This Article Free upon publication Full Text (PDF) All Versions of this Article: 114/16/1761    most recent CIRCULATIONAHA.106.178458v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Budoff, M. J. Articles by Wiegers, S. E. Search for Related Content PubMed PubMed Citation Articles by Budoff, M. J. Articles by Wiegers, S. E. Pubmed/NCBI databases Medline Plus Health Information Coronary Artery Disease CT Scans Related Collections Acute coronary syndromes Chronic ischemic heart disease Cardiovascular imaging agents/Techniques CT and MRI

(Circulation. 2006;114:1761-1791.)
© 2006 American Heart Association, Inc.

AHA Scientific Statement

Assessment of Coronary Artery Disease by Cardiac Computed Tomography

A Scientific Statement From the American Heart Association Committee on Cardiovascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on Clinical Cardiology

Matthew J. Budoff, MD, FAHA; Stephan Achenbach, MD; Roger S. Blumenthal, MD, FAHA; J. Jeffrey Carr, MD, MSCE; Jonathan G. Goldin, MD, PhD; Philip Greenland, MD, FAHA; Alan D. Guerci, MD; Joao A.C. Lima, MD, FAHA; Daniel J. Rader, MD, FAHA; Geoffrey D. Rubin, MD; Leslee J. Shaw, PhD; Susan E. Wiegers, MD

Key Words: AHA Scientific Statements • calcium • atherosclerosis • coronary disease • noninvasive coronary angiography

	TABLE OF CONTENTS

Top TABLE OF CONTENTS Executive Summary Introduction 1. Coronary Artery Calcification... 2. Clinical Utility of... 3. Future Directions Conclusion References

 

Executive Summary...1761

Introduction...1764

1. Coronary Artery Calcification and Epidemiology of Coronary Calcium...1764

   1.1. Calcium Detection Methods ...1765

      1.1.1. EBCT Methods...1765

      1.1.2. MDCT Methods ...1766

   1.2. Coronary Artery Calcified Plaque ...1766

   1.3. Speed/Temporal Resolution...1767

   1.4. Studies Comparing EBCT and MDCT for Calcium Scoring...1767

   1.5. Reproducibility and Validity of Calcium Scoring ...1767

   1.6. Radiation Dose for Cardiac CT...1768

      1.6.1. Radiation Exposure During CT Angiography ...1769

2. Clinical Utility of CACP Detection ...1771

   2.1. CT Coronary Calcium and Symptomatic Patients...1771

   2.2. CT Coronary Calcification and Clinical Outcomes in Asymptomatic Individuals...1772

   2.3. Limitations...1774

   2.4. Recommendations of Professional Societies...1775

   2.5. Utilizing Coronary Calcium Measure to Improve Outcomes...1776

   2.6. Limitations of the Use of Coronary Calcium for Detecting Obstructive Disease in Asymptomatic Persons ...1777

3. Future Directions...1777

   3.1. Tracking Progression of Subclinical Atherosclerosis...1777

   3.2. Hybrid Nuclear/CT Imaging ...1779

   3.3. Contrast-Enhanced CT of the Coronary Arteries...1779

      3.3.1. Electron Beam CT ...1779

      3.3.2. Multidetector CT ...1780

   3.4. CT Angiography Applications in a Clinical Context...1782

      3.4.1. Suspected CAD...1782

      3.4.2. Follow-Up of Percutaneous Coronary Intervention...1782

      3.4.3. Follow-Up After Bypass Surgery...1782

      3.4.4. Anomalous Coronary Arteries ...1782

   3.5. Assessment of NCP ...1782

Conclusion ...1783

Disclosures ...1784

References ...1784

	Executive Summary

Top TABLE OF CONTENTS Executive Summary Introduction 1. Coronary Artery Calcification... 2. Clinical Utility of... 3. Future Directions Conclusion References

 
This scientific statement reviews the scientific data for cardiac computed tomography (CT) related to imaging of coronary artery disease (CAD) and atherosclerosis. Cardiac CT is a CT imaging technique that accounts for cardiac motion, typically through the use of ECG gating. The utility and limitations of generations of cardiac CT systems are reviewed in this statement with emphasis on CT measurement of CAD and coronary artery calcified plaque (CACP) and noncalcified plaque. Successive generations of CT technology have been applied to cardiac imaging beginning in the early 1980s with conventional CT, electron beam CT (EBCT) in 1987, and multidetector CT (MDCT) in 1999. Compared with other imaging modalities, cardiac CT has undergone an accelerated progression in imaging capabilities over the past decade, and this is expected to continue for the foreseeable future. As a result, the diagnostic capabilities at times have preceded the critical evaluation of clinical application. In this statement, the American Heart Association (AHA) Writing Group evaluates the available data for the application of cardiac CT for CAD.

Cardiac CT uses natural contrast within subjects (utilizing the different brightness of fat, tissue, contrast, and air). Noncontrast CT is a low-radiation exposure technique and, even without premedication or intravenous contrast, can determine the presence or absence of CACP in <10 minutes. The amount of CACP can be measured to provide a reasonable estimate of total coronary atheroma including calcified and noncalcified plaque. The data supporting detection of CACP as a measure of CAD are extensive. Imaging applications that detect CACP include conventional chest radiographs, cinefluoroscopy, conventional and helical CT, EBCT, and MDCT.

The majority of published studies have reported that the total amount of coronary calcium (usually expressed as the "Agatston score") predicts coronary disease events beyond standard risk factors. Although some registries used self-reported risk factor data, data from EBCT reports using measured risk factors demonstrate incremental risk stratification beyond the Framingham Risk Score (FRS). These studies demonstrate that CACP is both independent of and incremental with respect to traditional risk factors in the prediction of cardiac events. Data from Greenland et al¹ demonstrated that intermediate-risk patients with an elevated coronary artery calcium (CAC) score (intermediate FRS and CAC score >300) had an annual hard event rate of 2.8%, or a 10-year rate of 28%, and thus would be considered high risk. The best estimates of the relative risk (RR) from this study indicated that a CAC score >300 had a hazard ratio (HR) of about 4 compared with a score of 0. This would mean that the estimated risk in the intermediate-risk patient with a CAC score of 0 might be reduced by at least 2-fold while the risk of a person with a CAC score of 300+ would be increased by about 2-fold. Thus, the person with a high CAC score and intermediate FRS is now reclassified as high risk. CT information may then be used to guide primary prevention strategies, especially among individuals within the intermediate-risk category, in whom, as suggested by the AHA Prevention Conference V,² clinical decision-making is most uncertain. Individuals determined to be at intermediate risk of a cardiovascular disease (CVD) event on the basis of traditional risk factors may benefit from further characterization of their risk through measurement of their atherosclerotic burden with cardiac CT. This AHA Writing Group agrees with the statement from the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III): "In persons with multiple risk factors, high coronary calcium scores (eg, >75th percentile for age and sex) denote advanced coronary atherosclerosis and provide a rationale for intensified LDL-lowering therapy."³ Guidelines and expert consensus documents⁴ have extended the recommendation for use of coronary calcium measurements in clinically selected patients at intermediate risk for CAD (eg, those with a 10% to 20% 10-year FRS) to refining clinical risk predictions and to assessing whether more aggressive target values for lipid-lowering therapies are indicated for select patients.⁵ Asymptomatic persons should be assessed for their cardiovascular risk with such tools as the FRS. Individuals found to be at low risk (<10% 10-year risk) or at high risk (>20% 10-year risk) do not benefit from coronary calcium assessment (Class III, Level of Evidence: B). In clinically selected, intermediate-risk patients, it may be reasonable to measure the atherosclerosis burden using EBCT or MDCT to refine clinical risk prediction and to select patients for more aggressive target values for lipid-lowering therapies (Class IIb, Level of Evidence: B).

