Since its introduction into medical imaging in the 1970s,1 computed tomography (CT) has significantly contributed to novel diagnostic approaches in a wide variety of clinical conditions across multiple medical specialities. However, increased use has also been associated with a significant increase in radiation exposure, with uncertain long-term implications.2,3 Cardiovascular multidetector computed tomography (MDCT) has been in the spotlight of the cardiovascular community since imaging of the coronary arteries became possible using coronary CT angiography (CTA)4–6 (see Figure 1).
Single- and multicentre studies have compared CTA with conventional coronary angiography and consistently demonstrated a high negative predictive value (i.e. ability to rule out significant disease), but a lower positive predictive value in identifying high-risk lesions.7,8 The limitations in assessing and quantifying severe, advanced coronary lesions are related to the lower spatial and temporal resolution of CTA and lesion calcification. Plaque calcification, which is an integral part of advanced atherosclerotic lesions, frequently precludes accurate visualisation of the artery lumen because dense calcification of the plaque leads to overestimation of luminal stenosis (calcium ‘blooming’ artefact) (see Figure 2).
Based on these results, a consensus about the appropriate diagnostic use of CTA is evolving.9 In contrast to conventional coronary angiography, which is typically reserved for symptomatic high-risk populations (high pre-test likelihood of significant lesions) with suspected high-grade stenosis, current guidelines describe CTA as being appropriate in intermediate-risk populations (intermediate pre-test likelihood of significant stenosis). In these patient populations, which are alternatively examined with stress testing, the diagnostic goal is two-fold: identification or exclusion of luminal stenosis and assessment of long-term cardiovascular risk. An established example for this diagnostic approach is nuclear stress testing, which has well-documented prognostic value in addition to the role of identifying ischaemia.10,11 The prognostic value of CT has been defined in non-contrast CT (calcium scoring) studies that show a prognostic role of calcium burden independent of conventional risk factors in intermediate-risk populations.12,13
More recently, CTA studies have demonstrated the potential of CTA to visualise calcified and non-calcified atherosclerotic plaque of the vessel wall.14–16 By simultaneously assessing luminal stenosis and plaque burden, CTA allows the description of atherosclerotic disease patterns, including absence of disease, non-obstructive disease and suspected obstructive disease. The pattern of absence of disease (no atherosclerotic plaque and no luminal stenosis) appears to be associated with a very low risk of future events.17 No further tests are necessary, and risk factor modification should follow established preventative guidelines. The pattern of non-obstructive disease (calcified or non-calcified plaque in the vessel wall with estimated <50% stenosis) is likely associated with intermediate-risk patients.17 Depending on clinical suspicion, additional functional stress tests can be justified, and this pattern should trigger a review of potentially more aggressive risk factor management.
In patients with a pattern of suspected obstructive disease (suspected >50% luminal stenosis), calcified lesions frequently preclude precise quantification of luminal stenosis, and further evaluation of haemodynamic significance is necessary. Considering the well-known discrepancy of anatomical and functional lesion significance,18 the exclusion of haemodynamically significant stenosis should be based on initial correlation with functional stress test results in most patients. Due to the overestimation of calcified lesions with CTA, direct referral to cardiac catheterisation is associated with a high number of unnecessary procedures in patients with calcified but haemodynamically insignificant lesions. If clinical symptoms and stress testing suggest a high likelihood of significant stenosis, cardiac catheterisation is justified. Only in a few, highly selected situations (in particular if proximal disease is identified) is cardiac catheterisation without stress testing appropriate.
The identification of significant plaque burden, regardless of the presence or absence of associated haemodynamic stenosis, requires a more aggressive approach to risk factor modification. The clinical significance of these disease patterns and appropriate clinical management approach needs to be further evaluated and evidence-based data are needed.19 In particular, there is a need for epidemiological data correlating results from CTA with stress testing20 and, eventually, clinical outcome.21 CTA is not recommended for low-risk populations (screening) because of its associated radiation exposure, contrast administration and risk of false-positive test results.
