associated with an increased likelihood of obstructive coronary artery
disease regardless of the number of risk factors [24].The EBCT calcium score, in association with risk factor analysis, can rule in or rule out angiographic three-vessel or left main disease. Based upon data from 291 patients, a noninvasive index (NI) using the separate calcium scores transformed into the natural logarithmic scale for the left anterior descending (Log(e)[LAD]) and left circumflex artery (Log(e)[LCx] was constructed [25]:
NI = Log(e)[LAD] + Log(e)[LCx] + 2[if diabetic] + 3 [if male]
Noninvasive values >14 increased the probability of angiographic three-vessel or left main disease from 23 percent (pretest) to 65 to 100 percent (posttest) and noninvasive values <10 increased the probability of no severe disease from 77 percent (pretest) to 95 to 100 percent (posttest). Various noninvasive cut points had sensitivities of 87 to 97 percent and specificities of 46 to 74 percent (show figure 5).
Use for prognosis ！ There is increasing evidence that a clinical measure of the EBCT calcium score has prognostic value in both symptomatic and asymptomatic patients [26,27,28,29]. This is illustrated by the following observations:
• In one study of 501 symptomatic patients, a calcium score of
100
was highly predictive of a cardiac event and, in a logistic regression which
included age, gender, and angiographic findings, only the calcium score
predicted events [26].• Another series evaluated 1173 initially asymptomatic patients who were followed for 19 months [27]. Calcium score thresholds of 100, 160, and 680 with EBCT had sensitivities for a subsequent event of 89, 89, and 53 percent, respectively, and specificities of 77, 82, and 95 percent, respectively. Negative predictive values were above 99 percent, and the odds ratios for development of symptomatic coronary disease ranged from 22:1 to 36:1.
These studies and others provide a foundation for the use of EBCT coronary artery calcium scanning as a rapidly emerging clinical tool to estimate the coronary atherosclerotic "plaque burden" in asymptomatic and symptomatic adults suspected of having premature or developing coronary disease. One study also indicated that EBCT scanning in the symptomatic patient provided an incremental value for assessing the severity of angiographic disease as compared to risk factor analysis or SPECT [30]. However, its role for establishing prognosis in asymptomatic low or high risk patients remains uncertain and may offer limited benefit over that determined by traditional risk factor assessment. Still, it has significant potential in allowing refinement of risk in those at traditional "intermediate" risk in whom Framingham models may be limited [31]. Although guidelines for scanning and interpretation of calcium scores in clinical practice have been put forward for such patients, they are in development and will require refinement as more data come forward; still, suggestions can be made for clinical applications based upon currently available literature (show figure 6 and show figure 7) [13,32]:
• Evaluation of patients with chest pain, with the results used in the decision to perform adjunctive or additional noninvasive stress testing, coronary angiography, or to continue medical therapy. New software can allow clinical calcium scoring in 5 to 10 minutes. A number of studies have demonstrated that coronary calcification assessment with EBCT has a sensitivity for significant angiographic stenosis (85 to 100 percent) that is comparable and possibly superior to that of exercise testing with or without thallium in symptomatic patients (show figure 6) [23,33,34,35]. Furthermore, the analysis is quite rapid, since current software can perform the scoring and provide a full report in five to ten minutes.
• Screening of asymptomatic subjects who are at high risk for coronary heart disease in order to identify those who require aggressive risk factor management, further diagnostic workup with exercise testing and angiography, and exclusion from high risk occupations. At present, there are insufficient data to recommend the use of EBCT as a single diagnostic modality for screening of low risk, asymptomatic subjects; however, there are data to support its use in the "intermediate" risk asymptomatic patient in which the contribution of multiple sub-threshold risk factors is difficult to determine using conventional risk assessment strategies.
