The application of nuclear medicine techniques to cardiology is based on the identification of the functional consequences of coronary stenoses, i.e. of myocardial ischaemia. In nuclear cardiology, the evaluation of myocardial perfusion with single-photon-emission computed tomography (SPECT) is the most commonly performed procedure. The SPECT study is currently performed with electrocardiogram (ECG) gating, which enables a simultaneous evaluation of myocardial perfusion and left ventricular function. Nuclear medicine techniques play an important role in the non-invasive diagnosis of coronary heart disease (CHD) and enable the identification of patients at high risk of cardiac events to be guided towards the appropriate therapy.
Coronary angiography is the reference standard for the evaluation of the coronary arteries, for both its high spatial and temporal resolution and the possibility for percutaneous revascularisation procedures. However, although morbidity derived from angiography-related complications is low, the invasiveness of the technique and its relatively high costs has orientated research towards non-invasive coronary imaging modalities.
In recent years, intensive research effort has been invested in the development of cardiac computed tomography (CT), a technique able to measure coronary calcium and obtain information related to coronary tree anatomy. In addition, the development of hybrid systems (dual-modality imaging) offers the possibility of co-registration of anatomical data obtained from CT and functional information from SPECT or positron emission tomography (PET) in a single procedure.
Evaluation of Atherosclerosis
The content of coronary calcium is associated with vascular damage and the formation of atherosclerotic plaque. The formation of calcified atherosclerotic plaque is an active process that can be identified during the various stages of its development. At the beginning of the 1900s it was observed that atherosclerosis is the only vascular disease associated with coronary calcification. Coronary calcium is very common in patients with known CHD and is correlated with patient age, increasing drastically after 50 years of age in men and 60 years of age in women.1 Cardiac CT is able to obtain a quantitative measure of coronary calcium. Calcified atherosclerotic plaque is defined as a lesion above the threshold of 130 Hounsfield units (HU) with an area ≥3 adjacent pixels (a minimum of 1mm2). The two systems for measuring coronary calcium are the Agatston score and the ‘volume’ score.
The Agatston score is the result of the product of the area of each calcified focus with a numeric density peak at CT (score 1 for 13–199 HU, score 2 for 200–299 HU, score 3 for 300–399 HU and score 4 for ≥400 HU). The sum of all of these values in all the identified lesions provides the total coronary calcium or calcium score.2 The following categories have been standardised: minimal (score 1–10), mild (score 11–100), moderate (score 101–400) and severe (score >400). In clinical practice the score is often given as a percentile based on age and sex. The calcium score can be reported for each individual coronary artery and for the entire coronary tree, even though most studies report the data as the total score of the coronary tree.
The calcium volume score is a parameter independent of calcium density, and is therefore considered more appropriate in the serial evaluation of progression or regression of atherosclerosis. Comparing the results of CT with histological findings of plaque and the severity of vessel stenosis, it has been shown that the total area of the calcified atherosclerotic plaque is correlated with the total area of the plaque.3 In addition, not all plaques are calcified and the total area of the calcium accounts for approximately 20% of the total area of the atherosclerotic plaque. The comparison between intravascular ultrasound (IVUS) and CT confirms the direct association in vivo of the calcium score with the location and extension of atherosclerotic plaque.4 The prevalence of calcified atherosclerotic plaque, measured with the calcium score, reflects the prevalence of coronary atherosclerosis and increases with increasing age. However, atherosclerotic plaque and calcification are weakly correlated with the extension of coronary stenosis at histopathology examination.5 The calcified segments are not necessarily stenotic and vice versa.6 In addition, non-calcified atherosclerotic plaques cannot be detected with low-density X-rays. It has been shown that the calcium score has an elevated sensitivity but low specificity.