CTA is also not recommended in high-risk patients, in whom conventional angiography remains the test of choice and CTA would only confirm the subsequent need for catheterisation. Other indications are evolving. Examples are the potential use of CTA in patients presenting with chest pain in the emergency department.22
These consensus recommendations will change with increasing experience, but are also inter-related with changes in scanner technology. The development of CT has been characterised by fast technical developments. Current multidetector scanners allow subsecond data acquisition during continuous rotation of the gantry and continuous movement of the patient table. In the resulting spiral acquisition, data are acquired throughout the entire cardiac cycle during simultaneous recording of the electricardiogram (ECG) signal. Current standard 64-detector scanners cover a few centimetres per rotation and acquire the entire coronary tree with three to five gantry rotations. Subsequently, data from specific periods of the cardiac cycle (most commonly late diastole, when cardiac motion is the lowest) are reconstructed by retrospective referencing to the ECG signal (spiral or helical scanning with retrospective ECG-gating).23
Although modern scanners reduce the tube current (and therefore radiation exposure) outside the selected phase (‘dose modulation’), the continuous X-ray exposure during the entire cardiac cycle results in an increased patient radiation dose.24 A significant increase in the number of detectors will allow the imaging of the entire heart in one rotation, obviating the need to move the patient table. With such a scanner, data acquisition can be performed by selectively turning the X-ray tube on during the selected phase only, triggered by the ECG signal. This is basically ‘prospective triggering’.25 The lower radiation dose is an important advantage.
The combination of fast gantry rotation time and a large number of detectors appears to make sequential imaging with prospective triggering feasible for cardiac scanning with these systems. A recent study describes the initial experience with a 320-detector-row CT system.26 The system has a craniocaudal coverage of 16cm in a single gantry rotation, which allows coronary imaging in a single heartbeat in most patients. This is in contrast to current state-of-the-art 64-slice systems, which acquire the data from several heartbeats, introducing potential artefacts at the transitions zone between gantry rotations. Coupled with prospective image acquisition,25 the radiation exposure appears favourable with current CT systems.
Other recent technological developments include faster gantry rotation times, dual-source technology,27 and more efficient detectors, and these improvements are associated with a better image quality. Despite this rapid technical evolution, imaging of the small, rapidly moving coronary arteries with non-invasive modalities remains challenging because of the limited spatial and temporal resolution and impaired visibility of densely calcified segments. This is reflected in recent discussions about appropriate clinical use and national coverage decisions, and is evidence of the evolving role of coronary CT angiography.28,29
While much interest is focused on coronary imaging, there is a much wider field of clinical applications of cardiovascular MDCT, many of which are already well accepted.9 Examples are imaging of the aorta and pulmonary artery.30,31 A particularly innovative area is the planning of endovascular and surgical procedures based on 3D and 4D reconstructions of CT data (see Figure 3). A distinct advantage of coronary computed tomographic angiography (CTA) is the routine acquisition of high-resolution 3D/4D data sets, which allows visualisation of vascular anatomy, including the lumen, vessel wall and their relationships with the surrounding structures. Modern computer-based analysis software allows unlimited oblique reconstruction for precise measurement in the axial plane and along the centre line of vascular structures. This information allows the optimisation of fluoroscopic view selection or surgical access plane, guidance of device selection and device customisation. While data for coronary intervention are still limited,32–34 the potential of image guidance has already been demonstrated by the experience with aortic endovascular stent procedures. In experienced centres, pre-procedural planning with CT is a critical part of clinical routine,35,36 and in particular is used for the design of custom-made stents37 that accommodate vessel tortuosity and side branches, e.g. in the aorta arch and proximal abdominal aorta. More recently, this approach has been described in the context of novel surgical and interventional procedures, including hybrid surgical/ endovascular procedures and robotic surgery.38,39 Other clinical applications are stenting of the pulmonary veins after atrial fibrillation treatment40 and, most recently, percutaneous aortic valve replacement41,42 (see Figure 3).
However, it is important to emphasise that the demonstration of technical feasibility does not automatically imply clinical utility. It is critical to match the rapid technological developments with a rigorous scientific evaluation to demonstrate the impact on clinical decision-making and, eventually, patient outcome. While such evidence-based data are still sparse, recent papers have begun to evaluate these questions (e.g. for outcome after bypass surgery).21,43 Based on accumulating clinical data, appropriate clinical use of cardiovascular CT will be defined in comparison with other imaging modalities, including coronary angiography, echocardiography, magnetic resonance imaging and nuclear imaging. Imaging research will identify novel applications supporting innovative approaches to cardiovascular disease.