Dr. Mollet and Dr. de Feyter are cardiologists in the Department of Cardiology, Thorax Center and the Department of Radiology; Dr. Nieman is a cardiologist in the Department of Cardiology, Thorax Center; Dr. Cademartiri and Dr. Pattynama are radiologists and Mr. Raaijmakers is a radiology technician in the Department of Radiology, Erasmus Medical Center, Rotterdam, The Netherlands.

C onventional X-ray coronary angiography is the undisputed gold standard for detection and assessment of coronary artery disease. However, it is an invasive technique with a low risk of procedure-related complications, such as arrhythmia, stroke, coronary artery dissection, and, very rarely, death. In the past decade, several imaging modalities have been developed for noninvasive coronary imaging, such as magnetic resonance imaging (MRI), electron-beam computed tomography (EBCT), and multislice computed tomography (MSCT). Promising results have been reported using MRI and EBCT. ^1-3 However, reliable noninvasive angiography was usually restricted to the proximal and middle part of the main branches of the coronary arteries, and a significant number of these branches were excluded due to insufficient image quality. Additionally, while impressive images and promising results have been published using 4-row MSCT scanners, ^4-7 clinical use of this technique has primarily been restricted to use in patients with slow heart rates (<65 bpm) who are able to hold their breath for up to 40 seconds.

Recently, a new generation of MSCT scanners has been introduced, using an extended number of thinner detector rows, as well as an increased X-ray tube rotation speed. The robustness of the 16-row MSCT scanner may boost this technology to become an alternative to invasive coronary angiography.

Multislice spiral CT coronary angiography procedure

Patient preparation

Heart rate control remains important for high image quality and reliable diagnostic MSCT angiography. In case
of a prescan heart rate >=65 bpm, and in the absence of contraindications, a single oral dose of 50 to 100 mg metoprolol
is administered 1 hour prior to the ex-amination. Additionally, patients are in-structed not to drink coffee or tea prior
to the examination.

Contrast material protocol

Iodinated contrast material (CM) is routinely administered through an antecubital vein. Alternatively, femoral or central venous injection of CM can be applied. After initiation of the contrast injection, and a preset delay of 10 seconds, axial slices at the level of the aortic root are acquired to monitor the arrival of CM. When the CM reaches a predefined threshold (+100 HU) inside a region of interest set inside the ascending aorta, the patient is automatically instructed to maintain an inspiratory breath-hold for approximately 20 seconds. The CM (320 to 400 mgI/mL) can be injected as a monophasic bolus (120 mL at 3.5 to 4.0 mL/sec), a biphasic bolus (120 mL split in 50 mL at 4.0 to 4.5 mL/sec, and 70 mL at 2.5 to 3.5 mL/sec), or using a bolus chaser (100 mL at 4.0 mL/sec, followed by 50 mL of saline).

Scan protocol

The minimal slice width of the 16-row MSCT scanner (Somatom Sensation 16, Siemens Medical Solutions, Forchheim, Germany) is 0.75 mm. For extended coverage per second, a 1.5-mm slice width can be selected. A 16- * 0.75-mm detector collimation is used for this purpose, while 16 * 1.5 mm is used for electrocardiography-synchronized scanning of the thoracic vasculature. Scan parameters of the 4-row and 16-row MSCT scanner are listed in Table 1 (Somatom Sensation 4 and Somatom Sensation 16, Siemens Medical Solutions). In patients with a regular heart rate, X-tube modulation can be applied prospectively. This feature reduces the radiation output of the scanner during the less relevant systolic phase, thereby reducing the total radiation dose by >50% in patients with a low heart rate, leading to an effective radiation dose of 4 to 5 mSv. ⁸ The spiral CT data is acquired in the craniocaudal direction, usually in <20 seconds.

Image reconstruction

When images are reconstructed retrospectively within the diastolic phase, nearly motion-free images of the coronary arteries can be obtained. Therefore, images are routinely reconstructed at different reconstruction intervals to select the best images, although a reconstruction interval starting at 400 or 450 ms before the next R-wave generally provides the best results. With heart rates >70 bpm, a bisegmental reconstruction algorithm can be used to combine data from consecutive heart cycles. This algorithm improves the temporal resolution by decreasing the effective slice reconstruction time down to 105 msec, depending on the heart rate. ⁹ A three-dimensional
(3D) data set is prepared by reconstruction of up to 250 images with a reconstructed slice thickness of 0.75 to 1.0 mm and a reconstruction increment of 0.6 mm. This dataset can be further processed with 3D-volume rendering software.