Coronary calcium evaluated with the calcium score indicates the presence of atherosclerosis, but not necessarily the presence of significant stenosis. The process of atherosclerotic progression described by Glagov can explain this phenomenon.7 In the early stages of the atherosclerotic process, the accumulation of atherosclerotic plaque induces vascular remodelling, with an outward expansion of the external elastic membrane. During this phase there is only minimal compression of the vessel lumen, but with progression of the atherosclerotic process the vessel reaches a limit of outward expansion. This phenomenon explains the weak correlation between calcium score and the presence of coronary stenosis ≥50% in the three major coronary arteries. If we were to consider coronary stenoses ≥20%, the calcium score would show a much closer correlation with the angiographic extension of coronary atherosclerosis. This suggests that the evaluation of the calcium score enables the diagnosis of atherosclerotic lesions before they become haemodynamically significant. The probability of having haemodynamically significant CHD increases proportionally with the increase in the calcium score, as this parameter is correlated with the atherosclerosis present in the coronary plaque.8
A multicentre study with a population of 1,851 symptomatic patients evaluated the diagnostic value of CT, reporting a sensitivity of 79% and specificity of 62%.1 Another study with 1,764 symptomatic patients compared the calcium score with angiography, reporting for calcium score values >100 a sensitivity of 95% and specificity of 79%.8 In symptomatic patients, ruling out the presence of coronary calcium can be important for avoiding invasive angiography. Indeed, studies with over 7,600 symptomatic patients demonstrated a negative predictive value of 96–100%, and in the presence of a calcium score <100 the probability of a stenosis >50% at coronary angiography is <1%.1,8 The American Heart Association has proposed that the evaluation of the calcium score with CT be indicated (class IIb) in the following situations in symptomatic patients: in the presence of an equivocal exercise test or functional test (SPECT), to determine the cause of cardiomyopathy and in patients with acute chest pain and a non-diagnostic ECG and negative biochemical markers.9
Recently, there has been growing interest in the visualisation and characterisation of coronary plaque with non-invasive techniques. Various studies have demonstrated the ability of multidetector CT (MDCT) to visualise atherosclerotic plaque. One pilot study demonstrated that the technique can classify atherosclerotic plaques as calcified, non-calcified and mixed with the same level of accuracy as IVUS. MDCT has demonstrated a good correlation with IVUS in terms of both plaque area (r=0.73) and the percentage of lumen obstruction (r=0.61), with a sensitivity of 84% and specificity of 91% in identifying non-obstructive atherosclerotic plaques even in the most distal segments.10 Some plaques can be at high risk of erosion or rupture even when these lesions are not associated with a significant degree of coronary stenosis, and can play a crucial role in the development of acute coronary events. It has been noted that the unstable plaque has elevated lipid content, and MDCT can be useful in characterising plaque composition. Various studies have compared ultrafast CT and MDCT with IVUS in identifying fibrous and soft atheromas.11,12 In these studies, the sensitivity in diagnosing calcified atherosclerotic plaques was higher than that of non-calcified plaques, varying between 88 and 95%. As the segments containing only non-calcified atherosclerotic plaque were identified with a sensitivity of 53%, it follows that MDCT tends to underestimate atherosclerotic plaque burden, particularly when the plaques are located in the most distal vessels.11 The evaluation of the progression of calcified atherosclerotic plaque and the identification of non-calcified atherosclerotic plaque requires further clinical evidence. The limitations of using CT to diagnose non-calcified plaques is not so much a question of the limited sensitivity of the technique, but rather a question of a lack of repeatability and prognostic data compared with conventional risk factors, the evaluation of the calcium score and the severity of coronary stenosis. Currently, therefore, its use is not recommended in this setting.9
Evaluation of Ischaemia
Nuclear medicine techniques are based on the identification of myocardial ischaemia by inducing myocardial perfusion anomalies or left ventricular contractile dysfunction with physical or pharmacological stress. In the presence of haemodynamically significant coronary stenosis, blood flow at rest is still preserved, whereas the increase in oxygen demand during stress induces an imbalance between the demand and supply of oxygen to the myocardial tissue, thus triggering a gradual sequence of changes defined as the ischaemic cascade. The initial signs of the ischaemic cascade are myocardial perfusion abnormalities, followed by diastolic function abnormalities and then systolic and left ventricular abnormalities. The electrical changes (visible at the ECG) and angina are later manifestations. The imaging modalities available for identifying the initial perfusion changes induced by the ischaemic event are ECG-gated SPECT and PET.
