Since the introduction of multislice computed tomography (MSCT) in 1998, non-invasive cardiac imaging has developed rapidly. In the recent past, two different objectives can be observed in the technological development of cardiac MSCT. On the one hand, the number of detector rows was increased up to 320 in single-source scanners in order to ensure simultaneous coverage of the entire heart with a high spatial resolution. On the other hand, a higher temporal resolution was attained, which was achieved by introducing a dual-source scanner with a maximum of 2x64 rows (2x128 slices). Currently, the maximum temporal resolution is 75ms on a dual-source system and the maximum spatial resolution is 0.5mm on a single-source system. The high spatial resolution of MSCT results in excellent morphological depiction of coronary arteries, bypasses, myocardium and even heart valves, which can be simultaneously reconstructed from a 3D data set. While motion artefacts at heart rates up to 65bpm have generally been reduced, higher heart rates, arrhythmias, severe calcification, coronary stents and an unfavourable signal-to-noise ratio in obese patients remain challenges.
Computed Tomography Coronary Angiography
Coronary heart disease (CHD) is one of the most common causes of death in industrialised western nations.1 Invasive coronary angiography (ICA) as the valid standard of reference allows diagnosis and, if necessary, treatment of CHD in the same session. More than 50% of ICAs are purely diagnostic, meaning that there is no coronary intervention.2 Since complications of this invasive procedure are very rare but potentially serious, a non-invasive diagnostic procedure such as CT coronary angiography is desirable. In other vascular regions, non-invasive diagnostic procedures such as magnetic resonance imaging (MRI) and CT angiography have mainly replaced invasive angiography. However, the requirements in terms of spatial and temporal resolution with CT coronary angiography are much higher due to the movement of the heart, the size of the coronary arteries, the veins running parallel and the surrounding tissue.
However, in recent years the technological development of cardiac MSCT has increased the importance of this procedure for the assessment of CHD in comparison with cardiac catheterisation. Recent studies of CT coronary angiography performed using 64-slice scanners describe a sensitivity of up to 98% and a specificity of up to 91% over all coronary segments.3–8 Due to a high negative predictive value of 98–100%, CT coronary angiography is particularly suitable for non-invasive exclusion of CHD4–8 (see Figure 1). Limitations with regard to the diagnostic reliability depend on the patient group, the equipment used and the heart rates and bodyweight of the examined patients. However, the only published prospective multicentre study of CT coronary angiography using 64-slice scanners that included symptomatic patients at high risk of obstructive CHD described lower negative and positive predictive values.9 Consequently, CT coronary angiography cannot replace ICA at this stage.
CT coronary angiography visualises the coronary lumen and the vessel wall at the same time and thus also allows plaque characterisation. With the introduction of the Agatston score in 1990, reliable non-invasive quantification of coronary plaque burden became available for the first time; however, this was based solely on the detection of coronary calcification. In today’s contrast-enhanced CT coronary angiography, intermediate and lipid plaques as well as calcifications can be differentiated reliably in close correlation to intravascular ultrasound (IVUS).10 Using this plaque characterisation, in the future it may be possible to differentiate consolidated stable from vulnerable plaques for individual risk stratification for the occurrence of a possible acute coronary syndrome.11,12
The depiction of arterial and venous coronary artery bypasses as a part of a cardiac CT is feasible and reliable (see Figure 2). This technique is used regularly in many institutions. The sensitivity and specificity for the assessment of coronary bypasses, as well as the negative predictive value, are similar to those of CT coronary angiography.13 The limitations of this procedure primarily concern the assessment of the distal bypass anastomosis and the distal native vessel. Coronary stents can be visualised depending on their diameter and material, but often the stent lumen eludes a reliable assessment (see Figure 2C). As a rule, coronary stents with a diameter less than 3mm cannot be assessed using CT coronary angiography.14
Evaluation of Myocardium, Heart Valves and Pulmonary Veins
The morphology of the left ventricle and its functional parameters can be determined reliably in cardiac CT using retrospective gating. Using MRI as the standard of reference, cardiac CT showed similar good results with regard to the assessment of left ventricular function parameters but at a lower temporal resolution than MRI15–17 (see Figure 3). CT has a special role for the combined assessment of left ventricular morphology or function and the coronary arteries, for example in patients with left ventricular aneurysms. In addition, MSCT allows the assessment of myocardial perfusion dynamically or as a first-pass perfusion. The dynamic perfusion measurement principally enables an absolute quantification of myocardial blood flow, but is reasonable only when wide-area detectors are used. The distribution of the contrast agent in first-pass perfusion correlates strongly with single photon emission CT (SPECT) myocardial perfusion imaging. Sensitivity and specificity for the detection of abnormal perfusions are 75 and 72%, respectively.18 Furthermore, it is technically possible to use a ‘delayed enhancement’ of the myocardium in CT, similar to the MRI for scar detection. For example, a typical subepicardial delayed enhancement can be depicted in patients with acute myocarditis.19 However, CT evaluation of the myocardium has not yet been fully validated,20 and MRI is preferred in clinical practice for reasons of radiation protection.
The spiral acquisition of a cardiac CT contains the necessary data to evaluate the heart valves, if the study protocol provides sufficient contrast in the associated cardiac cavities. This may reliably include valve calcification and planimetrical measurements of the aortic valve area in systole or the regurgitant orifice area in diastole.21 In combination with CT coronary angiography, MSCT may be an alternative to pre-operative ICA before aortic valve replacement in the future. The mitral and tricuspid valves are much more susceptible to motion artefacts, but in principle can also be reconstructed to provide good-quality results. Direct flow measurement of the valves as performed in MRI is not possible with CT.
