7 Nuclear Cardiology

Non-invasive images of the myocardium that reflect myocardial perfusion can be obtained either by using conventional nuclear medicine radiopharmaceuticals and cameras or by positron emission tomography (PET). Myocardial perfusion scintigraphy (MPS) with thallium-201 and/or technetium (Tc)-99m-labelled sestamibi and tetrofosmin, in combination with single photon emission computed tomography (SPECT), is a robust and well validated technique for the identification of myocardial ischaemia and infarction with high sensitivity and specificity. 99mTc-labelled myocardial perfusion agents have a high-count density which enables acquisition of electrocardiogram-gated images. Spatial and temporal changes in activity during the cardiac cycle reflect regional myocardial motion and thickening and this technique allows left ventricular volume, ejection fraction, and myocardial motion and thickening to be measured in addition to the information on perfusion. Since the main feature of an acute coronary syndrome is reduced myocardial perfusion, MPS can provide important diagnostic and prognostic information in the emergency department and allows patient stratification in the post-infarction phase. PET provides absolute measurement of myocardial blood flow and has enabled the demonstration of coronary microvascular dysfunction. This has highlighted the potential contribution of the microcirculation to myocardial ischaemia in patients with angiographically normal coronary arteries. Both SPECT and PET are invaluable tools for the identification of viable myocardium in patients with coronary artery disease and congestive heart failure. The constant technological developments of non-invasive cardiac imaging over the past years, including the advent of hybrid nuclear and computer tomography (CT) scanners now allow image fusion of CT coronary angiography and nuclear imaging which can be achieved by using hybrid scanners or software fusion of data sets obtained from stand-alone scanners. Although the potential of such comprehensive non-invasive coronary artery disease assessment appears great, the clinical impact of this tool remains to be established.

A range of investigations is normally used in patients with suspected coronary artery disease (CAD), the simplest ‘investigation’ being the history. Typical angina is a good indicator of myocardial ischaemia and abolition of symptoms is the primary aim of treatment. Symptoms however can be indeterminate and they do not indicate the site or extent of underlying ischaemia. It is therefore often helpful to proceed to further investigations to aid the diagnosis and guide future management. MPS is a robust, non-invasive, and widely available method of assessing regional myocardial perfusion, and has an obvious role in the clinical setting. Many studies have assessed the sensitivity and specificity of this technique for the detection of CAD—coronary arteriography usually being used as the standard by which the accuracy of scintigraphy is judged. The wisdom of this approach can be debated, but at least the arteriogram provides a universal standard for coronary anatomy even if it is less suited to assess coronary arterial function. Published figures for sensitivity and specificity of MPS vary widely and depend upon the characteristics of the population studied (gender, presenting symptoms, medication, previous infarction, etc.), the imaging technique used (planar or SPECT, qualitative or semi-quantitative analysis), and the experience of the centre. Using modern techniques with tomographic imaging good accuracy can be achieved with sensitivity and specificity as high as 91% and 87%, respectively [1]. This is significantly better than exercise electrocardiography for which a large meta-analysis has shown sensitivity of 68% and specificity 77%.

It is reasonable to ask therefore whether MPS should not replace exercise electrocardiography ( Chapter 25) in patients with suspected CAD. Several factors militate against this. The most important is the relative availability of the two techniques, but radiation burden and cost are also relevant. Although the cost of myocardial perfusion imaging (MPI) is higher than that of the exercise electrocardiogram (ECG) this is more than outweighed by its greater effectiveness [2]. Studies of cost-effectiveness have shown significant advantages for strategies of investigation using MPS, with savings in total diagnostic and management costs over two years in the region of 20% in centres routinely using scintigraphy.

Many centres use a staged approach with the exercise ECG being the initial stress test and MPS next if the likelihood of disease is indeterminate after the exercise ECG, or if further information on myocardial perfusion is required to assist management decisions. MPS should be the initial investigation in patients who are unlikely to exercise adequately, in women (because of the very high number of false-positive ECGs), and if the exercise ECG will be uninterpretable because of resting abnormalities such as left bundle branch block, pre-excitation, left ventricular hypertrophy, or drug effects.