• Following the progression of coronary atherosclerosis with serial scans to help determine the efficacy of pharmacologic or nutritional intervention aimed at retarding the progression of atherosclerosis. As an example, a retrospective study evaluated 149 patients, 105 of whom received a statin drug, who underwent EBCT at baseline and after a minimum of 12 months: there was a net reduction in the calcium-volume score only in treated patients whose final serum LDL-cholesterol concentration was less than 120 mg/dL (3.1 mmol/L) [36]. In contrast, the calcium-volume score increased in both untreated patients and, to a lesser degree, in treated patients whose LDL-cholesterol was
120
mg/dL (show
figure 8). (See
"Mechanisms of benefit of lipid lowering in patients with coronary heart
disease").The use of EBCT for this indication is justified only if coronary calcium measurements accurately track atherosclerotic volume and have sufficient interstudy reproducibility, issues which have not yet been adequately established; however the use of the "calcium volume score [35] appears to offer considerable increase in precision over the more traditional Agatston calcium score used for EBCT calcium studies.
Much of the current confusion regarding the clinical application of EBCT coronary artery calcium scanning has been the paucity of clear suggestions for its application in clinical practice. As an example, one study examined 105 patients, most of whom were middle aged men and women without a prior diagnosis of heart disease, who had a normal or non-diagnostic electrocardiogram and negative cardiac enzymes [37]. These patients underwent a routine clinical work up, which included stress testing, imaging, and/or coronary angiography. Prior to discharge an EBCT scan was obtained on each patient to evaluate for coronary artery calcification, but this information was not included in the clinical evaluation. More than half the patients had no detectable coronary calcification; there were no patients with a diagnosis of coronary disease and a negative EBCT scan. Retrospective review of the clinical data showed that coronary calcium on EBCT had a sensitivity, specificity, and negative and positive predictive value of 100, 63, 100 and 100 percent, respectively. These data need to be extended towards a larger group of patients, but suggest that EBCT can be most helpful in ruling out obstructive coronary disease in patients that are clinically felt to be low to intermediate pre-test likelihood.
A review of the current literature has suggested guidelines in the clinical interpretation of EBCT calcium scans in asymptomatic but "at risk" individuals [38]. EBCT calcium scores are divided into several categories (0 to 10, 11 to 100, 101 to 400, >400) as well as taking note if the score was above or below the 75th percentile, adjusted for age and gender (show figure 6 and show figure 7) [32]. However, long term application of these guidelines will be necessary to determine if further refinements will be needed in the recommendations (show table 1).
Identification of silent ischemia ！ Calcium severity on EBCT can identify asymptomatic patients at high risk for coronary heart disease, but it is not certain if this translates into identification of asymptomatic patients who have silent ischemia, which is of importance since the presence of silent ischemia is predictive of a cardiac event. (See "Silent myocardial ischemia: Prognosis and therapy"). Noninvasive techniques currently available for establishing the presence of silent ischemia include ambulatory monitoring, exercise treadmill testing, and myocardial perfusion scan. (See "Silent myocardial ischemia: Diagnosis and screening").
The role of EBCT for identifying asymptomatic patients with myocardial ischemia was evaluated in one study of 411 patients who had a exercise stress test with myocardial perfusion imaging within a close time period of the EBCT [39]. Although most subjects (78 percent) with coronary calcium on EBCT did not have inducible ischemia with exercise testing, the likelihood of ischemia increased with calcium score, regardless of age or sex. No subject with a score <10 had ischemia, while 2.6 percent of those with a score of 11 to 100, 11.3 percent with a score of 101 to 399, and 46 percent of those with a score
400
did have ischemia.Ischemic versus nonischemic cardiomyopathy ！ Since EBCT can identify and quantitate coronary calcification which correlates well with angiographic stenosis, this noninvasive technique has the ability to distinguish between ischemic and nonischemic cardiomyopathy. One study of 125 patients with cardiomyopathy found that 99 percent of those with ischemic cardiomyopathy had coronary calcification on EBCT with a mean score of 798 [40]. In contrast, the mean score was significantly lower (17) in those with a nonischemic cardiomyopathy, 83 percent of whom had no demonstrable coronary calcification [40]. The specificity of EBCT for excluding coronary disease in patients with a cardiomyopathy was 83 percent when a threshold score of 0 was used and 92 percent for a score
80.
