Computed Tomography and Cardiac Magnetic Resonance in Ischemic Heart Disease

Ischemic heart disease is a complex disease process caused by the development of coronary atherosclerosis, with downstream effects on the left ventricular myocardium. It is characterized by a long preclinical phase, abrupt development of myocardial infarction, and more chronic disease states such as stable angina and ischemic cardiomyopathy. Recent advances in computed tomography (CT) and cardiac magnetic resonance (CMR) now allow detailed imaging of each of these different phases of the disease, potentially allowing ischemic heart disease to be tracked during a patient’s lifetime. In particular, CT has emerged as the noninvasive modality of choice for imaging the coronary arteries, whereas CMR offers detailed assessments of myocardial perfusion, viability, and function. The clinical utility of these techniques is increasingly being supported by robust randomized controlled trial data, although the widespread adoption of cardiac CT and CMR will require further evidence of clinical efficacy and cost effectiveness.

preclinical phase. The first pathological abnormality, the fatty streak, can be observed as early as the second decade of life (1). Ultimately, these streaks develop into mature atherosclerotic plaques consisting of a central lipid core covered by a fibrous cap.
As the plaque grows, the affected vessel expands in an outward direction, preserving both the luminal diameter and blood flow, in a process known as positive remodeling.
Consequently, even large plaques can be accommodated without producing symptoms and without being identified on invasive angiography or stress testing. Eventually, the plaque begins to grow into the lumen of the vessel, obstructing blood flow and causing myocardial ischemia and symptoms of angina pectoris. Importantly, although the degree of luminal stenosis is closely related to the development of myocardial ischemia, multiple other factors such as entrance effects, friction, and turbulence, also contribute to increased flow resistance across a particular stenosis (2). Moreover, it is now well established that most myocardial infarctions (MIs) arise from plaques that are nonobstructive on antecedent angiography, in part related to the much greater prevalence of these lesions (3,4).
Atherosclerotic plaques can remain quiescent for years; indeed, most will remain subclinical during a patient's lifetime. However, individual plaques can have a major clinical impact when their surface becomes disrupted, initiating thrombus formation and, potentially, vessel occlusion and acute coronary syndromes (ACS). Most commonly, ACS is triggered by acute fibrous cap rupture, which exposes the thrombogenic, tissue factor-rich lipid core to circulating blood. Alternatively, plaque erosion of the endothelium overlying the fibrous cap can lead to the formation of a platelet-rich thrombus, accounting for up to 30% of MIs (5). However, ACS is not an inevitable consequence of fibrous cap disruption. Indeed, subclinical plaque rupture appears common, with up to 70% of obstructive coronary plaques containing histological evidence of previous rupture and subsequent repair (6). The magnitude of the thrombotic response to plaque disruption is therefore also important and is governed by multiple factors, including the thrombogenicity of the blood, flow along the vessel, and constituents of the plaque.
Plaques that rupture and precipitate acute thrombotic events frequently have certain key characteristics on histological examination, including a thin fibrous cap (<65 mm), positive remodeling, a large necrotic core, inflammation, microcalcification, angiogenesis, and plaque hemorrhage. Each of these adverse plaque characteristics, therefore, represents a potential imaging target for identifying plaques at high risk of rupture. Although prospective data have suggested that most of these so-called vulnerable plaques either heal or rupture subclinically rather than cause ACS (7), such features rarely exist in isolation and can serve as a marker of patients with advanced and metabolically active atheroma (7,8).
Invasive imaging studies have demonstrated the presence of multiple high-risk plaques across the coronary vasculature in patients with ACS (9), and postmortem studies have demonstrated that multiple coronary thrombotic events are present at the time of fatal MIs (10). Both of these observations support a pancoronary vulnerability to atherosclerotic plaque rupture. On this basis, interest persists in detecting vulnerable plaque, not because these lesions will necessarily rupture themselves (11), but rather as a means of identifying vulnerable patients, those subjects with active atheroma and a propensity to develop multiple high-risk plaques over time, one of which may ultimately cause an event.
Both myocardial ischemia and infarction can have a profound effect on the structure and function of the left ventricular myocardium. MI results in tissue necrosis and ultimately irreversible areas of scarring, reducing the ability of the heart to both contract and relax. MI also results in areas of stunned myocardium, regions of hypokinesia that are damaged but not infarcted, with the potential to regain function.
Similarly, severe myocardial ischemia can result in hibernating myocardium, areas with impaired function that can be restored if blood flow is improved.
Ultimately, these insults to cardiac function can lead to congestive cardiac failure and development of ischemic cardiomyopathy.