When cardiac CT is used for CACP assessment, the AHA Writing Group strongly recommends a low-dose technique using prospective ECG gating. Although alternative techniques may provide improved resolution or increased precision in measurement, data to support an enhanced predictive ability given the higher radiation exposure are limited. A minimum CT-system configuration of EBCT C150 or more up to date or MDCT 4 channel with 0.5-second gantry rotation or faster is recommended. Although virtually all of the prognostic and epidemiological data derived for CACP have been performed with EBCT, several large prospective trials have documented that cardiac CT (both MDCT and EBCT) measurements can be similarly applied across multiple centers with equally high levels of patient satisfaction and acceptance.

The utility of CACP in symptomatic patients has been widely studied and has been discussed in depth in a previous ACC/AHA statement,⁴ as well as in the AHA Cardiac Imaging Committee scientific statement "The Role of Cardiac Imaging in the Clinical Evaluation of Women With Known or Suspected Coronary Artery Disease."⁵ The test has been shown to have a predictive accuracy equivalent to alternative methods for diagnosing CAD. These studies may have been subject to referral bias, as a positive test may have been the rationale for subjecting the patient to the invasive angiogram. More comparison work between modalities is clearly needed. A positive cardiac CT examination in which any CACP is identified is nearly 100% specific for atheromatous coronary plaque. CACP can develop early in the course of subclinical atherosclerosis and can be identified histologically after fatty streak formation. CACP is present in the intima of both obstructive and nonobstructive lesions, and thus, the presence of calcified plaque on cardiac CT is not specific to an obstructive lesion. Studies using intracoronary ultrasound have documented a strong association between patterns of CACP and culprit lesions in the setting of acute coronary syndromes.

Cardiac CT studies correlating calcified plaque using EBCT technology and various methods of coronary angiography in more than 7600 symptomatic patients demonstrate negative predictive values of 96% to 100%, providing physicians with a high level of confidence that an individual without CACP (total calcium score=0) does not have obstructive angiographic CAD. The presence of CACP is extremely sensitive, albeit with reduced specificity, for diagnosing obstructive CAD (95% to 99%) in patients >40 years of age. A recent study of 1195 patients who underwent CACP measurement with EBCT and myocardial perfusion single photon emission CT (SPECT) assessment demonstrated that CACP was often present in the absence of myocardial perfusion scintigraphy (MPS) abnormalities (normal nuclear test) and that <2% of all patients with CACP <100 had positive MPS studies.⁶ This is supported by other published reports and is synthesized in a recent appropriateness criteria statement from the American Society of Nuclear Cardiology and the American College of Cardiology.^7,8 CACP measured by cardiac CT has a high sensitivity and negative predictive power for obstructive CAD but markedly limited specificity. Because calcified plaque may be present in nonobstructive lesions, the presence of CACP in asymptomatic persons does not provide a rationale for revascularization but rather for risk factor modification and possible further functional assessment. Clinicians must understand that a positive calcium scan indicates atherosclerosis but most often no significant stenosis. With exceptions, high-risk calcium scores (such as an Agatston score 400) are associated with an increased frequency of perfusion ischemia and obstructive CAD. The absence of coronary calcium is most often associated with a normal nuclear test and no obstructive disease on angiography. Coronary calcium assessment may be reasonable for the assessment of symptomatic patients, especially in the setting of equivocal treadmill or functional testing (Class IIb, Level of Evidence: B). There are other situations when CAC assessment might be reasonable. CACP measurement may be considered in the symptomatic patient to determine the cause of cardiomyopathy (Class IIb, Level of Evidence: B). Also, patients with chest pain with equivocal or normal ECGs and negative cardiac enzyme studies may be considered for CAC assessment (Class IIb, Level of Evidence: B).

Coronary calcium assessment for diagnosis of atherosclerosis and obstructive disease and for risk stratification for future cardiac events has undergone significant validation over the past 20 years. CT angiography is a noninvasive technique, performed by either EBCT or MDCT, to evaluate the lumen and wall of the coronary artery. Especially in the context of ruling out stenosis in patients with low to intermediate pretest likelihood of disease, CT coronary angiography may develop into a clinically useful tool. CT coronary angiography is reasonable for the assessment of obstructive disease in symptomatic patients (Class IIa, Level of Evidence: B). Several small studies have assessed the value of EBCT and MDCT for detecting restenosis after stent placement. At this time, however, imaging of patients to follow up stent placement cannot be recommended (Class III, Level of Evidence: C).

Where MDCT is used for CT angiography, the AHA Writing Group currently recommends a minimum of 16-slice capability, submillimeter collimation, and 0.42-second gantry rotation with retrospective ECG gating. If EBCT is used, 1.5-mm slice thickness should be used. A limitation of EBCT relative to MDCT is its lower power, with EBCT limited to 63 or 100 milliamperes/second (mAs), depending on scanner generation, which becomes important in larger patients because image quality can be affected by noise. Another advantage of MDCT is thinner slice imaging, with section thickness as small as 0.5 mm, whereas EBCT is limited to 1.5 mm. An advantage of EBCT, however, is the lower radiation dose associated with this procedure (1.1 to 1.5 mSv), compared with MDCT angiography (5 to 13 mSv).^9,10 The use of both CT modalities to evaluate noncalcified plaque (NCP) is promising but premature. There are limited data on variability but none on the prognostic implications of CT angiography for NCP assessment or on the utility of these measures to track atherosclerosis or stenosis over time; therefore, their use for these purposes is not recommended (Class III, Level of Evidence: C).

CT technology is evolving rapidly, and these radiation dose estimates are likely to decrease with modification of the hardware and scanning protocols. The clinical relevance of the radiation dose that is administered with cardiac CT is unknown. However, higher radiation doses in general are associated with a small but defined increase in cancer risk later in life. The AHA Writing Group reviewing the available literature endorses the use of a prospective ECG trigger for measurement of CACP with a slice collimation of 2.5 to 3 mm for clinical practice. EBCT systems have an effective dose of 0.7 to 1 mSv (male) and 0.9 to 1.3 mSv (female), and MDCT systems have an effective dose of 1 to 1.5 mSv (male) and 1.1 to 1.9 mSv (female). Higher radiation exposures with retrospective gating for CACP assessment preclude its use for screening. Similarly, for CT angiography, the higher radiation doses (up to 1.5 mSv for EBCT and up to 13 mSv for MDCT) prohibit the use of this test as a screening tool for asymptomatic patients. CT coronary angiography is not recommended in asymptomatic persons for the assessment of occult CAD (Class III, Level of Evidence: C).

The role of cardiac CT in measuring clinically or prognostically meaningful changes in calcified plaque over time and its correlation with other measures of coronary heart disease (CHD) is currently an area of intense investigation. Reductions in the test-to-test variability and improvements in the interreader reliability of the calcium score may allow for serial assessment of coronary calcium scores; however, more studies are required. It is difficult to justify the incremental population exposure to radiation and the cost associated with a repeat CT test to assess "change," until it is better understood what therapies may be of benefit and how clinicians should utilize this data in clinical practice. There is conflicting evidence as to whether vigorous cholesterol-lowering therapy with statins retards the rate of progression of CACP. The AHA Writing Group concluded that this potential use of cardiac CT will require additional validation before any recommendation. Serial imaging for assessment of progression of coronary calcification is not indicated at this time (Class III, Level of Evidence: C).