The overall accuracy for determining the etiology of cardiomyopathy was 92
percent.Cardiac transplantation ！ The diagnosis of transplant vasculopathy is often difficult to establish, and the disease generally progresses rapidly and silently. Although coronary angiography and intracoronary ultrasound are the most widely used techniques, they are invasive and may not be accurate because of the diffuse nature of the disease. (See "Diagnosis and treatment of cardiac transplant vasculopathy"). EBCT may provide useful information in such patients noninvasively. An initial study evaluated EBCT and coronary angiography in 102 cardiac transplant recipients: 40 percent had a stenosis of
24
percent in at least one artery with coronary angiography, while 46 percent had
coronary calcium on EBCT [41].Use in the emergency room for patients with chest pain ！ Patients presenting to the emergency room with chest pain often undergo a thorough evaluation to assess the possible presence of anginal pain due to acute coronary ischemia. Patients with a good clinical history who do not clearly have a Q-wave or non-Q-wave MI are considered to have undifferentiated chest pain. Such patients may still have an acute coronary syndrome and further evaluation is indicated; EBCT may be a potentially useful technique. (See "Evaluation of suspected acute coronary ischemia in the emergency department"). One study of 181 such patients who had a normal or nondiagnostic ECG found that a negative EBCT, ie, a coronary artery calcium score of 0, perfectly predicted which patients could be safely discharged; it had a negative predictive value of 100 percent after excluding one patient who was a cocaine user [42]. In contrast, the 30 day event rate was 8 percent in those with a positive score.
Cost and limitations ！ Assessment of coronary calcification by EBCT can be obtained in any subject, but it provides anatomic rather than physiologic information. The testing is noninvasive, requires minimal patient cooperation and preparation, and medication does not need to be discontinued prior to study. Results are immediately available for qualitative evaluation, while a quantitative calcium score requires 10 minutes or less to obtain. The radiation dose for a single screening EBCT scan is 82 mrem for men and 150 mrem for women, which is at least 10 to 100 times less than that of conventional coronary arteriography [13].
The cost of EBCT varies, but ranges between $300 to $400 [13]. This is similar to the cost for a routine nurse-monitored treadmill exercise test and about one-half the cost for a stress echocardiogram and one-third the cost of a nuclear imaging stress test.
EBCT and the use of the calcium score is cost-effective for the diagnosis of obstructive coronary artery disease. In one study, an EBCT calcium score of 168 was the least costly and most cost-effective noninvasive method when compared to treadmill exercise testing, exercise echocardiography, or exercise thallium scanning [43]. For ambulatory patients with a low to moderate disease prevalence (pretest likelihood of disease
70
percent), the most cost-effective initial noninvasive testing approach was the
calcium score by EBCT, while for those with a high prevalence of disease
(>70 percent), direct angiography was the most cost effective as the first
and only test.These findings were corroborated by a second study that compared EBCT to exercise testing for the initial evaluation of 207 patients with low to intermediate probability of coronary disease [44]. Patients who underwent EBCT first and had a calcium score
150
on EBCT, which was considered diagnostic of obstructive coronary disease,
underwent exercise testing, followed by angiography if the exercise test was
positive. Patients who initially underwent exercise testing and had a positive
or equivocal test underwent myocardial perfusion imaging, followed by
angiography if imaging was suggestive of ischemia. The diagnostic pathway
starting with EBCT provided a 45 to 65 percent cost saving over the the
pathway beginning with exercise testing; the cost benefit decreased as the
prevalence of the disease increased.Noninvasive coronary angiography ！ Intravenous contrast-enhanced, thin and overlapping tomographic sections of the heart are now possible during a single breath-hold using software recently incorporated into the EBCT scanner (show figure 9). Three-dimensional registration of data at end-diastole facilitates true 3-D images of the major epicardial coronary arteries [45,46,47].
