Cardiac computed tomography (CT) has evolved greatly over the last 20 years. This article reviews its current clinical uses, and describes some of the potential for even greater utility in the near future. Electron beam tomography (EBT) was developed 20 years ago specifically for cardiac imaging. Although the technique can quantify ventricular anatomy and function1 as well as myocardial perfusion2 it is currently best known for defining and measuring coronary calcified plaque (coronary artery calcification (CAC)). Over the past decade, there have been well over 1,000 articles published regarding EBT and coronary artery imaging. Standardized methods for imaging, identification and quantification of CAC using EBT have been established.3 The scanner is operated in the high-resolution, single slice mode with continuous, non-overlapping slices of 3mm thickness and an acquisition time of 100msec per tomogram.4 Electrocardiographic triggering is carried out during end-systole or early diastole at a time determined from the continuous ECG tracing carried out during scanning.
Since 1999, multidetector CT has been utilized to evaluate the heart. Advances in temporal resolution and imaging protocols have allowed cardiac applications to become more routine. The newer generations of multi-detector- row (MDCT) systems are capable of acquiring 16-64 sections of the heart simultaneously with ECG gating in either a prospective or retrospective mode. In the current 16-row MDCT systems, 16 sections can be acquired at either 0.5-0.75mm or 1-1.5mm section widths or eight sections 2.5mm thick. Thin slices (sub-mm) as well as rapid acquisition allow for cardiac imaging with reduced motion artifacts.
Coronary Calcium Assessment
During the past decade, there has been a progressive increase in the clinical use of EBT scanners to identify and quantify the amount of calcified plaque in the coronary arteries. CAC can be quantified and calcium scores can be related to extent and severity of atherosclerotic disease and improving coronary heart disease (CHD) risk prediction. A positive EBT study (presence of CAC) is nearly 100% specific for atheromatous coronary plaque. Since both obstructive and non-obstructive lesions have calcification present in the intima, CAC is not specific to obstructive disease; thus, a positive EBT does not always imply significant stenosis. However, the presence of CAC is extremely sensitive for obstructive (more than 50% luminal stenosis) coronary artery disease ((CAD) 95% to 99%).5 For clinicians, evidence of CAC is highly sensitive but less specific for obstructive CAD. Importantly, EBT studies of over 5,500 symptomatic patients demonstrate negative predictive values (NPVs) of 96% to 100%, allowing physicians a high level of confidence that an individual with no coronary calcium (score=0) has no obstructive angiographic disease.5
CAC occurs in approximate proportion to the severity and extent of coronary atherosclerosis.6 From a synthesis of both retrospective and prospective cohort studies, there appears to be a directly proportional relationship between the risk and the extent of CAC.7 The risk associated with increasing CAC is approximately 10-fold, found to be both incremental and independent of traditional cardiac risk factors. All published studies have reported that the total amount of coronary calcium predicts coronary disease events beyond standard risk factors.8
Tracking Progression of Subclinical Atherosclerosis
Several studies have shown that serial EBT scanning can be utilized to follow the progression of atherosclerosis.A study measured the change in CAC in 495 asymptomatic subjects submitted to sequential EBT scanning.9 Statins were initiated in all patients after their initial EBT scan, with a follow-up of 3.2┬▒0.7 years. Interestingly, mean LDL level did not differ between patients experiencing myocardial infarction (MI) compared with those event-free patients (118┬▒25mg/dL compared with 122┬▒30mg/dL, MI versus no MI). On average, MI subjects demonstrated an annual rate of CAC change of 42%┬▒23%, while event-free subjects showed a 17%┬▒25% yearly change (p=0.0001).The associated relative risk for acute MI for patients exhibiting more than 15% CAC progression was elevated 17.2-fold (95% confidence interval (CI): 4.1 to 71.2) when compared with those without CAC progression (p<0.0001). Continued progression of CAC appears to be an independent risk factor for future events; however, future studies are needed. On-going studies with the National Institutes of Health (NIH) should help to answer these questions.
Non-invasive CT Angiography
The most exciting application of cardiac CT (CCT) is the use as a non-invasive angiogram. While magnetic resonance imaging (MRI) for this application has been limited (due to spatial and temporal resolution issues), CT provides thin slice acquisition and high-contrast resolution, a unique combination that permits visualization of the coronary artery lumen, coronary atherosclerotic plaque, and coronary stenoses. Multiple studies have evaluated the accuracy of EBT and MDCT 'coronary angiographyÔÇÖ for the assessment of coronary artery stenoses, and generally found sensitivities and specificity in the 90% to 92% range. Newer studies with 64 MDCT confirm this high accuracy.9 The increased collimation width (more coverage of the heart per rotation) and greater number of slices obtained simultaneously allow for shorter examination times, reducing both the breath-hold and contrast requirements.
CT angiography (CTA) is increasingly being clinically used to rule out significant disease, and to decide when patients do not need further invasive angiography (see Figure 1).Virtually all published studies (dating back to as early as 1995) have demonstrated a very high negative predictive value (NPV).10 Thus, a 'normalÔÇÖ CT coronary angiography permits the ruling out of the presence of hemodynamically relevant coronary artery stenoses with a high degree of reliability.