Cardiac CT technology is rapidly evolving. On the basis of the substantial validation data, EBCT remains the reference standard for CACP measurement.¹¹ MDCT-64 is the current standard for coronary CT angiography and NCP characterization based on publications to date.¹² The trend for improved image quality with cardiac CT is consistent. It is critical that the cardiac imaging scientific community continue to integrate evolving technological advances with best clinical practices for treatment and prevention of CVD.^7,13

An area of ongoing clinical research is the application of hybrid positron emission tomography CT (PET-CT) and SPECT-CT scanners that are currently available. This research will allow for the acquisition of metabolic and/or perfusion information as well as anatomic data, including angiographic data and data on coronary calcification. The incremental benefit of hybrid imaging strategies will need to be demonstrated before clinical implementation, as radiation exposure may be significant with dual nuclear/CT imaging. At this time, there are no data supporting the use of hybrid scanning to assess cardiovascular risk or presence of obstructive disease (Class III, Level of Evidence: C).

In summary, cardiac CT has been demonstrated to provide quantitative measures of CACP and NCP. CACP, as determined by cardiac CT, documents the presence of coronary atherosclerosis, identifies individuals at elevated risk for myocardial infarction (MI) and CVD death, and adds significant predictive ability to the Framingham Score (an index of traditional CVD risk factors). Data suggest that cardiac CT may improve risk prediction, especially in individuals determined to be at intermediate risk according to the NCEP ATP III criteria and for whom decisions concerning prevention strategies may be altered based on the test results. The use of cardiac CT angiography for noninvasive assessment of lumen stenosis in symptomatic individuals has the potential to significantly alter the management of CAD and current diagnostic testing patterns. The assessment of progression of CACP and the detection of nonobstructive NCP by cardiac CT angiography warrant further investigation.

	Introduction

Top TABLE OF CONTENTS Executive Summary Introduction 1. Coronary Artery Calcification... 2. Clinical Utility of... 3. Future Directions Conclusion References

 
The AHA has issued 2 prior statements on CAC scanning; one in 1996¹⁴ and a second (in conjunction with the American College of Cardiology [ACC]) in 2000 specifically related to EBCT.⁴ The AHA also sponsored the Prevention V Conference, which focused on the identification of the asymptomatic high-risk patient and discussed the potential role of CAC scanning.² In light of a rapidly evolving literature since the last ACC/AHA expert consensus statement (2000), the current statement will focus on new data available on using EBCT and MDCT to identify patients with coronary atherosclerosis defined by quantification of coronary artery calcification. EBCT is an especially fast form of x-ray imaging technology that can detect and measure calcium deposits in the coronary arteries.⁵ The amount of calcium detected by EBCT is related to the amount of underlying coronary atherosclerosis. During the past decade, there has been a progressive increase in the clinical use of both EBCT and MDCT scanners to identify and quantify the amount of calcified plaque in the coronary arteries. This approach has generated much interest and scrutiny for several reasons. Although coronary calcification can be quantified and calcium scores can be related to the extent and severity of atherosclerotic disease and improving CHD risk prediction, misuse or abuse of these methods as a broad-based "screening" tool has created considerable controversy.

Recently, CT scanners with subsecond image acquisition and MDCT (also referred to as multirow or multislice) capability have been studied and proposed as an alternative approach to EBCT for detecting coronary calcification owing to the greater availability of such CT scanners. This scientific statement will compare MDCT and EBCT and serve as a clinical update for the use of CACP in clinical decision-making regarding evaluations for CHD in the asymptomatic individual. Current evidence regarding noninvasive angiography using CT, as well as the future role of these techniques in monitoring atherosclerosis over time and in detecting NCP, will be reviewed.

	1. Coronary Artery Calcification and Epidemiology of Coronary Calcium

Top TABLE OF CONTENTS Executive Summary Introduction 1. Coronary Artery Calcification... 2. Clinical Utility of... 3. Future Directions Conclusion References

 
Arterial calcium development is intimately associated with vascular injury and atherosclerotic plaque. CACP is an active process and can be seen at all stages of atherosclerotic plaque development.^15–17 The long-held notion of so-called "degenerative" calcification of the coronary arteries with aging is incorrect. Since Faber¹⁸ noted in 1912 that Mönckeberg’s calcific medial sclerosis did not occur in the coronary arteries, atherosclerosis is the only vascular disease known to be associated with coronary calcification.^{4,11,14,19,20} Thus, CACP in the absence of luminal stenosis is not a "false-positive" result but rather evidence of coronary atherosclerosis.²⁰

Coronary calcification is nearly ubiquitous in patients with documented CAD^21–23 and is strongly related to age, increasing dramatically after age 50 in men and after age 60 in women (Tables 1 and 2).^24,25 However, coronary plaque and its associated coronary calcification may have only a weak correlation with the extent of histopathologic stenosis.^26,27 The degree of encroachment on the vessel lumen by the atherosclerotic plaque is largely determined by individual variations in coronary artery remodeling. However, the presence of CACP is associated with atherosclerotic plaque size.²⁶

View this table: [in this window] [in a new window]   TABLE 1. Descriptive Characteristics of the Total Electron Beam Tomographic CAC Scores in Asymptomatic Men and Women

View this table: [in this window] [in a new window]   TABLE 2. Electron Beam Tomographic CAC Score Percentiles for Men and Women Within Each Age Stratum

Rumberger and colleagues^28,29 examined 13 autopsied hearts and compared measures of CACP using EBCT as compared with direct histological plaque areas and percent luminal stenosis. These studies determined that the total area of CACP quantified by EBCT is linearly and highly correlated (r=0.90) with the total area of histological coronary artery plaque. Although the total atherosclerotic plaque burden was tracked by the total calcium burden, not all plaques were found to be calcified, and the total calcium area was approximately 20% of the total atherosclerotic plaque area. Baumgart et al³⁰ and Schmermund et al³¹ compared direct intracoronary ultrasound measures during angiography with EBCT scanning and confirmed a direct association, in vivo, of CACP score with localization and extent of atherosclerotic plaques.

The prevalence of CACP mirrors the prevalence of coronary atherosclerosis in both men and women.³² The data show the following: (1) the prevalence of CACP increases from only a small percentage in the second decade of life to nearly 100% by the eighth decade in men and women; (2) the prevalence of CACP in women is similar to that in men who are a decade younger; (3) the gender difference in prevalence with age is eliminated by approximately age 65 to 70, when the prevalence of coronary calcium in women is similar to that in men of the same age. The prevalence of CACP increases with age, paralleling the increased prevalence of coronary atherosclerosis with advancing age.

1.1. Calcium Detection Methods
This section will discuss methods related to CACP identification.

1.1.1. EBCT Methods
EBCT is a tomographic imaging device developed nearly 20 years ago specifically for cardiac imaging. Although the technique can quantify ventricular anatomy and function³³ as well as myocardial perfusion,³⁴ it is currently best known for defining and measuring CACP. Over the past decade, there have been more than 1000 articles published regarding EBCT and coronary artery imaging.

