Cardiac computed tomography (CT) has become a useful non-invasive technique for the study of cardiac pathology in daily clinical practice due to its high negative predictive value for detecting coronary artery disease (CAD). Its established main role in clinical practice is ruling out CAD in symptomatic patients with intermediate risk of CAD, who remain a diagnostic problem after other non-invasive investigations (stress test, myocardial perfusion single-photon-emission CT [SPECT]). The British Cardiovascular Society Working Group has estimated a 40-fold increase in the number of cardiac CT tests by 2020, not accounting for the impact of technological advances.1
More than 20 years of collaboration between the departments of cardiology and nuclear medicine at our institution have resulted in several clinical studies2 and textbooks3 in cardiac SPECT imaging. This successfully shared experience paved the way for the integration of new imaging techniques such as magnetic resonance imaging (MRI) and CT in cardiac imaging.4
Cardiac CT, including coronary arteriograms, became technically feasible around 2003, after the introduction of multidetector CT (MDCT) scanners with retrospective electrocardiographic (ECG) gating techniques. Initially, radiologists were actively learning how to use a new technology in a different field – the beating heart and its coronary arteries – previously seen as a large resting grey spot surrounded by a pericardium in the middle of thoracic CT images. Specific technical training for cardiac CT requires many hours. Despite basically being a CT scanner, the acquisition and sophisticated volume-rendering capabilities of ECG-gating applications in comparison with conventional CT equipment is similar to a new multimedia computer system compared with a classic television set. Currently, radiologists and cardiologists are evaluating ways of using cardiac CT more efficiently, obtaining the best from the technique in more appropriate indications and in a safer manner. It should be borne in mind that in most radiology and cardiology departments, time is at a premium. Multiplanar reformations and volume rendering are fundamental but time-consuming techniques for exploiting acquired data. Specialised personnel training and newer and better software applications are still required for better optimisation of post-procedure routines.
Selecting Patients for Cardiac Computed Tomography
Cardiac CT is not to be performed in every patient suspected of cardiac pathology due to technical limitations with standard 64 MDCT equipment, inappropriate indications with no proven benefit and, last but not least, exposure to ionising radiation.
The spatial and temporal resolution of standard 64 MDCT equipment frequently results in tests of suboptimal quality in:
- obese patients with lower signal–noise ratio;
- elderly patients with excessive coronary calcifications that preclude luminal evaluation; and
- patients with fast heart rates or arrhythmia, as retrospective gating relies on several identical cardiac beats with long diastolic periods.5
Non-conclusive tests will probably result in additional imaging or functional tests.
As a medical discipline, diagnostic imaging is overloaded by inappropriate indications, which drain a huge amount of limited human and material resources, thereby obstructing the access of appropriate indications for the test.6
Adequate indications for performing cardiac CT have been reviewed in a multidisciplinary scientific statement7 and a document of appropriateness criteria from an expert panel.8 These documents are based on known evidence beneficial for clinical decision-making and should be routinely used for patient selection.
The main and clearly appropriate indications for coronary imaging with cardiac CT7,8 are to rule out CAD in symptomatic patients with low to intermediate risk, and to evaluate coronary artery anomalies and coronary artery bypasses.
As cardiac CT is a new and fast-evolving technique that is in competition with other established imaging tests for the diagnosis of CAD, there are some uncertain indications that are still a matter of debate and controversy.7,8 Some of the indications for coronary angiography using cardiac CT have not been included among appropriate indications established by experts, but should be discussed in future years.9 One new potential indication is non-invasive coronary angiography of candidates for valvular replacement, where preliminary results from ongoing prospective studies seem quite favourable to CT. This indication could be extended to general use prior to cardiovascular surgery in patients with a low- to intermediate-risk profile.
Other indications where risk of invasive coronary angiography for planning a surgical procedure outweighs the benefit are acute aortic dissection, large vegetations in front of coronary ostia or occlusive prosthetic thrombosis leading to an unstable haemodynamic condition.10 As can be easily inferred, many of these indications, for which non-invasive angiography is much more suitable, will be requested after the prior employment of another diagnostic test such as echocardiogram or MRI, which would allow the detection of special circumstances that contraindicate conventional coronary angiography.
Promising results in small patient groups have been reported for coronary stent imaging, as well as fusion SPECT–CT imaging. An example with both modalities is shown in Figure 1.
There is no doubt that the employment of cardiac CT angiography in other special populations, such as children, will require further evaluation; for example, specific conditions such as proximal coronary aneurisms in Kawasaki disease, currently diagnosed and followed up by catheterisation, will probably shift to a non-invasive approach wherever experience is available.
Cardiac CT also allows adequate visualisation of the morphological features and function of a wide range of valve diseases and congenital conditions, which may prove valuable in pre-surgical planning. The main objection against CT is that most of the functional (gradients, shunt amount and direction) and morphological information in these conditions can be achieved by means of ECG or MRI, without the necessity to irradiate patients even with small amounts. In spite of irradiation, the superiority of cardiac CT for adequate evaluation in some extra-cardiac connections (natural, e.g. anomalous pulmonary venous drainage, or artificial, e.g. Glenn or Blalock connection) will lead to CT indications, especially when MRI is contraindicated (pacemakers) or severely affected by artefacts (prostheses).