IMAGING WITH CT AND CMR
The successful application of CT and CMR imaging to the heart was delayed compared with application to static organ systems, principally due to the heart's complex motion during both the cardiac and the respiratory cycles. However, advances in scanner technology now offer robust methods for motion correction and much improved spatial and temporal resolution, heralding a new era of noninvasive cardiac imaging. Each technology has different strengths and weaknesses ( Table 1) that can potentially provide complementary information regarding ischemic heart disease. We focus here on the ability of these technologies to image plaque burden, high-risk plaque characteristics, and

CORONARY PLAQUE BURDEN
Coronary plaque burden assessments are useful in identifying the subclinical phase of the disease and providing powerful prediction of adverse cardiovascular events. Although most patients identified on imaging as having coronary atherosclerotic plaque do not subsequently suffer adverse events, the more plaques a subject has, the higher their risk, presumably because this increases the chances of 1 plaque becoming disrupted and causing an event. Imaging plaque burden is, therefore, potentially attractive in terms of both population screening and risk stratification of asymptomatic patients.
Coronary artery calcium (CAC) scoring uses noncontrast electrocardiographic (ECG)-gated CT to provide accurate and simple measurements of the coronary atherosclerotic burden (Figure 1). In particular, CAC quantifies macroscopic calcium within these vessels by using the Agatston score (12). Coronary macrocalcification is highly specific for atherosclerosis but is usually associated with stable plaques at low risk of rupture. This is supported by autopsy and imaging studies demonstrating that stable plaques are associated with advanced macroscopic calcification, whereas ruptured coronary plaques tend to be associated with the very early stages of micro-or spotty calcification (often with extensive calcification elsewhere in the coronary vasculature) (13)(14)(15). Similarly, other CT studies have demonstrated that more dense coronary calcification is associated with a lower risk of cardiovascular events than less dense calcium (16). Nevertheless, CT calcium scoring provides powerful prognostic information because it provides a surrogate for the total plaque burden that will correlate with the number of unstable, adjacent plaques. A CAC score >300 Agatston units (AU) is associated with a 4-fold higher risk of cardiovascular events than a CAC score of zero (17), which is itself associated with an excellent prognosis.
Indeed, "the power of zero" is so good that it can provide a <1% annual mortality rate for up to 15 years in asymptomatic and otherwise low-or intermediaterisk patients (18), justifying the downscaling of preventive treatments (19). High calcium scores are also of use, with several population-based studies demonstrating that addition of calcium scoring to Framingham risk scores (20,21) improves the prediction of coronary events. On that basis, the 2013 American College of Cardiology/American Heart Association Guideline on the Management of High Cholesterol (22) recommended that a CAC score of >300 AU be used as a modifier to justify statin therapy for primary prevention in adults between 40 and 75 years of age without diabetes and with low-density lipoprotein cholesterol 70 to 189 mg/dl. Although the radiation dose associated with CT calcium scoring is small (2 to 3 mSv), it does remain of concern in the context of multiple repeat assessments or wide-    Both CT and CMR offer detailed and comprehensive imaging of ischemic heart disease. Gold-outlined boxes highlight preferred modalities. CT currently holds the advantage for coronary imaging, whereas CMR offers more detailed assessments of the myocardium. Orange areas on the coronary CTA outline areas of low attenuation    Figure 2). In post-MI      CMR perfusion has been tested in several large clinical trials, demonstrating high diagnostic accuracy (69) that was at least equivalent to that of SPECT perfusion imaging in the recent CE-MARC (Cardiovascular magnetic resonance and single-photon emission computed tomography for diagnosis of coronary heart disease) clinical trial (70). This trial also recently demonstrated that CMR perfusion provides prognostic information that is improved compared with that of SPECT (71). A recent large meta-analysis demonstrated a pooled sensitivity of 89% and a specificity of 78% for the detection of obstructive coronary artery disease (72) and that a negative CMR stress perfusion test was associated with an excellent prognosis (73). Although SPECT is much more widely available in the United States and supported by strong prognostic observational data, CMR perfusion has the crucial advantage of being radiation-free and appears to be cost effective    (Figure 4).
The assessment of global and regional function in the left ventricle is performed routinely using echocardiography. This provides important prognostic and diagnostic information in the assessment of patients with ischemic heart disease and is used to guide implantation of automatic implantable cardioverterdefibrillators and cardiac resynchronization therapy.
In patients with poor echocardiographic windows or in cases of diagnostic uncertainty, CMR can be used to assess left ventricular function. Indeed, CMR is widely considered the noninvasive gold standard for these measurements (62). Similar assessments are also available with coronary CTA using images acquired throughout the cardiac cycle, although this again involves radiation exposure (3 to 10 mSv) and is rarely performed in clinical practice.