EBCT (also referred to as "EBT" and "Ultrafast-CT," General Electric, South San Francisco, Calif) uses unique technology enabling ultrafast scan acquisition times currently of 50 ms, 100 ms, and multiples of 100 ms (up to 1.5 seconds) per slice (Table 3). There have been 3 iterations of EBCT systems since their clinical introduction in the early 1980s. The core imaging methods have remained unchanged, but there have been improvements in image acquisition; in data storage, manipulation, management, and display; and in spatial resolution. The original C-100 scanner was replaced in 1993 by the C-150, which was replaced by the C-300 in 2000. The current EBCT scanner, the "e-speed" (GE/Imatron, South San Francisco, Calif) was introduced in 2003. The e-speed is a multislice scanner and currently can perform a heart or body scan in half the total examination time required by the C-150 and C-300 scanners. In addition to the standard 50-ms and 100-ms scan modes common to all EBCT scanners, the e-speed is capable of high-resolution imaging speeds as fast as 50 ms. This very short acquisition time leads to fewer motion artifacts and improved contrast-to-noise ratios.³⁵

View this table: [in this window] [in a new window]   TABLE 3. Basic Description of CT System Components

EBCT uses a stationary multisource/split-detector combination coupled to a rotating electron beam and produces serial, contiguous, thin-section tomographic scans in synchrony with the heart cycle. EBCT is distinguished by its use of a scanning electron beam rather than the traditional x-ray tube and mechanical rotation device used in current "spiral," single, and multiple-detector scanners. The electron beam is steered by an electromagnetic deflection system that sweeps it across the distant anode, a series of 4 fixed tungsten "target" rings. A stationary, single-level or dual-level arc of detectors lies in apposition to the tungsten target rings. In contrast, MDCT physically moves the x-ray tube in a circle about the patient; with EBCT, only the electron beam is moved.

Standardized methods for imaging, identification, and quantification of CAC using EBCT have been established.^4,36 The scanner is operated in the high-resolution, single-slice mode with continuous, nonoverlapping slices of 3-mm thickness and an acquisition time of 100 ms/tomogram.³⁷ Electrocardiographic triggering is done during end-systole or early diastole at a time determined from the continuous ECG tracing done during scanning.

Historically, the most common trigger time used is 80% of the R-R interval. However, this trigger occurs on or near the P wave during atrial systole, and the least cardiac motion among all heart rates occurs at 40% to 60% of the R-R interval.³⁸ Therefore, it has been demonstrated that the protocol of triggering at 80% of the R-R interval is not optimal for imaging of the coronary segments near the right or left atrium. Mao et al³⁹ compared 40% and 80% trigger delay (imaging during early compared with late diastole) and obtained an interscan variability of 11.5% versus 17.4%, respectively. For a more complete discussion on gating, see section 1.5.

1.1.2. MDCT Methods
The current generation of MDCT systems is capable of acquiring 4 to 64 sections of the heart simultaneously with ECG gating in either a prospective or retrospective mode. MDCT differs from single detector–row helical or spiral CT systems principally by the design of the detector arrays and data acquisition systems, which allows the detector arrays to be configured electronically to acquire multiple adjacent sections simultaneously (Table 3). In the current 16-row MDCT systems, 16 sections can be acquired at either 0.5- to 0.75-mm or 1- to-1.5-mm section widths, or 8 sections 2.5-mm thick.

In MDCT systems, like the preceding generation of single detector–row helical scanners, the x-ray photons are generated within a specialized x-ray tube mounted on a rotating gantry. The patient is centered within the bore of the gantry such that the array of detectors is positioned to record incident photons after they have traversed the patient. Within the x-ray tube, a tungsten filament allows the tube current to be increased (mA), which proportionately increases the number of x-ray photons for producing an image. This is a design difference with current generation EBCT systems, which use a fixed tube current.

MDCT systems have 2 principal modes of scanning, which depend on whether the patient on the CT couch is advanced in a step-wise fashion (axial, sequential, or conventional mode) or continuously moved at a fixed speed relative to the gantry rotation (helical or spiral mode). The axial mode is analogous to EBCT in using prospective ECG triggering at predetermined offset from the ECG-detected R wave and is the current mode for measuring coronary calcium at most centers using MDCT.

When prospective gating is performed, the temporal resolution of a helical or MDCT system is proportional to the gantry speed, which determines the time to complete one 360° rotation. To reconstruct each slice, data from a minimum of 180° plus the angle of the fan beam are required, typically approximately 220° of the total 360° rotation. Unless data from several consecutive heartbeats are combined, the temporal resolution is 257 ms for a 50-cm display FOV (field of view) when using a 16-row system with 0.42-second rotation. The newest 64-slice scanners now have rotation gantry speeds up to 330 ms.

1.2. Coronary Artery Calcified Plaque
Calcified plaque or calcified atheroma are the terms used in the AHA consensus paper on the definition of the advanced lesions of atherosclerosis (ie, AHA IVb lesion)—calcified plaque is a subcomponent of atheroma, not a surrogate measure.⁴⁰ CACP, as measured on cardiac CT, is defined as a hyperattenuating lesion above a threshold of 130 "Hounsfield Units" (HU) with an area 3 adjacent pixels (at least 1 mm²).

There are currently 2 CT calcium scoring systems widely used: the original Agatston method and the "volume" score method. The Agatston score method involves multiplication of the calcium area by a number related to CT density and, in the presence of partial volume artifacts, can be variable. Also, the Agatston system was designed and is properly used only when the slice thickness of the scan is 3 mm. A calcium score is reported for a given coronary artery and for the entire coronary system; however, most research studies have reported data related to the summed or total "score" for the entire epicardial coronary system.

The Agatston scoring³⁷ scale is rule based: Calculate an area for all pixels above a threshold of 130 HU, do so every 3 mm (the slice thickness and spacing used by Agatston et al), and multiply it by a density factor. Partial volume effects lead to higher peak values for small lesions (but not for large ones). If the change in peak value happens to be such that it changes the density factor, then it can, theoretically, change the score by a factor of 4. The volume method of Callister et al⁴¹ somewhat resolves the issue of slice thickness and spacing by computing a volume above threshold. The volume score is much less dependent on minor changes in slice thickness.

Current EBCT systems are now able to perform scanning at 1.5 mm, and the latest MDCT systems can provide slice thicknesses that are <1 mm. Use of thinner slices leads to higher radiation doses. In the future, a more universal scoring system may be possible that would be machine independent but, at present, data derived from MDCT should be compared with caution with those derived from EBCT. While the portability of the volume method is affected by the same issues that affect the Agatston method (slice thickness, calcium content), most studies demonstrate improved interscan reproducibility using volumetric scores for both MDCT and EBCT.

The calcium mass score has recently been reported. Basically, the mass score consists of integration of the signal for pixels above a given threshold. For a well-calibrated CT scanner, in the absence of noise, this integration (scaled by pixel volume) will give the total mineral content independent of slice thickness and spatial resolution. Although theoretically better for portability between scanners, this score has not yet undergone sufficient validation (autopsy, histology, outcomes, progression, or angiographic comparison), so its use clinically is premature.^42,43

The retention of the Agatston score has been predicated on the availability of databases for these scores, which include the availability of outcome data so clinicians understand the significance of a certain score. Volume scores are similar, while mass scores tend to be much lower values for a given patient. Adoption of newer scoring methods will depend on the availability of similar risk stratification and outcome data. Data published by Rumberger et al⁴⁴ showed that the Agatston, volume, and mass scores, when applied properly, can provide similar characterization.

1.3. Speed/Temporal Resolution
Cardiac CT is dependent on having a high temporal resolution to minimize coronary motion–related imaging artifacts. Coupling rapid image acquisition with ECG gating makes it possible to acquire images in specific phases of the cardiac cycle. Studies have indicated that temporal resolutions of 19 ms would be needed to suppress all pulmonary and cardiac motion throughout the complete cardiac cycle.⁴⁵ Current-generation cardiac CT systems can create individual images at 50 to 100 ms (EBCT) and 83 to 210 ms (MDCT), a level of resolution that cannot totally eliminate coronary artery motion in all individuals.

