Coronary Computed Tomography Angiography
Marc Dewey
Lucia J.M. Kroft
CLINICAL APPLICATION OF CORONARY CT ANGIOGRAPHY
Noninvasive coronary angiography using computed tomography (CT) results in high sensitivity for the detection of significant coronary artery stenosis that is far above 90%. Meta-analyses of patients with suspected coronary artery disease have shown a pooled per-patient sensitivity of 97% (with 95% confidence intervals of 96% to 98%) (1) (Table 25.1). As a result also the negative likelihood rate of coronary CT angiography, that is, the odds that a negative result is actually false, is very low (0.03 with confidence intervals of 0.02 to 0.04) (1). This makes coronary CT angiography the most accurate noninvasive imaging test for ruling out coronary artery stenosis, and it is mainly recommended for this purpose in patients with low-to-intermediate pretest likelihood of disease (2,3) (Fig. 25.1). The pretest likelihood for coronary artery disease is determined beforehand by the clinical presentation, risk factors, and diagnostic test results (such as exercise ECG) (4).
However, the moderate specificity of coronary CT angiography in patients with suspected coronary artery disease of 89% (95% confidence intervals of 86% to 92%) (1) makes CT an imperfect test to confirm significant coronary artery stenosis. Nevertheless, the probability that a diseased patient has a positive coronary CT angiography is about nine times higher than a positive result in a patient without coronary disease (i.e., positive likelihood ratio). Therefore, further refinements are needed to increase the specificity, for example, by adding functional testing such as myocardial stress CT perfusion (CTP) to the armamentarium of cardiac CT (5,6 and 7). Besides accurately ruling out coronary artery disease (Fig. 25.1), cardiac CT is also extremely valuable for the comprehensive evaluation of the course of anomalous coronary arteries (Figs. 25.2 and 25.3) and abnormal terminations of coronary arteries, such as in the case of fistulae (Fig. 25.3).
TECHNICAL IMPLEMENTATION
Setting up coronary CT angiography in clinical practice requires two basic conditions.
First, there is the need of appropriate training of technicians to conduct the scan and of physicians to read the images. This sounds trivial but the learning curve for institutions and individuals is rather long with a minimum of about 6 and 12 months, respectively (8,9). Hands-on courses as well as longer-lasting fellowships with individual supervision provide the best available opportunity to prepare for the high demands of cardiac CT (10). Second, the technical features of the CT system used should allow achieving diagnostic image quality in almost every patient. Reassuringly, on a worldwide perspective such technical requirements are achieved in most sites (11).
One of the basic technical requirements for cardiac CT is the possibility to trigger the acquisition based on the ECG of the patient (Fig. 25.4). Prospectively ECG-triggered acquisitions which are currently feasible on 64-row coronary CT angiography in patients with slow and stable heart rates enable significant reductions in radiation exposure when compared with traditional retrospectively gated acquisitions. The slice collimation for coronary CT angiography needs to be ≤0.625 mm, and for improvement of rendering the slice increment should be set to smaller values, such as 0.5 mm or less. With current standards, CT systems should be equipped with at least 64 detector rows and should have a gantry rotation time of below 0.4 seconds (12,13,14 and 15). Nevertheless, in certain patients 64-row CT does not allow reliable delineation of coronary anatomy and ruling out of coronary artery disease. The main causes for non-diagnostic
coronary CT angiography are CT-dependent technical limitations in spatial and temporal resolution. Partial volume effect and beam hardening cause blooming artifacts especially in the presence of severe calcifications that hamper evaluation of the coronary artery lumen. Motion artifacts can be caused by postural motion (e.g., if the patient cannot hold the breath), by high heart rates that result in too short coronary rest-phase for motionless imaging, and by arrhythmia (causing reconstruction artifacts, Fig. 25.5) (16,17). The problems of arrhythmia can be overcome by fast coverage of the entire cardiac volume within a single heartbeat (Fig. 25.6). Arrhythmia rejection performed prospectively by the CT scanner is helpful to improve image quality while reducing effective dose as compared to retrospective gating techniques. Improving the coverage of coronary CT angiography and single heartbeat scanning can be done using either 320-row volume or fast dual-source CT (Fig. 25.7).
coronary CT angiography are CT-dependent technical limitations in spatial and temporal resolution. Partial volume effect and beam hardening cause blooming artifacts especially in the presence of severe calcifications that hamper evaluation of the coronary artery lumen. Motion artifacts can be caused by postural motion (e.g., if the patient cannot hold the breath), by high heart rates that result in too short coronary rest-phase for motionless imaging, and by arrhythmia (causing reconstruction artifacts, Fig. 25.5) (16,17). The problems of arrhythmia can be overcome by fast coverage of the entire cardiac volume within a single heartbeat (Fig. 25.6). Arrhythmia rejection performed prospectively by the CT scanner is helpful to improve image quality while reducing effective dose as compared to retrospective gating techniques. Improving the coverage of coronary CT angiography and single heartbeat scanning can be done using either 320-row volume or fast dual-source CT (Fig. 25.7).