Commercially developed software specifically available for EBCT permits rapid 3-D rendering, making this technique a potentially useful adjunct to conventional coronary angiography. Studies in approximately 200 patients worldwide using EBCT coronary angiography have demonstrated high sensitivity (77 to 82 percent), specificity (92 to 94 percent) and overall accuracy (87 percent) for establishing luminal stenosis of
50
percent when compared directly with selective, invasive coronary angiography (show
figure 10) [48].Despite the high sensitivity and specificity of EBCT noninvasive coronary angiography for identifying high grade coronary stenoses, a major limitation is inadequate image quality. In one study of 125 patients, for example, 25 percent of 500 coronary segments were excluded from analysis for this reason [49]. The most frequent vessels which were poorly visualized were the right coronary and left circumflex arteries because of their position in the coronary groove, resulting in an increased diastolic motion during atrial contraction. Interpretation may be further compromised by artifacts of respiration if patients are unable to hold their breath and by heavily calcified vessels, which are a frequent cause of false negative or false positive results [45,50]. However, other studies have shown a much greater ability to visualize the RCA and LCX vessels, suggesting that there continue to be methodological differences between laboratories that must be resolved before this method can be placed into general clinical use.
Myocardial perfusion ！ In patients with coronary artery disease, assessment of the hemodynamic or "physiologic" significance of a specific stenotic lesion may be as important as knowledge of the angiographic estimate of stenosis severity. One important physiologic measurement is myocardial flow or perfusion. (See "Exercise perfusion testing in the diagnosis of coronary heart disease").
Assessment of alterations in regional myocardial perfusion produced by coronary artery stenoses contributes to a greater understanding of stenosis significance and can be helpful in directing or guiding the choice of therapeutic interventions. EBCT and other fast-CT methods have accurately quantitated regional myocardial perfusion [51,52,53]. This technique involves a bolus intravenous injection of contrast followed by scanning on successive heart beats in multiple cardiac planes at predefined phases of the cardiac cycle [54]. By assessing the first pass of contrast media through a given myocardial region of interest, a "time-density" curve can be generated. The peak opacification and mean transit time of contrast through the region are directly proportional to the mean blood flow or perfusion. Ongoing studies suggest that clinically useful data on absolute regional myocardial perfusion and flow reserve are possible using EBCT.
The administration of coronary vasodilators has improved the ability of EBCT to assess coronary blood flow. One preliminary investigation of EBCT scanning with nonionic contrast was performed in 14 normal subjects studied at rest and 10 patients during maximum coronary vasodilatation produced with an intravenous infusion of adenosine (140 mg/min) [54]. Myocardial perfusion was established by determining the areas under the flow curves in the myocardium and left ventricular cavity and then directly applying the Steward Hamilton equations. The mean global resting flow, as estimated by EBCT, was 106 mL/min per 100 gm and increased to 301 mL/min per 100 gm during adenosine infusion; the flow reserve was 3.1 (show figure 11). Flow reserve is equal to the maximum flow after vasodilation divided by the resting flow.
Regional absolute flow and flow reserve were also assessed in the anterior, lateral, and septal myocardial walls. These values were relatively uniform at rest and had similar increases in flow during intravenous adenosine infusion. Flow reserve varied from 2.7 in the anterior wall to 3.3 in the septal wall. These findings in normals are in agreement with more invasive methods such as intracoronary Doppler flow wire measurements obtained during catheterization.
Comparison with myocardial perfusion scanning ！ Exercise testing with myocardial perfusion scanning is a well established noninvasive method for detecting coronary artery disease. One study compared rest and exercise EBCT with stress sestamibi SPECT in 33 patients with chest pain who underwent coronary angiography [55]. Exercise EBCT, analyzed using a global ejection fraction, had a sensitivity and specificity for coronary disease of 81 and 76 percent, respectively, compared with angiography. When the development of a new regional wall motion abnormality was considered evidence for obstructive coronary artery disease, the sensitivity and specificity of EBCT were 88 and 100 percent, respectively, versus angiography. In comparison, reversible defects on SPECT had a sensitivity and specificity of 75 and 71 percent.