Motion artifacts are especially prominent in the mid right coronary artery, where the ballistic movement of the vessel may be as much as 5 to 6 times its diameter during the twisting and torsion of the heart during the cardiac cycle. Blurring of cardiac structures secondary to coronary motion increases in systems with slower acquisition speeds. It should be noted that utilizing more detectors (ie, 4 versus 8 versus 16 versus 64 detector/channel systems) does not improve the temporal resolution of the images (the rotation speed of the scanners does not change) but reduces scan time (ie, breath-hold time) and section misregistration. Generally, the higher x-ray flux (mAs = tube current x scan time) and greater number and efficiency of x-ray detectors available with MDCT devices leads to images with better signal-to-noise ratio and higher spatial resolution when compared with current EBCT scanners.

1.4. Studies Comparing EBCT and MDCT for Calcium Scoring
Several studies comparing these modalities have been published. Becker et al⁴⁶ studied 100 patients comparing MDCT with EBCT and reported a variability of 32% between the 2 modalities. Knez et al⁴⁷ studied the diagnostic accuracy of MDCT compared with EBCT in 99 symptomatic male patients (60±10 years). The mean variability between the MDCT- and EBCT-derived scores was 17%. The findings of extensive calcification and a good correlation over a large range of values do not fully address the need to measure CACP scores accurately and reproducibly in a given individual. These high correlations may not apply as well to a younger, "asymptomatic" population with generally much lower scores.⁴⁸

Carr et al⁴⁹ found agreement could be further improved by calibration of the Agatston score to an external standard. It should be emphasized that the clinical value for CAC determination is to facilitate individual risk assessment, and thus scoring for a given individual should be as accurate as possible. In epidemiologic studies of CACP in broad population groups, measures by MDCT and EBCT may well provide important insight into the atherosclerotic process, a hypothesis currently under investigation in large, population-based studies (Multi-Ethnic Study of Atherosclerosis [MESA]⁵⁰ and the Heinz Nixdorf RECALL study⁵¹).

1.5. Reproducibility and Validity of Calcium Scoring
A potential of these technologies is to estimate atherosclerosis burden and to track changes over time in order to assess efficacy of therapy.⁵² This ability to assess progression is dependent on the reproducibility of the technologies. With EBCT, the mean interscan variability, with improved methodology (early diastolic or end-systolic triggering) and hardware improvements available since 1997, has been shown to be approximately 15%, with interreader variability approximately 3% and intrareader variability <1%.^39,53–58 Achenbach demonstrated the median variability to be 5.7% using EBCT.⁵⁹

The interscan variability in several early studies using noncardiac gated MDCT (dual slice) scanners was 32% to 43%.^60,61 The literature clearly supports the use of cardiac gating to improve the measurement of CACP. A study of 75 persons using 4-slice MDCT demonstrated a mean variability of 25% for overlapping images with volume scoring, as compared with 46% for Agatston scoring without overlap.⁶² A study of 537 patients undergoing 2 studies on 4-slice MDCT with cardiac gating demonstrated a mean variability of 36% for volume scoring and 43% for Agatston scoring.⁶³ Other small studies demonstrated variabilities of 20% to 37% for Agatston scoring and 14% to 33% for volume scoring.^64–66

The National Institutes of Health/National Heart, Lung, and Blood Institute MESA is a population-based study in which 6814 men and women 45 to 84 years of age and free of clinically apparent CVD were recruited from portions of 6 US communities. Cardiac CT (EBCT-C150 and MDCT-4) examinations for measuring CACP were performed during the baseline examination between July 2000 and August 2002 using a standardized protocol.⁶⁷ Dual scans were obtained in 3551 MESA participants on an EBCT-C150 and in 3190 participants on an MDCT-4-channel system to evaluate reproducibility of the CT systems for measuring CACP. Both systems were highly concordant on the paired scan series (96% EBCT and 96% MDCT) for the presence or absence of calcified plaque.⁶⁸ Chance corrected agreement for both technologies was high with an identical kappa statistic of 0.92. When the mean absolute rescan differences were compared, adjusted for body mass index and extent of CACP, no significant difference was seen between EBCT and MDCT-4 with absolute Agatston unit values (95% confidence intervals [CIs]) by scanner type for GE-Imatron C-150 (EBCT), Siemens Volume Zoom (MDCT-4), and GE LightSpeed Plus (MDCT-4) being 15.8 (15.1,16.6), 17.5 (16.5, 18.5) and 15.7 (14.5,17.1), respectively.

One important limitation of this study was the difference in methodologies used by the scanners. The triggers in this study used 80% gating for EBCT and 50% gating for MDCT.⁶⁷ Mao et al⁵⁶ demonstrated that the Agatston score variability with EBCT decreases from 24% to 15% with use of an early diastolic trigger rather than the 80% trigger employed in the MESA study (P<0.05). The measure of CACP volume in MESA had a mean relative difference of 18% with both technologies, and this 2% improvement as compared with the Agatston score was statistically significant. This improvement in reproducibility with the volume score is consistent with this measure not accounting for information related to plaque density (ie, calcium mass). The results from MESA demonstrate good performance by both cardiac CT technologies with regard to presence, absence, and amount of CACP.

There has been some debate about using retrospective gating instead of prospective gating with MDCT to further improve reproducibility, despite the increased radiation exposure. Ohnesorge et al⁶⁹ studied 50 patients using retrospective gating, demonstrating mean variability of 23% (Agatston score) and 21% (volume score) when using nonoverlapping increments of 3 mm. A considerable reduction in rescan variability can be achieved by overlapping the slices obtained (Agatston 12%, volume 8%) with P<0.01. Considerably higher mean variability is present for the patient subgroup with low to mild calcification if image data with nonoverlapping increments are used (Agatston 42%, volume 34%). The radiation dose reported for this methodology was >2.6 mSv per patient, representing a 2-fold increase as compared with prospectively gated MDCT studies.

Van Hoe et al⁶⁵ evaluated 50 patients and reconstructed the retrospective datasets at 3 different time intervals to try to minimize interscan variability. The mean percentage interscan variability was 30±31% with the use of an image reconstruction window of 40%, 33±37% with use of an image reconstruction window of 50%, and 27±22% with use of the optimal image reconstruction window. The authors stated, "Although we used the same technique as that of Ohnesorge et al,⁶⁹ we found mean interscan variability values that were 2 to 3 times higher. No obvious explanation can be given for these striking differences."

Use of retrospective gating in an attempt to improve reproducibility with MDCT is associated with a higher radiation exposure, increased interreader variability, and markedly increased interpretation times. In 1 study of 30 patients, Agatston and volumetric scores were assessed by using 16-detector retrospectively gated MDCT.⁷⁰ For each patient, 10 datasets were created that were evenly spaced throughout the cardiac cycle. Nineteen (63%) of 30 patients could be assigned to >1 risk group depending on the reconstruction interval used to measure the calcium score. Agatston and volumetric scores both proved highly dependent on the reconstruction interval used (coefficient of variation 63%), even with the most advanced CT scanners. Accurate and reproducible quantification of coronary calcium using retrospective gating seems to require analysis of multiple reconstructions.

The AHA Writing Group proposes that the following minimum requirements be met in scanning for CAC⁷¹:

1. Use of an electron beam scanner or a 4-level (or greater) MDCT scanner
   
2. Cardiac gating
   
3. Prospective triggering for reducing radiation exposure
   
4. A gantry rotation of at least 500 ms
   
5. Reconstructed slice thickness of 2.5 to 3 mm to minimize radiation in asymptomatic persons (and to provide consistency with established results)
   
6. Early to mid-diastolic gating
   

1.6. Radiation Dose for Cardiac CT
CT uses x-rays, a form of ionizing radiation, to produce the information required for generating CT images. Although all individuals are exposed to ionizing radiation from natural sources on a daily basis, healthcare professionals involved in medical imaging must understand the potential risks of a test and balance them against the potential benefits. This is particularly important for diagnostic tests that will be given to healthy individuals as part of a disease-screening or risk-stratification program. For healthcare professionals to effectively advise individuals, they must have an understanding of the exposure involved.