The three commercially available perfusion tracers have equal accuracy for the detection of CA [3]. Thallium (²⁰¹Tl) has better uptake characteristics and, in theory, provides defects with greater contrast, but ^99mTc-labelled sestamibi and tetrofosmin images are superior in terms of resolution and susceptibility to attenuation artefact. The net effect of these technical differences in clinical practice is negligible, but the technetium tracers are preferred in obese patients or when ECG-gating is required. In fact, ECG-gating can aid the distinction between artefact and perfusion defect ( Fig. 7.1) and can increase confidence in reporting [4]. Attenuation correction is another technique that can reduce artefact although it is controversial whether this can be achieved without loss of sensitivity and attenuation correction is not used routinely in most centres.

Soft-tissue attenuation in the chest produces regional inhomogeneities in the normal pattern of tracer uptake and is one of the most frequent causes of artefact in MPS. Attenuation refers to the combined effects of photoelectric absorption and Compton scattering. The former occurs, when a photon interacts with an orbital electron in the tissue and the total energy of the photon is lost. The latter indicates interaction of a photon in the patient prior to detection, which makes the photon change direction. If patient positioning for rest and stress acquisition is kept constant, soft-tissue attenuation appears as a fixed defect. The resulting uncertainty in differentiating between a fixed defect due to attenuation artefact and myocardial infarction can reduce the specificity of the test for detecting CAD. Several methods of non-uniform attenuation correction are now available commercially, albeit with variable clinical success. Most of these methods use attenuation maps based on radionuclide line sources. Recently, however, the use of CT attenuation correction has been introduced and established for SPECT ( Fig. 7.2) as well as for quantitative myocardial perfusion measurement with PET. The transmission scan can by CT can be acquired in a much shorter time and with higher quality than can be obtained from a conventional radionuclide transmission scan, because CT achieves a far higher spatial resolution and higher photon flux. It has been shown that the respiration-dependent change in attenuation maps to correct SPECT is a possible drawback of CT-based attenuation correction. The attenuation map obtained with the lower resolution X-ray-based CT data of hybrid SPECT with low-end CT systems represent an average over many breathing cycles, similar to the attenuation maps acquired with Germanium sources in conventional PET scanners. By contrast, data acquisition with a modern multislice CT scanner occurs within fractions of a breath-hold. As a consequence, even slight misalignments may have a large adverse impact on image quality. Therefore, careful verification of alignment and meticulous manual correction for any misalignment is important ( Fig. 7.3). Under this condition CT yields good results. Although several studies of attenuation-corrected SPECT have demonstrated improved specificity with no change in overall sensitivity, attenuation correction is not yet widely used. The relative capabilities of gated SPECT and attenuation correction to improve diagnostic specificity are still uncertain. The fact that from most vendors SPECT/CT scanners, combining a multihead gamma camera with a CT facility, are now available will contribute to more widespread use and increase in knowledge about the clinical validity of X-ray-based CT attenuation correction.

^99mTc-labelled myocardial perfusion agents are valid alternatives to ²⁰¹Tl for the assessment of CAD. Their high-count density has enabled acquisition of ECG-gated SPECT studies. ECG-gating represents a great step forward in the evolution of functional myocardial imaging as it allows simultaneous assessment of resting ventricular function and either stress or rest perfusion. Due to the limited resolution of MPS with SPECT it is not possible to assess the left ventricular wall thickness by geometric methods. Because ^99mTc-labelled tracer (tetrofosmin and sestamibi) distribution in the myocardium is stable, spatial and temporal changes in activity during the cardiac cycle reflect regional myocardial motion. An increase in regional myocardial activity from diastole to systole is proportional to wall thickening. Thus, in addition to perfusion data, gated SPECT offers the potential to measure left ventricular volume, ejection fraction, and myocardial motion and thickening [5].