Image evaluation

For assessment of coronary artery disease, images are evaluated using (curved) multiplanar reconstructions (MPRs) and thin-slab maximum intensity projections (MIPs). Maximum intensity projections provide fast and accurate information for an overview of coronary artery patency. However, in the case of pronounced calcification, evaluation of stent patency, or plaque imaging, MPRs are preferred over MIPs. Volume-rendering techniques (VRT) provide additional information, which can be useful to evaluate anatomy and give an overview to the referring clinician. However, these images are not yet optimal for detection and semi-quantification of coronary artery stenoses. Further imaging tools (eg, automatic vessel tracking) are under development.

The imaging performance
of 16-row MSCT scanners

Clinical impact of technical improvements

The technical improvements implemented within 16-row MSCT scanners have several practical consequences. Faster rotation time reduces motion artifacts and increases scan interpretability. The increased number of slices results in reduction of scan time. A shorter scan time has several practical advantages, such as reduction of CM volume, a manageable breath-hold, reduction of breath-hold­related heart rate acceleration, and reduced voluntary patient motion. The use of thinner slices increases through-plane spatial resolution, which results in near-isotropic submillimeter resolution and allows better assessment of smaller branches. Another technical improvement that recently has become available is the use of bolus tracking (eg, CARE bolus, Siemens Medical Solutions). Bolus tracking eliminates the need for a test bolus, which reduces the total amount of CM by 20%. Combined with automatic patient instruction, bolus tracking improves workflow.

MSCT angiography versus conventional angiography

Several studies have compared conventional X-ray coronary angiography with MSCT coronary angiography using 4-row MSCT scanners. ^4-7 Sensitivity and specificity for detection of coronary stenoses >=50% ranged from 75% to 88%, and 84% to 97%, respectively (Table 2). However, a significant number of vessels were excluded (up to 32%) because of reduced image quality, most frequently owing to motion artifacts and extensive vessel calcification. First results obtained with a 16-row MSCT scanner indicate very promising results for detection of coronary stenoses >=50% with an overall sensitivity and specificity of 95% and 86%, respectively, and, most importantly, without excluding vessels from the analysis. ¹⁰ Results of other validation studies will become available in the near future.

Clinical relevance of
MSCT angiography

Until now, noninvasive coronary imaging for detection of coronary stensoses had not reached an important role in clinical practice. Reliable coronary angiography using MRI, EBCT, or 4-row MSCT had been primarily restricted to proximal and mid-segments, while a significant number of coronary vessels were excluded. ^1-3 With the introduction of 16-row MSCT scanners, noninvasive coronary imaging has reached a new level of performance; 16-row MSCT coronary angiography offers an acceptable alternative in a restricted patient population to conventional angiography with respect to the detection of stenoses (Figure 1). This population involves patients with regular sinus rhythm, a heart rate of <65 bpm, and age <70 years to limit the image-degrading effect of coronary calcium. Additionally, patients should be willing to cooperate and be capable of maintaining a breath-hold of 20 seconds, and no contraindications should be present concerning roentgen exposure, iodinated intravenous CM, or, when
necessary, the use of beta-blockers.

No studies have been published exploring how this new technology could fit into clinical practice. However, clinical applications could be: 1) exclusion of sten-otic lesions in high-risk patients; 2) work-up of patients with chest pain; 3) follow-up after bypass surgery or angioplasty; and 4) risk stratification of patients with or suspected to have coronary artery disease. The latter application involves not only quantification of coronary calcium (calcium scoring), but also plaque characterization. Preliminary data obtained with the use of 4-slice MSCT scanners found a strong correlation between MSCT plaque density measurements (in HU) and the qualitative intracoronary ultrasound classification of soft, intermediate, and calcified plaques. ¹¹ Noninvasive plaque character-
ization may become clinically important and may aid in the detection of vulnerable plaques. MSCT angiography may, in the near future, mature into an important diagnostic tool and serve as a reliable alternative to conventional invasive diagnostic coronary angiography.

Conclusion

Sixteen-row MSCT scanners allow for fast and robust imaging of the coronary arteries. First results show a good diagnostic accuracy with respect to detection of coronary stenoses in a defined patient population. Further expected advances in this technology will soon expand the role of noninvasive coronary angiography in cardiovascular medicine.