The FDA, in describing the radiation risks from CT screening,⁷² used the following language:

In the field of radiation protection, it is commonly assumed that the risk for adverse health effects from cancer is proportional to the amount of radiation dose absorbed and the amount of dose depends on the type of x-ray examination. A CT examination with an effective dose of 10 millisieverts (abbreviated mSv; 1 mSv=1 mGy in the case of x-rays) may be associated with an increase in the possibility of fatal cancer of approximately 1 chance in 2000. This increase in the possibility of a fatal cancer from radiation can be compared with the natural incidence of fatal cancer in the US population, about 1 chance in 5. Nevertheless, this small increase in radiation-associated cancer risk for an individual can become a public health concern if large numbers of the population undergo increased numbers of CT procedures for screening purposes. It must be noted that there is uncertainty regarding the risk estimates for low levels of radiation exposure as commonly experienced in diagnostic radiology procedures. There are some authorities who question whether there is adequate evidence for a risk of cancer induction at low doses. However, this position has not been adopted by most authoritative bodies in the radiation protection and medical arenas.

Effective dose is an estimate of the dose to patients during an ionizing radiation procedure. It measures the total energy entered into the body and then takes into account the sensitivity of the organs irradiated. Although it has many limitations, it is often used to compare the dose from a CT examination or other examination using ionizing radiation to the background radiation a patient experiences in a year. Units are either millirem (mrem) or millisieverts (mSv); 100 mrem is 1 mSv. The estimated dose from chest x-ray is 0.04 to 0.10 mSv, and the average annual background radiation in the United States is 3 to 3.6 mSv.¹⁰

One drawback of MDCT as compared with EBCT is the higher radiation exposure to the patient (Table 4).^{10,11,73–84} The x-ray photon flux expressed by the product of x-ray tube current and exposure time (mAs) is generally higher with MDCT. For example, 200 mA with 0.5-second exposure time yields 100 mAs in MDCT versus 614 mA (fixed tube current) with 0.1-second exposure time yields 61.4 mAs in EBCT.

View this table: [in this window] [in a new window]   TABLE 4. Radiation Doses With EBCT and MDCT Coronary Angiography

Hunold et al¹⁰ performed a study of radiation doses during cardiac examinations. Coronary calcium scanning was performed with EBCT and 4-level MDCT using prospective triggering to assess each patient’s effective radiation exposure, which was then compared with measurements made during cardiac catheterization. EBCT yielded effective doses of 1.0 and 1.3 mSv for men and women, whereas MDCT using 100 mA, 140 kV, and 500-ms rotation yielded 1.5 mSv for men and 1.8 mSv for women. Invasive coronary angiography yielded effective doses of 2.1 and 2.5 mSv for men and women, respectively.

When similar protocols using single-detector-row CT (SD CT) and MDCT were compared, MDCT resulted in a dose profile approximately 27% higher than that from SD CT in the plane of imaging (8.0 versus 6.3 mGy) and 69% higher adjacent to the plane of imaging (6.8 versus 4.0 mGy).⁷⁴ The individual doses to the kidneys, uterus, ovaries, and pelvic bone marrow were 92% to 180% higher with MDCT than with SD CT. The authors concluded, "With image noise constant between SD CT and MDCT, the radiation dose profile both inside and outside the plane of imaging was higher with MDCT than with SD CT. Organ dose also was higher with MDCT than with SD CT."

Because retrospective gating exposes the patient to significantly higher radiation, several techniques have been implemented to reduce those exposures. Mahnken et al⁷⁵ studied body-weight dosing (reducing the radiation exposure based on body size) and measured the mean of the effective radiation dose with and without dose modulation. The radiation dose for a calcium scan using MDCT was 4.44 mSv (range, 3.28 to 5.88 mSv) for women and 3.01 mSv (range, 2.52 to 4.18 mSv) for men, whereas with dose modulation, the mean of the calculated radiation dose was 3.34 mSv (range, 2.39 to 3.83 mSv) for women and 2.66 mSv (range, 2.09 to 3.53) for men.

1.6.1. Radiation Exposure During CT Angiography
MDCT angiography requires retrospective gating, associated with significantly greater radiation exposures, to acquire images. Radiation doses of cardiac MDCT scans reported in the literature vary a great deal depending on the scan parameter settings. The tube voltages in the published protocols vary from 120 to 140 kVp, and the tube currents vary from 150 to 600 mA.⁷⁶ In contrast, the scan settings of EBCT used for cardiac imaging were fixed, in the older technology, to 130 kVp, 630 mA, and 100-ms exposure time. These EBCT settings have been somewhat altered, however, by the newer e-Speed technology, with both higher kVp and mA potential (140 kVp, 1000 mA). Newer protocols for MDCT angiography allow for increased power utilization, with settings as high as 900 mA possible. These higher settings will further increase the radiation dose, which is an issue to be considered when performing these protocols.

Pitch is calculated as table speed divided by collimator width. A low pitch (low table speed) allows for overlapping data from adjacent detectors. Most commonly, physicians use a low table speed and thin collimation width, leading to a large number of very thin axial slices, which are of great value for imaging the heart with high resolution. The tradeoff for these overlapping images is a markedly higher radiation exposure.⁷⁶ These protocols are also responsible for substantial increases in radiation doses, especially for the MDCT systems, with dose estimates of up to 11 to 13 mSv per study (Table 4).

Two studies have measured the radiation doses for CT angiography, comparing EBCT and 4-slice MDCT. The first reported EBCT angiography doses of 1.5 to 2.0 mSv, MDCT angiography doses of 8 to 13 mSv, and coronary angiography doses of 2.1 to 2.3 mSv, while the second reported EBCT angiography doses of 1.1 mSv and MDCT doses of 9.3 to 11.3 mSv.^9,10 Newer MDCT studies report that radiation doses are similar with 16-level multidetector scanners and higher with 64 MDCT.^77,78 Studies estimate radiation exposure for 16-row MDCT at 8.8 mSev for a 16x0.75-mm scan protocol with a pitch of 0.28 and power of 370 mA⁷⁹ and at 13 and 18 mSv (for men and women, respectively) with 64-row MDCT.⁸⁰ It should be noted that nuclear imaging has similar radiation exposure doses for cardiac studies (8 to 12 mSv).⁸¹ Specifically, technetium studies are on the lower end of this spectrum (6 to 8 mSev on average), and thallium studies have been reported as high as 27 mSv.⁸²

With the retrospective ECG-gating mode, scan data are acquired and available for the entire phase of the cardiac cycle. In most cases, however, the scan data used for image reconstruction are selected only during the diastolic phase. This implies that a high tube current is required only during the diastolic phase and that a low tube current is acceptable during the remaining cardiac phase. Modulating the tube current online with prospective ECG control (dose modulation) is reported to help reduce radiation exposure substantially without decreasing diagnostic image quality.^83,84

For MDCT coronary angiography, dose modulation techniques reduce radiation exposures⁸⁴ and should be employed whenever possible. The effects of dose reduction are more pronounced for lower heart rates. Also, using the lowest necessary mA during each study will also help limit radiation exposures during these procedures. For MDCT, increased numbers of detectors allow for better collimation and spatial reconstructions. Having more of the heart visualized simultaneously will also allow for reductions in the contrast requirements and breathholding for the patient, further improving the methodology.