Currently, cardiac CT tests for inappropriate indications are more likely to require additional imaging or functional tests, resulting in inadequate patient management with test layering and a decreased cost-effectiveness ratio for all the tests involved.4,6
As expected, cardiologists have proved more likely to request appropriate examinations than non-cardiologists.11 At our institution, the radiology department only accepts cardiac CT tests ordered or approved by a cardiologist or cardiac surgeon, in order to keep high standards of appropriateness.
Not only are inappropriate tests are an economic liability, but also cardiac CT with 64 MDCT technology is associated with a relatively high radiation exposure.12 The lifetime attributable risk of cancer incidence associated with a cardiac CT is highly variable, but is believed to be considerably greater for women and younger patients.13 Awareness of the frequency and the radiation dose of radiological and nuclear medicine procedures should be an integral consideration in prescribing examinations.14 Even though the latest MDCT technology will significantly reduce dose exposure for cardiac CT, the ordering physician must always carefully balance the risks and benefits.
Performing Cardiac Computed Tomography
CT equipment is frequently located in and handled by radiology departments with their teams of radiologists, radiology nurses and technicians. Only a few institutions specialised in cardiac care have access to CT scanners exclusively dedicated to cardiac imaging.4 In Spain, most major tertiary care institutions own at least one 64 MDCT in the radiology department partially dedicated to non-invasive angiography in neurology, cardiovascular and abdominal imaging. This CT scanner is also frequently used for emergency CT imaging and standard CT tests, which interrupt outpatient cardiac CT planning and patient preparation, making it difficult for the cardiologist to attend the session. It was estimated in 43 university hospitals that only one in four has a CT-trained cardiologist present during the study acquisition. This is in marked contrast to cardiac MRI, where the virtual absence of emergency tests allows several patients to be booked in a cardiac MR session, thus enabling the presence of a cardiologist to be scheduled. Ideally, cardiac patients should be examined with the latest technology scanner allotted for specific cardiac CT sessions of in- and outpatients as well as emergency cardiac CT imaging.4
Cardiac CT performed with 64-detector technology has suboptimal temporal resolution for cardiac imaging (from 165 to 200ms depending on vendor and generation). Long diastolic phases with motionless coronary arteries are thus required in order to obtain a quality diagnostic cardiac CT scan, with ideal heart rates <65 beats per minute.
Thus, during the patient’s most recent examination, blood pressure and an ECG should be obtained, and heart rate during apnoea should be monitored. If the expected heart rate is >65, or between 60 and 65 but irregular and no contraindications are detected, beta-blockers are given, either orally or intravenously.15 Additionally, sublingual nitroglycerine is given to the patient on the scanner table to achieve coronary dilatation.
On the other hand, the relatively high exposure to ionising radiation during the procedure (effective doses around 10mSv with optimal dose-saving technique using 64 MDCT)16 has been the main drawback compared with conventional diagnostic invasive coronariography (±4mSv).
The latest CT technology (256 detector CT and dual-source CT) has increased the temporal resolution to 70–110ms, with wider detector arrays, resulting in much shorter scanning times with constant image quality in faster and irregular heart rates. Nevertheless, most dose-saving measures, such as prospective gating or fast helical one-beat scanning, require slow heart rates, so beta-blockers are still necessary to achieve doses similar or inferior to conventional coronariography.17
While the administration of beta-blockers and nitroglycerine is frequently performed by the radiologist,15 a cardiologist should always be consulted in those patients prone to instability or with complex pathologies such as cardiomyopathies and congenital or valvular heart disease. In institutions with a high proportion of complex studies, such as in many university hospitals, an on-site cardiologist trained in the technique seems quite convenient, not only for examining patients and administering drugs prior to the test but also for planning and obtaining the best results from the study.
Extensive experience with technical factors allows a radiologist to tailor the cardiac CT examination to each patient and specific scanner in order to obtain the best quality available with the least radiation exposure possible.18 The respective assets of radiologists and cardiologists is shown in Table 1.
Interpreting Cardiac Computed Tomography
Radiologists have extensive morphological imaging skills based on cross-sectional methods that have been applied to non-invasive MDCT angiography of large and small vessels, from the aorta to the polygon of Willis.19 They are experienced at interpreting complex anatomical arterial courses, analysing lumen size and vessel wall composition and manipulating imaging software to gain the maximum information form the source images.20
On the other hand, cardiologists have optimal knowledge of cardiac physiology and pathology, and are usually experienced in interpreting dynamic and functional cardiac images, which are obtained ‘for free’ with retrospective gating techniques. A cardiologist‘s opinion may enhance the diagnostic power of the technique in some cases. It has to be noted that the low-dose prospective gating technique excludes functional analysis. In addition, cardiologists experienced in coronary catheterisation often provide the invasive quantitative gold standard (angiographic or by means of other methods such as fractional flow reserve) for coronary lesions that are employed to evaluate the diagnostic performance of CT coronary angiography in a specific centre.