According to current guidelines, the implementation of an imaging stress test in patients with suspected CHD depends on the pre-test probability of disease, assessed by Bayesian analysis of the patient’s cardiovascular risk factors. A stress test (with or without imaging) is especially indicated in patients with an intermediate probability of disease. Patients with intermediate probability who are classified as low probability after a stress test require primary or secondary prevention, whereas those classified as having a high probability of CHD are indicated for coronary angiography. However, recent studies suggest that patients with a high pre-test probability of disease can benefit from undergoing a SPECT study. In fact, a normal SPECT identifies patients at low risk of future cardiac events, whereas an abnormal SPECT identifies patients at high risk who should undergo coronary revascularisation. SPECT performed with Tc-99m-labelled agents (sestamibi or tetrofosmin) is the most widely used nuclear medicine imaging technique in clinical practice for the evaluation of myocardial perfusion. A meta-analysis combining 79 studies for a total of 8,964 patients reported a sensitivity of 86% and specificity of 74% of myocardial SPECT for the diagnosis of CHD. The specificity is lower than the sensitivity, in agreement with the early manifestation of perfusion abnormalities.13 Furthermore, the low specificity values are in part correlated with an important post-test limitation, considering that most patients with a normal test are not indicated for coronary angiography. This means that the number of patients who are truly negative at the test will be relatively lower than the number of false-positive patients, with a consequent lowering of specificity. To overcome this problem, the evaluation of the normality rate, which indicates the frequency of normal tests in patients with a low to intermediate probability of CHD, can be used as a surrogate of specificity. This yields values corresponding to approximately 89%.13 Another frequent cause of low specificity values of SPECT is the presence of a relatively high number of false-positives due to soft-tissue attenuation artefacts, which are interpreted as perfusion defects. The most common attenuation artefacts originate from the diaphragm or breast tissue, or adipose tissue in obese subjects. Therefore, dedicated hardware and software have been developed to enable correct image reconstruction for different types of attenuation. The use of gating, with the simultaneous evaluation of perfusion and left ventricular function, has produced an improvement in the diagnostic accuracy of SPECT. In fact, with gating there has been a reduction in the number of false-positive tests, thus producing an improvement in specificity. Moreover, the addition of information in terms of systolic function to the perfusion data produces a substantial increase in the normality rate from 74 to 93% and a reduction in equivocal tests from 31 to 10%.13
The most important advantages of PET over SPECT are the systematic use of attenuation correction and the ability to obtain absolute quantification of perfusion in millilitre/minute/gram. PET is the ideal technique for evaluating cardiac physiology in that it has high spatial and temporal resolution and an excellent sensitivity and diagnostic accuracy. The most commonly used tracers for the perfusion study are N-13 ammonia, Rb-82 and [O-15] water. These tracers have a high-energy emission, so are particularly indicated in obese patients, and have a short half-life, thus guaranteeing that the tissues are exposed to radiation for a short time. PET is the only technique that allows the quantification of coronary flow at rest and coronary reserve during hyperaemia. The initial stages of atherosclerosis are marked by endothelial dysfunction, which can be evaluated with the absolute quantification of coronary reserve. PET can therefore be used to measure early atherosclerosis and identify patients at elevated risk of coronary impairment. A meta-analysis including 663 patients demonstrated a sensitivity of 89% and specificity of 86% of PET for the diagnosis of CHD.14 In addition, in patients with balanced myocardial ischaemia or diffuse CHD, the evaluation of coronary reserve with PET is able to identify those areas of the myocardium at risk that may remain undiagnosed at SPECT. The quantification of coronary reserve with Rb-82 has been shown to be more accurate in identifying the extension of CHD than SPECT in patients with three-vessel disease.15 The data in the literature demonstrate the superiority of PET over SPECT in diagnosing CHD, even though there are no data available comparing the two techniques since the recent developments of SPECT equipment. Various studies have shown that the two techniques have a comparable diagnostic capacity with the addition of attenuation correction of the SPECT images.16
Prognosis and Risk Assessment
The techniques of nuclear medicine provide important information in terms of patient prognosis. On the basis of the results of the test patients can be classified according to one of three categories: low, intermediate and high risk of future cardiovascular events. Patients at low to intermediate risk can benefit from conservative therapy, whereas high-risk patients are recommended to undergo more intensive diagnostic or therapeutic procedures.