In summary, CT technology is evolving rapidly and radiation exposures are likely to be reduced with modification of the hardware and scanning protocols. The clinical relevance of differences in radiation dose between different technologies is unknown, but most would agree that less radiation is better for patients than more radiation. The AHA Writing Group, reviewing the available literature, endorses the use of a prospective ECG trigger for measurement of CACP with a slice collimation of 1.5 to 3 mm for clinical practice. EBCT systems have an effective dose of 0.7 to 1 mSv (for men) and 0.9 to 1.3 mSv (for women), and MDCT systems have an effective dose of 1 to 1.5 mSv (for men) and 1 to 1.8 mSv (for women).^9,10,76 For CT angiography, the higher radiation doses suggest the need for greater forethought when using these tests, and use of these higher radiation exposure tests in asymptomatic persons for screening purposes is not currently recommended.

	2. Clinical Utility of CACP Detection

Top TABLE OF CONTENTS Executive Summary Introduction 1. Coronary Artery Calcification... 2. Clinical Utility of... 3. Future Directions Conclusion References

 
This is the first time that the AHA evidence-based scoring system (see http://circ.ahajournals.org/manual/manual_IIstep6.shtml) has been incorporated into the AHA’s evaluation of cardiac CT. The purpose of the scoring system is to assist the clinician in interpreting these recommendations and formulating treatment decisions. The system is based on both a classification of recommendations and the level of evidence. Each treatment recommendation has been assigned a class and a level of evidence. The use of this system should support but not supplant the clinician’s decision making in the management of individual patients’ cases.

Classification of Recommendations

• Class I: Conditions for which there is evidence, general agreement, or both that a given procedure or treatment is useful and effective.
  
• Class II: Conditions for which there is conflicting evidence, a divergence of opinion, or both about the usefulness/efficacy of a procedure or treatment.
  •    Class IIa: Weight of evidence/opinion is in favor of usefulness/efficacy.
    
  •    Class IIb: Usefulness/efficacy is less well established by evidence/opinion.
    
  
  
• Class III: Conditions for which there is evidence, general agreement, or both that the procedure/treatment is not useful/effective and in some cases may be harmful.
  

Level of Evidence

• Level of Evidence A: Data derived from multiple randomized clinical trials
  
• Level of Evidence B: Data derived from a single randomized trial or nonrandomized studies
  
• Level of Evidence C: Consensus opinion of experts
  

2.1. CT Coronary Calcium and Symptomatic Patients
The utility of measuring CAC in symptomatic patients has been widely studied and discussed in depth in a previous ACC/AHA statement,⁴ as well as in the recent AHA Cardiac Imaging Committee consensus statement, "The Role of Cardiac Imaging in the Clinical Evaluation of Women With Known or Suspected Coronary Artery Disease."⁵ A positive EBCT study (indicating the presence of CACP) is nearly 100% specific for atheromatous coronary plaque^4,5,26–28 but is not highly specific for obstructive disease, as both obstructive and nonobstructive lesions have calcification present in the intima. The presence of CACP by EBCT is extremely sensitive, however, for obstructive (>50% luminal stenosis) CAD (95% to 99%).^4,5,20–22 This has led to much confusion over the interpretation of CACP as a diagnostic test.

A large multicenter study on EBCT for diagnosis of obstructive CAD in symptomatic persons (n=1851) found that the sensitivity and specificity of CACP were 96% and 40%, respectively.²² However, increasing the cutpoint for calcification markedly improves the specificity. In this same study, increasing the CACP cutpoint to >80 decreased the sensitivity to 79%, while increasing the specificity to 72%. In another large study (n=1764) comparing CACP to angiographic disease, use of a CACP score >100 led to a sensitivity of 95% and a specificity of 79% for the detection of significant obstructive disease by angiography.²³ Summing these 2 large studies (n=3615) leads to a sensitivity of 85% with a specificity of 75%. In a meta-analysis of 44 studies, technetium stress was found to have a mean sensitivity of 87% and mean specificity of 64%,⁸⁵ similar to the results of CACP. Thus, CACP measurements have a similar accuracy to other commonly accepted modalities for diagnosis of obstructive CAD by angiography (Table 5). For all diagnostic accuracy literature, one must be concerned about posttest referral bias, whereby positive tests are the cause for the referral to the catheterization laboratory. If the test is allowed to be part of the referral pattern, the sensitivity will increase and the specificity will decrease. However, for the 3 studies of EBCT, imaging was performed after the patient was referred for an invasive angiogram. The reason for the low specificity with CAC testing is the presence of CAC in nonobstructive as well as obstructive lesions.

View this table: [in this window] [in a new window]   TABLE 5. Sensitivity and Specificity of Diagnostic Tests for Evaluation of CAD

In direct comparison studies, EBCT coronary calcium has been shown to be comparable to nuclear exercise testing in the detection of obstructive CAD.^87,88 The accuracy of EBCT is not limited by concurrent medication, the patient’s ability to exercise, baseline ECG abnormalities, or existing wall motion abnormalities. Patients whose studies prove negative would be less likely to undergo invasive angiography. More comparison work between modalities is clearly needed.

Data also support a complementary role for coronary calcium and MPS measurements. A recent study of 1195 patients who underwent CACP measurement and MPS assessment demonstrated that the presence of CACP was the most powerful predictor that a nuclear test would be positive for ischemia and that <2% of all patients with CACP <100 had positive MPS studies.⁶ EBCT, owing to its high sensitivity for flow-limiting CAD, may be useful as a filter before angiography or stress nuclear imaging, with more caution in younger patients. Knez et al⁸⁹ studied 2115 consecutive symptomatic patients (n=1404 men, mean 62±19 years of age) with no prior diagnosis of CAD, finding CAC in more than 99% of patients with obstructive CAD. No calcium was found in 7 of 872 men (0.7%) and in 1 of 383 women (0.02%) who had significant luminal stenosis on coronary angiography. Seven of these 8 patients with missed obstructive disease and scores of 0 were <45 years of age.

Recent ACC/ASNC appropriateness criteria support that a low calcium score precludes the need for MPS assessment and a high score warrants further assessment.⁸ These appropriateness criteria suggest nuclear testing may generally be inappropriate in patients with calcium scores <100, as the probability of obstruction or abnormal scan is very low. However, more recent evidence suggests that MPS may be indicated in patients with diabetes and those with a family history of CAD who have a calcium score <100.^90–92 For the remaining asymptomatic patients, a person with an Agatston score >400 may benefit from functional testing to detect occult ischemia. The use of functional testing is paramount in determining the need for revascularization, as functionally insignificant lesions do not benefit from revascularization.

CACP may also be considered in determining the etiology of cardiomyopathy (Class IIb, Level of Evidence: B). The clinical manifestations of patients with ischemic cardiomyopathy are often indistinguishable from those with primary dilated cardiomyopathy. One large study of 120 patients with heart failure of unknown etiology demonstrated that the presence of CACP was associated with 99% sensitivity for ischemic cardiomyopathy.⁹³ Another study demonstrated similarly high sensitivity using dual CT to differentiate ischemic from nonischemic cardiomyopathy.⁹⁴ Direct comparison studies have demonstrated this methodology to be more accurate than echocardiography and MPS techniques.^95,96 Additional comparative prognostic and diagnostic evidence is required to evaluate the role of CT as compared with conventional stress imaging techniques, as well as an assessment developing marginal cost-effectiveness models.