Table 25.1 Diagnostic Accuracy of Coronary CT Angiography on the Per-Patient Level | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 Figure 25.1. Three-dimensional display of an example of coronary CT angiography in which coronary artery disease could be reliably ruled out. |
PERFORMING AND INTERPRETING CORONARY CT ANGIOGRAPHY
Preparation for coronary CT angiography includes administering oral and/or intravenous beta-blocker in patients with heart rates exceeding 60 bpm. About 91% of cardiac CT sites regularly use beta-blockade (11). Alternative medication (e.g., if channel blockers, calcium-antagonists and midazolam) may be used in patients with contraindications for beta-blockers. Slowing down the heart rate is important for avoiding motion artifacts and for achieving diagnostic image quality. In addition, with slow and stable heart rates, radiation exposure can be reduced, as acquisition can be limited to a short prospectively triggered exposure interval during diastole.
Nitroglycerin is used by about 80% of cardiac CT sites worldwide (11) and has the following advantages. First, dilatation of the coronary arteries by nitroglycerin improves the relative spatial resolution of the CT images. Second, nitroglycerin is commonly used as an intracoronary injection during conventional coronary angiography and thus comparison with CT results is improved when nitroglycerin is also used for this test.
Performing cardiac CT is optimally done in a systematically unified way within the institution. Figure 25.8 gives an overview of the data acquisition procedure as typically done in
64-row cardiac CT. Bolus tracking is most often used for initiating the scan after contrast agent injection (Fig. 25.8) (11). Alternatively, test bolus injection can be used to identify the optimal scan point but requires a slightly higher contrast agent volume (i.e., that of the test bolus itself). Bolus tracking is performed during the actual scan and thus has less variability. By using single-beat cardiac CT, the contrast agent volume can be further reduced in comparison to multi-beat 64-row CT. Further reductions in contrast agent volume can be achieved using fast bolus tracking approaches available with single-beat imaging.
The flow and contrast agent amount is adjusted to the individual patient weight to unify image quality.
64-row cardiac CT. Bolus tracking is most often used for initiating the scan after contrast agent injection (Fig. 25.8) (11). Alternatively, test bolus injection can be used to identify the optimal scan point but requires a slightly higher contrast agent volume (i.e., that of the test bolus itself). Bolus tracking is performed during the actual scan and thus has less variability. By using single-beat cardiac CT, the contrast agent volume can be further reduced in comparison to multi-beat 64-row CT. Further reductions in contrast agent volume can be achieved using fast bolus tracking approaches available with single-beat imaging.
The flow and contrast agent amount is adjusted to the individual patient weight to unify image quality.
 Figure 25.2. Malignant-type coronary artery anomaly with the right coronary artery arising from the left sinus of Valsalva (arrow, A,B). The anomalous right coronary artery may be compressed between the aorta and pulmonary trunk on its interarterial way. This is a case with a low interarterial course of the anomalous right coronary artery. Such low courses have been shown to lead less commonly than high courses to major adverse cardiac events. LM, left main coronary artery. |
 Figure 25.3. Left anterior descending coronary artery to pulmonary artery fistula and aberrant left circumflex coronary artery in a 74-year-old female patient. A,B: A small arteriovenous coronary artery fistula (arrow) from the proximal left anterior descending coronary artery to the pulmonary artery (PA). C: The course of an aberrant left circumflex artery (LCX) arising from the right sinus of Valsalva with a retro-aortic course continuing in its normal location in the left atrioventricular groove. Arteriovenous coronary artery fistulas are present in approximately 1/1,000 coronary angiographies. CT can be very useful and superior to invasive coronary angiography in determining the anatomic relationships between the involved structures. Arteriovenous fistulas bypass the myocardial tissue, which may cause symptoms (present in about half of the patients), depending on the size of the fistula and the shunt volume. |
 Figure 25.4. Effect of untriggered (A) versus ECG-triggered imaging (B) in a 52-year-old male patient with chronic atrial fibrillation. CT imaging of the heart was performed before radiofrequency ablation therapy of the pulmonary veins. The CT images are used as roadmap for guiding the procedure. This scan was repeated because the procedure had to be performed a second time. Before prospective triggering techniques became available, we used an untriggered (non-ECG-synchronized or gated) protocol for pulmonary vein imaging (A). With untriggered protocols, the pulmonary veins are already sharply displayed (arrows, A) because they hardly move in time. This is different from the ascending aorta (Ao) and coronary arteries that move substantially throughout the cardiac cycle and are only displayed without motion on ECG-synchronized imaging (B). Now that low-dose prospective ECG-triggering techniques have become available, we use these approaches for sharp imaging of the pulmonary veins and the coronary arteries as well (arrowhead, B), at approximately the same radiation dose. |
 Figure 25.5. Issues with irregular heart rate in a 67-year-old female patient with suspected coronary artery disease who underwent 64-row CT that resulted in helical step-artifacts. The patient had an irregular heart rate during scanning (red circle, A), resulting in step-artifact through at the level of the aortic root (B-D). Some data are missing during image reconstruction because of a too short RR interval, whereas other data are projected twice because of one too large RR interval (i.e., double projected left main coronary artery; arrows, D). The result is a poor-quality scan with nondiagnostic results. |
 Figure 25.6. Successful arrhythmia rejection with 320-row CT (arrow, A) in a 49-year-old female patient with suspected coronary artery disease and irregular heart rate. Issues in patients with irregular heart rates shown in Figure 25.5 can be overcome using single-beat scanning of the entire heart with prospective triggering and arrhythmia rejection that allows for sharp imaging of the coronary arteries (B-D
Get Clinical Tree app for offline access
 Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
|