The prognostic accuracy of gated SPECT in risk stratification is based on the fact that with a single test some of the most important prognostic factors of CHD can be obtained, namely the extension of necrotic myocardial tissue, the extension and the severity of inducible ischaemia and the measurement of left ventricular volumes and function. Numerous studies over the last two decades have reported that the extension and severity of perfusion defects in patients with suspected or known CHD have an important relation with major adverse cardiac events (MACE). A recent meta-analysis including 69,655 patients demonstrated that a normal SPECT is associated with a rate of MACE (death or non-fatal myocardial infarction) of 0.85% per year, which is comparable to that of the general population without signs of CHD. By contrast, in patients with an abnormal SPECT classified as moderate to severe, the rate of MACE is 5.9% per year. These data suggest that the risk of MACE is <1% per year in the presence of a normal SPECT and that this risk increases in relation to the extension and severity of the perfusion abnormality in the post-stress images.17 This concept has been confirmed in all patient groups and over the entire spectrum of procedures available in nuclear cardiology, including both pharmacological and physical exercise stress tests and different imaging protocols.18 In addition, the prognostic information obtained with SPECT completes and adds to the information obtained from clinical or exercise tests.19,20
The prognostic assessment of an imaging study should pay considerable attention to the type of stress test used. It has been observed that the pharmacological test is associated with a higher rate of MACE than exercise testing, thus suggesting that patients selected for the pharmacological test are more elderly and have a higher prevalence of risk factors (e.g. diabetes or prior myocardial infarction).21 With the introduction of the gated method, the integrated evaluation of left ventricular function with perfusion data has provided additional prognostic information in terms of the data derived from perfusion imaging alone. In a prognostic evaluation with gating, the post-stress ejection fraction is the best predictor of cardiac death, whereas the extension of perfusion defects is the best predictor of non-fatal myocardial infarction. In addition, end-diastolic and end-systolic volumes can provide additional prognostic information to the evaluation of ejection fraction.22 It has also been observed that the transitory dilatation of the left ventricle post-stress is associated with an unfavourable prognosis even in the presence of a normal SPECT. Gated SPECT provides information in terms of the benefit of treatment after risk stratification. Survival has been observed to improve with revascularisation if at least 10–15% of myocardial tissue shows signs of inducible ischaemia at SPECT, whereas no benefit has been observed if the patient undergoes revascularisation in the presence of mild ischaemia.23 This technique has a good cost–benefit ratio in the diagnostic algorithm of patients with an intermediate or high probability of CHD in comparison with the direct invasive approach. It has in fact been shown that in patients with an intermediate and high pre-test probability, the use of SPECT in association with selective coronary angiography produces a substantial reduction in costs (30–50%) in terms of an approach that only involves angiography.24
PET has also demonstrated a significant prognostic value in a number of studies. In a study population of 685 patients who underwent Rb-82 PET with dipyridamole, a survival of 90% was observed in patients with normal PET, whereas the survival for patients with a mild, moderate or severe test was 87, 75 and 76%, respectively. PET also provides prognostic information in addition to clinical and angiographic data.25 In contrast to SPECT, where an ischaemic response at ECG in the absence of a perfusion abnormality is associated with an increase in MACE, PET is associated with a favourable prognosis even in the presence of ischaemia at ECG.
Future Prospects
The recent development of hybrid systems offers the possibility of obtaining information in terms of coronary calcium and coronary anatomy with CT, and functional information in terms of cardiac perfusion and/or metabolism with SPECT or PET.26 SPECT can present artefacts associated with the attenuation of the diaphragm or breast tissue or suboptimal image quality in obese patients. Attenuation correction of SPECT images using dedicated hardware and software for image reconstruction is the best method for eliminating these artefacts. However, the images generated by these systems tend to have low resolution, which is why CT has been introduced to perform attenuation correction. The development of MDCT has orientated research towards hybrid devices, which integrate the functional data provided by SPECT or PET.
Currently, there are limited available data that demonstrate the additional benefits of hybrid systems and that justify the clinical use of the dual nuclear/CT imaging modality in the diagnosis of CHD or in cardiovascular risk assessment.27 Thus, further studies are required to establish the precise role of each of these techniques in the clinical evaluation of patients, especially in view of the possible integration of the data they provide that will play an increasingly important role in the management of these patients.