Another potential application of CACP relates to the triage of patients with chest pain. Three studies have documented that CACP is a rapid and efficient screening tool for patients admitted to the emergency department with chest pain and for whom ECG findings have been nonspecific.^97–99 These studies show sensitivities of 98% to 100% for identifying patients with acute MI and very low subsequent event rates for persons with negative test results. The high sensitivity and negative predictive value may allow early discharge of those patients with nondiagnostic ECG and negative CACP scans (score=0). Long-term follow-up of this cohort demonstrates patients without demonstrated CACP at the time of the emergency visit are at very low risk of subsequent events.⁹⁷

Recommendation: Patients with chest pain with equivocal or normal ECGs and negative cardiac enzyme studies may be considered for CAC assessment (Class IIb, Level of Evidence: B).

For the symptomatic patient, exclusion of coronary calcium may be an effective filter before invasive diagnostic procedures or hospital admission. EBCT studies of more than 7600 symptomatic patients undergoing cardiac catheterization demonstrate negative predictive values of 96% to 100%, allowing physicians a high level of confidence that an individual with no coronary calcium (score=0) does not have obstructive angiographic disease.^21–23,89 Calcium scores <100 are associated with a very low probability (<2%) of abnormal perfusion on nuclear stress tests⁶ and <1% probability of significant obstruction (>50% stenosis) on cardiac catheterization.^6,21–23,89 While models suggest this is a cost-effective algorithm, further testing and prospective analysis are required.^100,101

Recommendation: Coronary calcium assessment may be reasonable for the assessment of symptomatic patients, especially in the setting of equivocal treadmill or functional testing (Class IIb, Level of Evidence: B).

2.2. CT Coronary Calcification and Clinical Outcomes in Asymptomatic Individuals
Calcification of the coronary arteries occurs in approximate proportion to the severity and extent of coronary atherosclerosis.¹⁰² In a landmark study of atherosclerosis, persons dying of coronary disease were found to have 2-fold to 5-fold greater amounts of coronary calcification than age-matched controls dying accidentally or of other natural causes.¹⁰³ Eight studies have examined the prognostic accuracy of CACP score by EBCT.

The first and longest study of EBCT scanning of the coronary arteries, the South Bay Heart Watch study,^1,104–106 began in 1990 as a prospective study of the prognostic accuracy of cardiac fluoroscopy in 1461 asymptomatic, high-risk individuals. In 1992, 1289 study participants (mean age 66±8 years) underwent EBCT scanning. Although an early analysis revealed no incremental advantage of EBCT scanning over conventional risk factor assessment for hard coronary events,¹⁰⁴ long-term (median=7.0 years) follow-up has demonstrated that the CACP score adds predictive power beyond that of standard coronary risk factors and C-reactive protein.^1,105 In multivariable models, a CACP score >300 was highly statistically significant and independently predictive of fatal or nonfatal MI, compared with a score of 0 (HR=3.9, P<0.001). In this study, patients with an FRS of 16% to 20% and a CAC score 300 had an annual event rate of 2.8%. This patient group would therefore have the 10-year event rate 20% that indicates high risk by current NCEP criteria.

From a retrospective cohort study of 632 asymptomatic persons (mean age 52±9 years, mean follow-up =2.7 years), the annual rate of nonfatal MI or CHD death increased from 0.045% in the lowest quartile of calcium scores to 2.7% among subjects in the highest quartile of calcium scores (a 59-fold increase).¹⁰⁷ Thus, patients with high calcific plaque burden did exceed the high-risk threshold (>2% per year hard cardiac event rate). These investigators demonstrated that EBCT added incremental benefit over and above standard coronary risk factors for risk prediction.¹⁰⁸

Another study of 1172 asymptomatic persons (mean age 53±11 years, follow-up=3.6 years) demonstrated that a calcium score >160 was highly predictive of nonfatal MI or CHD death with an elevated risk 23.3-fold higher for CACP scores >160 versus CACP <160.¹⁰⁹ This study did not measure risk factors but did multivariable analysis to adjust for self-reported cardiovascular risk factors.

Wong et al¹¹⁰ reported on 3.3-year follow-up in 926 asymptomatic persons (mean age 54±10 years). The calcium score predicted events independently of age, gender, and other cardiovascular risk factors (risk-adjusted RR=8.8 for scores in the fourth versus first quartile). Kondos et al¹¹¹ reported 37-month follow-up in 5635 initially asymptomatic low-risk to intermediate-risk adults (mean age 51±9 years). While follow-up was only obtained in 64% of patients, multivariable modeling demonstrated that patients with scores >170 (top quartile of scores) had an RR for hard cardiac events of 7.24-fold (95% CI, 2.01 to 26.15) as compared with patients without CACP. Finally, in a larger cohort of 10 377 asymptomatic individuals undergoing cardiac risk factor evaluation and CACP measurement with EBCT, a study with a mean follow-up of 5.0 years¹¹² used a risk-adjusted model to show that CACP was an independent predictor of all-cause mortality (P<0.001).

Shemesh et al¹¹³ reported on a 3.8-year follow-up of 446 hypertensive patients prospectively followed up after risk factor measurement and CACP. CACP (total coronary calcium score >0) independently predicted cardiovascular events with an odds ratio (OR) of 2.76 (95% CI 1.09 to 6.99, P=0.032). Of note, this was the first prognostic study with MDCT (using a dual-slice CT system).

A significant limitation to a number of the early studies, with the exception of the South Bay Heart Watch Study, is that they were retrospective and did not include measured risk factors. However, 6 recently reported prospective studies, all with measured risk factors, now demonstrate the independent and incremental prognostic value of CAC measurement over the FRS.

The St. Francis Heart Study is a prospective observational study of 4613 subjects (59±5 years of age) with 4.3 years of follow-up.¹¹⁴ A calcium score >100 predicted cardiovascular events, all coronary events, and the sum of nonfatal MI or CHD death events with RR ratios ranging from 9.2 to 11.1. Of note in this prospective series, the calcium score predicted cardiovascular events independently of standard risk factors and high-sensitivity C-reactive protein (P<0.004). Additionally, the calcium score also had improved event classification when compared with the FRS (area under the ROC curve 0.79±0.03 versus 0.68±0.03, P=0.0006).

Similarly, in a younger cohort of asymptomatic persons, the Prospective Army Coronary Calcium (PACC) Project¹¹⁵ reported 3-year mean follow-up in 2000 participants (mean age 43 years). Participants were evaluated with measured coronary risk variables and coronary calcium detected by EBCT. Coronary calcium was associated with an 11.8-fold increased risk for incident CHD (P<0.002) in a Cox model controlling for the FRS. Among those with CAC, the risk of coronary events increased incrementally across tertiles of coronary calcium severity (HR 4.3 per tertile). A family history of premature CHD was also predictive of incident events. A major limitation of this study is that coronary events occurred in only 9 of the men who participated, with no events reported in women. Thus, the CIs around the RR estimates were rather large. The authors concluded, "In young, asymptomatic men, the presence of CAC provides substantial, cost-effective, independent prognostic value in predicting incident CHD that is incremental to measured coronary risk factors."

The Rotterdam Heart Study¹¹⁶ investigated a general, asymptomatic population of 1795 elderly subjects. Participants who were followed up prospectively (mean age=71 years) had CAC and measured risk factors. During a mean follow-up of 3.3 years, 88 cardiovascular events, including 50 coronary events, occurred. The multivariable-adjusted RR of coronary events was 3.1 (95% CI, 1.2 to 7.9) for calcium scores of 101 to 400, 4.6 (95% CI, 1.8 to 11.8) for calcium scores of 401 to 1000, and 8.3 (95% CI, 3.3 to 21.1) for calci