With the increasing use of radiofrequency ablation for the treatment of patients with chronic atrial fibrillation in recent years, pre-interventional imaging of the anatomy in CT has proved to be very useful. The pulmonary veins can be easily segmented and are less susceptible to motion artefacts. The integration of the 3D reconstruction of the pulmonary veins as an overlay in fluoroscopy during the ablation procedure may shorten the fluoroscopy time and the number of required ablation points.22
Contrast Agents and Injection Protocols
Almost all studies of MSCT of the heart require the administration of a contrast agent. Usually, a non-ionic radiopaque contrast agent such as iopromide with iodine concentrations of 300–400mg/ml is applied. To obtain a diagnostically sufficient contrast ratio of coronary arteries to the surrounding tissue, an application with a high flow rate of 4–6ml/second is required, preferably via a peripheral, sufficiently strong antecubital arm vein, followed by a saline bolus. The scan delay should be defined by bolus tracking. An alternative to the two-phase injection protocols is a three-phase injection protocol, where an intermediate injection phase of 50% contrast and saline can be used to ensure sufficient contrast of the right heart cavities in cases concerning myocardial morphology and function. The quantity of contrast applied depends strongly on the scan time and is usually about 60–90ml. The bolus application of contrast is usually easily tolerated – as long as contraindications for iodine contrast agents are considered (especially allergies, renal restrictions and hyperthyroidism) – but in patients with very low ejection fraction and cardiac output a rapidly applied volume can lead to complications.
Radiation Dose Considerations
The radiation dose in cardiac CT using retrospective gating is relatively high in comparison with other non-invasive X-ray procedures. Previous studies estimated the effective dose using the CT dose index (CTDI) and dose length product (DLP) at a range of 6.4–14.7mSv.12,23 In purely diagnostic cardiac catheterisation, the average radiation exposure is estimated to be 3mSv.24 Dose measurements using anthropomorphic phantoms to 64-row scanners quantify the effective dose in the range of 8–31.8mSv.25,26 To reduce the radiation exposure, modern CT scanners can modulate the tube current during image acquisition and apply the maximum selected tube current during one cardiac phase only, for example the end-diastolic phase. This electrocardiogram (ECG)- modulated tube current leads, depending on the chosen parameters, to a dose reduction of up to 50%. Recently, a new approach to dose reduction, the so-called ‘step-and-shoot’ technique, was introduced, using an incremental image acquisition with prospective ECG triggering during a selected cardiac phase. Typically, end-diastolic images are used for the assessment of coronary arteries. On a 64-row CT, four acquisition steps usually suffice to cover the entire heart. Using a 320-row CT, the acquisition can even be performed in one step. For 64-slice scanners, a dose reduction of 77% is described according to the calculated DLP.27 Our own measurements on an anthropomorphic phantom confirmed these theoretical calculations, showing a reduction of organ doses by 65–87%.28
Appropriate Use of Cardiac Computed Tomography
In young patients, and especially in females, cardiac CT should be used extremely reluctantly and examinations should be performed using the maximum possible dose reduction options. CT coronary angiography is currently not appropriate as a screening examination in asymptomatic patients.29 The assessment of left ventricular function and morphology alone is not an indication for cardiac CT, as these parameters can be assessed by echocardiography and cardiac MRI without radiation exposure and with higher temporal resolution. Taking into account the interdisciplinary appropriateness criteria under the auspices of the American College of Cardiology Foundation (ACCF) and the American College of Radiology (ACR) and the Scientific Statement of the American Heart Association (AHA) for cardiac CT,29,30 the following indications are considered as accepted for current-generation 64-slice CT.
Exclusion of Coronary Heart Disease
Patients with an intermediate risk of CHD according to their symptoms, their age and their sex who have undergone one or more inconclusive stress tests are suitable candidates for cardiac CT. This includes patients with atypical angina pectoris and ambiguous results of previous stress tests (ACCF/ACR: A7–A8; AHA: Class IIa, Level of Evidence B). Patients at high risk of CHD, for example patients with atypical angina pectoris or a positive stress examination, do not benefit from cardiac CT because the need for coronary intervention is likely.
Acute Chest Pain
The first studies in this area have shown that cardiac or thoracic CT in patients with acute chest pain and an intermediate likelihood of CHD is of high value, and that CT could play a key role in the early management of acute chest pain patients.31–33 An indication in these patients may be given when ECG and cardiac enzymes are inconspicuous (ACCF/ACR: A7).
Assessment of Coronary Artery Anomalies, Coronary Artery Aneurysms/Fistulas and Complex Congenital Heart Defects
In these cases, cardiac CT and MRI are superior to ICA because the selective depiction of aberrant coronary arteries in ICA can be rather difficult. Because of radiation exposure considerations, MRI should be the first modality to be considered, although CT provides more reliable visualisation of the entire course of the vessel, including the distal sections (ACCF/ACR: A7–A9; AHA: Class IIa, Level of Evidence B).
Assessment of Coronary Bypasses
Patients undergoing re-surgery for revascularisation to assess intra- and extracardiac findings, including the A. mammariae, have an indication for cardiac CT (ACCF/ACR: A8; AHA: Class IIb, Level of Evidence C – state 2006,34 currently no evidence level assigned).
Radiofrequency Ablation and Pacemaker Implantation
Cardiac CT has an indication for the evaluation of pulmonary vein anatomy before invasive radiofrequency ablation in patients with chronic atrial fibrillation and evaluation of coronary veins before biventricular pacemaker implantation (ACCF/ACR: A8).
CT of the heart has developed into a robust method for the morphological and functional imaging of the heart. Currently, CT cannot replace diagnostic ICA, but it provides a useful non-invasive diagnostic alternative to ICA for appropriate indications.