Assessment of bypass graft patency ！ Assessment of coronary artery bypass graft (CABG) patency is often an important clinical issue. (See "Long-term outcome after coronary artery bypass graft surgery"). An early report evaluated the efficacy of EBCT in the "flow" mode to define patency of 127 mammary and saphenous bypass grafts; the results were compared to direct angiography [56]. The sensitivity of detecting an angiographically patent graft was 93.4 percent, the specificity of detecting an angiographic occluded graft was 89 percent, and the predictive accuracy was 92 percent.
Subsequent improvements in methodology and scanning techniques have resulted in an overall sensitivity of 94 to 96 percent, specificity of 100 percent, and accuracy of 95 percent for the detection of a patent saphenous vein and a free or in-situ internal mammary graft [57]. In addition, EBCT can provide noninvasive, three-dimensional visualization of coronary artery bypass grafts [58].
Restenosis after angioplasty ！ EBCT with intravenous injection of contrast is useful for the noninvasive diagnosis of restenosis after angioplasty. As an example, one study of 50 patients who underwent EBCT and coronary angiography at a mean of nine months after angioplasty reported that EBCT had a sensitivity and specificity of 94 and 82 percent, respectively [50].
Coronary artery stent patency ！ EBCT is useful for stent localization and may be an effective method for the noninvasive assessment of stent patency. As an example, one study of 202 patients with stents in 221 vessels who underwent EBCT during an intravenous bolus injection of contrast to evaluate multisection flow found that stents could be adequately visualized in 98 percent of vessels [59]. The sensitivity, specificity and positive and negative predictive values for detecting stent stenoses were 78, 98, 82, and 97 percent, respectively. (See "Intracoronary stent restenosis").
IMAGING OF THE GREAT VESSELS ！ Recent hardware and software improvements for the EBCT scanner have advanced its ability to conduct comprehensive and diagnostic examinations of the great vessels, ie, the pulmonary artery and aorta. For imaging of the great vessels, the scanner is used in the single slice high resolution or the newly introduced "spiral" mode. The ability to obtain a single scan in a short time (100 msec at or near end-diastole) has provided superior image quality by decreasing motion unsharpness of any object which might move during scan acquisition.
Aortic dissection ！ Because of the speed with which an examination can be done, EBCT, if available, is rapidly becoming the examination of choice for possible aortic dissection when the patient is clinically unstable (show figure 12 and show figure 13). (See "Clinical manifestations and diagnosis of aortic dissection").
Pulmonary embolism ！ EBCT is also of value for diagnosing pulmonary embolism. (See "Diagnostic strategies for acute pulmonary embolism"). One study used EBCT to image the pulmonary vasculature in 60 patients undergoing direct pulmonary arteriography for suspected pulmonary embolism [60]. The pulmonary vascular bed was divided into 12 zones and EBCT and angiographic findings were correlated on a patient-by-patient basis for each zone. EBCT and angiography were both negative in 36 patients and were both positive in 15. The sensitivity of EBCT was determined to be 65 percent, specificity 97 percent, positive predictive value 94 percent, and negative predictive value 82 percent. However, after review of the nine discordant cases, sensitivity and specificity of EBCT was found to approach 100 percent for clinically important acute pulmonary emboli. EBCT was equally applicable to depiction of central and peripheral emboli.
EBCT provides true visualization of pulmonary emboli and/or thrombi. As a result, serial imaging can be performed to examine changes in thrombus burden following an intervention (show figure 14).
DISEASES OF THE PERICARDIUM ！ EBCT can be used to define the entire anatomy of the pericardium and may be of greatest value in localization of effusions in the posterior areas of the heart, particularly over the right ventricle which is difficult to visualize using echocardiography. High-resolution images can define the anatomic localization and extent of pericardial thickening (show figure 15).
Additional information on the physiologic consequences of pericardial constriction may be provided from analysis of the character of the right or left ventricular time-dependent volume data when the conventional high-resolution examination is coupled with a cine study of left and right ventricular function and motion. Switching the dynamic display of the cine study to a "lung window" setting permits definition of cardiac motion transmitted to the surrounding pulmonary parenchyma. Failure of the immediately adjacent pulmonary structures to pulsate during the cardiac cycle, in the presence of a regionally or globally thickening pericardium, is virtually diagnostic of constrictive physiology.