Numerous studies have shown that EBT and MDCT permit assessment of coronary bypass graft occlusion and patency with high accuracy (see Figure 2). In most studies, the accuracy to detect bypass occlusion approached 100%.11 Another application is anomalous coronary arteries and congenital heart disease. The 3-dimensional (3-D) nature of CT coronary angiography data sets allow for an exact analysis of anomalous coronary arteries. Both for EBT and MDCT, published studies demonstrate that the analysis of coronary anomalies is straightforward and exact.12
Current uses of non-invasive CTA include:
- following the non-diagnostic stress test;
- with those people with intermediate likelihood of CAD (where the step to coronary angiography might be premature);
- with symptomatic people post-coronary angioplasty and possibly post-stent;
- evaluating graft patency post-coronary artery bypass grafting (CABG); and
- early detection of obstructive CAD in the high-risk person.
Given the current utility of these techniques, a rapid growth in both the knowledge and experience with non-invasive angiography, leading to much wider clinical applications for the assessment of obstructive CAD, can be expected.
'Soft PlaqueÔÇÖ Assessment
In addition to identifying lesions with significant luminal narrowing, there is also interest in visualizing and characterizing coronary artery plaques beyond the mere assessment of calcium or luminal stenosis. Improved spatial and temporal image acquisition with sub-mm slice collimation has facilitated atherosclerotic plaque detection with MDCT. Plaque with density below the vessel contrast is defined as non-calcified plaque. Conversely, structures with densities above the adjacent vessel lumen are considered calcified.13 The clinical implications of non-calcified plaque detection (soft plaque) on CT are currently unknown.
CTA allows for reproducible quantitative measurement of left ventricular ejection fraction (LVEF) and right ventricular ejection fraction (RVEF),14 ventricular volumes,15 ventricular mass,16 wall thickness, and regional wall motion17 simultaneously. These data can be integrated with coronary artery assessment for a comprehensive assessment of the role that CAD plays in the cardiomyopathic process in a single study.
Non-invasive quantitation of myocardial blood flow is also possible by evaluating flow patterns of iodinated contrast on CT.Myocardial blood flow is proportional to the peak iodine concentration in the myocardium after intravenous (IV) injection of contrast medium. Based on the principle that blood flow is proportional to iodine concentration during contrast medium infusion, acute MI can be imaged by CT as a region of little or no iodine. This technique has been used to detect MI, and quantitate the infarct size, as well as the patency of the infarct vessel, using both flow and 3-D techniques. Complications of MI, including ventricular septal defects, thrombi, aneurysms, and pericardial effusions, can all be detected by CT.
The same techniques that allow visualization of coronary arteries also visualize the coronary veins. The coronary venous anatomy has become increasingly important as many interventional procedures use the coronary veins to obtain venous access to the left atrium and LV. CTA is effective for visualization of the coronary sinus and its tributaries. CTA can provide detailed assessment of the coronary venous anatomy, with coronary sinus dimensions, branch vessel locations, diameters, angulations off from the coronary sinus/great cardiac vein, and associated myocardial segment location.18
Characterization of pulmonary venous anatomy is important to catheter-based therapies for atrial fibrillation. Procedural efforts have focussed on the ablation of pulmonary veins. CTA and MRI provide detailed information on pulmonary vein location, variation, size, and complexity, which are often difficult to visualize by other techniques.
This information is important for the ablation of pulmonary vein triggers and electrical isolation of pulmonary veins (see Figure 3).19 Endoscopic views of the left atrium can be achieved through software advances to visualize the complexity of each pulmonary vein. Several companies can now incorporate the images obtained with CCT (dicom format) into the catheterization laboratory and merge them with fluoroscopic images to help guide the procedure. The images provide the electrophysiologist with higher resolution images to perform ablations, potentially improving outcomes and shortening the procedure.
Another new application relates to the ability to see both cardiac valves, arteries, and veins simultaneously. New transcutanous devices are being implanted by cardiologists (including patent foramen and atrial septal closure devices), and the advance visualization of the size of the defect and the associated structures are assisting with proper device selection and decreasing the time (and possibly improving success rates) associated with the procedure. Several companies are developing new methods to treat heart failure (HF) and valve regurgitation, and these techniques require careful consideration of the pre-existing anatomy of the heart. CCT, by measuring both the baseline function (ejection fraction and mass) as well as the relationship of the cardiac veins and valve planes, greatly facilitates the performance of these new procedures, and is becoming routine prior to these new device implants. MRI, another technique commonly employed in this manner, is more challenging due to the exclusion of anyone with prior devices implanted, a situation that is becoming much more common in current cardiology patients, particularly those with congestive failure.
CCT is clearly the most quickly advancing and growing field within current cardiology. Non-invasive angiography, screening for heart disease, and applications related to EP and device implants will make this an integral part of most cardiology practices. A downside to CCT is the need for ionizing radiation. For calcium scoring, EBT systems have an effective dose of 0.7-1mSv and MDCT have an effective dose of 1-1.5mSv.20 For CTA, the higher radiation doses (up to 1.5mSev for EBT and up to 13mSev for MDCT) are similar to the current doses given during a cardiac nuclear test.21 CT technology is evolving rapidly and these radiation dose estimates are likely to fall with modification of the hardware and scanning protocols.