The interpretation of the test in a collaborative approach with experienced cardiologists allows the addition of clinical commentaries to plain imaging findings. Referring physicians unacquainted with the technique may find this extremely helpful for patient management in borderline or inconclusive findings, such as cases of moderate and/or tandem stenoses.
A high prevalence of unexpected extra-cardiac findings has been reported in the literature (≤18% in some reviews), ranging from pulmonary nodules to pathology. A thoracic radiologist should always interpret the whole test even if a cardiologist is reading the coronary arteries.21Figures 2 and 3 show examples of unexpected cardiac and extra-cardiac lesions found in cardiac CT.
Achieving this ideal state of collaboration requires a fluent relation between both departments, with mutual formation and interdisciplinary decision-making sessions and educational case reviews.
Reporting Cardiac Computed Tomography
As in other cardiac imaging techniques, the report of the cardiac CT represents the essence of a procedure that is performed to answer clinical queries, stated in a formal request, and reflects in an important manner the quality of the work. Bad reporting may destroy the value of magnificent imaging studies. The report should offer the results of a systematic approach using a standard acquisition and interpretation strategy addressing the main questions that motivated the study.22 Cardiologists with expertise in imaging techniques are aware of the importance of the organisation of quantitative information and often apply measurement references from other techniques, mainly coronary angiography and echocardiography, that have been widely validated for cavity dimensions, coronary stenoses, etc. There are different options for organising the results, but a logical structure should always be recommended in order to avoid confusion.23 Radiologists and cardiologists who have interpreted a study together in general tend to produce more complete reports and provide information that can be more easily understood by other physicians (not only cardiologists or radiologists). The Society of Cardiovascular Computed Tomography has recently published guidelines for the interpretation and reporting of coronary CT angiography to promote standardisation of the content of cardiac CT study reports.24
Training for Cardiac Computed Tomography
A recent report of the European Society of Cardiology writing group concluded that conventional training in either radiology or cardiology will not provide sufficient background to perform and evaluate cardiac CT imaging per se.25 Although some speciality fellowship programmes in cardiac imaging and cardiac CT are available in selected institutions, most cardiology and radiology training programmes do not incorporate mandatory exposure to cardiac CT at a level that would suffice to provide competent diagnostic services in cardiac CT.26 The American College of Radiology stated in 2005 that the expertise of a radiologist who supervises and interprets cardiac CT examinations should include supervision and interpretation of 75 cardiac CT cases within 36 months, excluding cases performed exclusively for calcium scoring.27 In 2008, a joint statement by the American College of Cardiology and American Heart Association stated that level 3 training – qualifying an individual to direct an independent cardiac CT programme – requires six months of training, 100 cases performed and 300 cases interpreted, plus 40 hours of continuing medical education.28
However, more recently, European analysis of one-year fellowships showed that acquiring expertise in coronary CT angiography was slow and could take more than one year (about 600 tests) to achieve sensitivities of 66–75%, specificities of 87–92% and diagnostic odds ratios (DORs) of 14.7–25.8.29 It is important to note that it has been demonstrated in other imaging techniques that training and accreditation of professionals involved in the technique are important factors to attenuate the inherent variability in measurements and to increase reproducibility.
Specific training requirements for cardiac imaging are being developed by the European Society of Cardiac Radiology and are expected for 2010. Meanwhile, the society is offering fellowships and advanced courses.
A senseless turf war for the monopoly of cardiac CT among cardiologists and radiologists should be avoided. The equipment and the patients should be equally shared, and withholding any of them will result in patients missing chances of a useful test. Radiologists must acknowledge that many cardiologists (at least those involved in echocardiography and invasive cardiology) are intrinsically imaging experts with an extensive knowledge of cardiovascular physiology and pathology. Cardiologists must acknowledge that radiologists are experienced in cross-sectional and vascular imaging, and are able to interpret cardiac and extra-cardiac imaging findings as well as being necessary for technical excellence.
It is an acknowledged fact that the future of diagnostic imaging lies in the integration of different imaging modalities, usually performed and/or interpreted by different physicians with widely differing backgrounds. The ability to fuse the myocardial perfusion data from a stress SPECT or positron-emission tomography (PET) test with the images of the coronary tree from the cardiac CT enhances the information obtained by both tests. SPECT–CT fusion has already proved to be useful for the detection of guilty lesions, especially in multi- or small-vessel disease, as well as for decision-making and planning of percutaneous coronary intervention (see Figure 1).24,25 In the end, SPECT–CT fusion imaging requires the collaboration of cardiologists, radiologists and nuclear physicians, thus setting the stage for molecular imaging.
It should be borne in mind that technology is important, but even more important for imaging are the professionals who indicate, perform and interpret those techniques.