A precise and reliable assessment of left ventricular function and LV dimensions is prognostically important in most cardiac diseases. Quantifying the degree and extent of left ventricular function offers an objective risk stratification and selection of therapeutic strategy and allows for the sequential follow-up of the therapeutic response. Prerequisites for successful ECG triggering allowing quantification of left ventricular volumes and function are adequate count density (low-dose protocols may not provide valid numbers and should not be triggered for clinical use) and a fairly regular heart rhythm (atrial fibrillation, sinus arrhythmia, frequent premature beats, intermittent pacing etc. are conditions that may render appropriate triggering impossible). The ECG lead should be carefully chosen so that the R wave is a marker of end-diastole and the R wave must be positive in most triggering systems. The cardiac cycle is usually divided into eight, and sometimes into sixteen intervals (frames, bins). The lower frame rate results in a slight underestimation of the ejection fraction by about four units. Sixteen intervals provide better determination of ejection fraction and end-systolic volume (enhanced temporal resolution) and offers some information on the diastolic function while eight intervals might provide a better assessment of regional wall motion (due to enhanced signal-to-noise ratio of gated SPECT images) and is probably the most frequently used protocol. Commercially available, widely used, standardized, and well established software solutions can be used for automated detection of endocardial and epicardial contours reliably and without user intervention. These fully automatic methods of measuring left ventricular function ( Fig. 7.4) have been extensively validated against a variety of techniques, such as equilibrium and first-pass radionuclide ventriculography, X-ray contrast ventriculography, magnetic resonance imaging, and two-dimensional echocardiography. Wall motion analysis aids the distinction between attenuation artefact and true perfusion abnormality because infarcted myocardium is unlikely to move or thicken normally and hence reporting confidence is increased and additional prognostic information is obtained. It appears that wall thickening is the best parameter derived from gated SPECT (and superior to wall motion alone) [4], which increases test specificity for CAD assessment by allowing better differentiation of scars from attenuation artefacts when interpreting the cause of fixed defects (see Fig. 7.1). Recently, phase analysis of gated SPECT MPI has been introduced and shown to compare well to tissue Doppler imaging for the assessment of left ventricular dyssynchrony. The latter is an important predictor of response to cardiac resynchronization therapy (CRT). Information on left ventricular dyssynchrony can be provided by gated SPECT with phase analysis of regional maximal count changes throughout the cardiac cycle, which tracks the onset of wall thickening [6]. This method may play a major role in the near future for predicting response to CRT in heart failure patients.

The most commonly used radiopharmaceuticals for PET imaging of myocardial perfusion are ¹⁵O-water, ¹³N-ammonia, and rubidium-82 (⁸²Rb). For the latter two tracers, sensitivities between 83–100% for the detection of CAD have been reported with specificities between 73–100%.

¹³N-ammonia and ¹⁵O-water are the most commonly used PET tracers for the quantification of regional myocardial perfusion. They have similar half-lives of 10min and 2min respectively and so they both require an on-site cyclotron, which limits their widespread use. ¹⁵O-water is superior to ¹³N-ammonia as a perfusion tracer because it is metabolically inert and it diffuses freely across capillary and sarcolemmal membranes. It equilibrates rapidly between the vascular and extravascular spaces and its myocardial uptake varies linearly with perfusion over a wide range. But ¹⁵O-water has an important shortcoming compared with ¹³N-ammonia: it does not accumulate in myocardial cells and it does not therefore provide images for clinical use. In contrast, ¹³N-ammonia accumulates in myocardial cells and provides high quality images of perfusion ( Fig. 7.5). Therefore, it is the preferred tracer for clinical use provided that a cyclotron is available. The problem of attenuation correction has been solved for PET by using external ⁶⁸Ge or X-rays sources as recently established in the hybrid PET/CT scanners [7].

The main advantages of ⁸²Rb are its short half-life of 78s and the fact that it is readily produced at the point of use by a ⁸²Rb generator without the need for a cyclotron. Although several methods of quantifying regional myocardial perfusion using ⁸²Rb have been described, their accuracy is limited by the dependence of myocardial extraction of this tracer on perfusion and on the metabolic state of the myocardium. The high energy of the positron emitted (3.15MeV) also reduces resolution of the images because of the long track of the positron before annihilation with an electron. Nevertheless, ⁸²Rb is now widely used in the USA while this tracer has had no commercial success in Europe so far. The fact that the prognostic value of PET perfusion scanning with ⁸²Rb has been fully established may help to increase its acceptance in Europe. This in turn may help to increase the use of perfusion PET, as with the widespread use of oncology PET scanning the availability of PET scanners has substantially improved [8]. Myocardial perfusion scanning with ⁸²Rb and PET has several advantages over conventional SPECT. It is possible to perform a complete stress and rest study within about 30min, aiding patient comfort and throughput. This straightforward ⁸²Rb protocol compares favourably with cardiac ^99mTc SPECT, as the latter generally is a multistage procedure which may take half a day or may require acquisitions on two separate days.

Because of the higher resolution of PET and its integrated attenuation correction, accuracy for the detection of CAD is thought to be superior to SPECT although only a small number of studies have directly compared the techniques [3] and it is not known if its higher cost outweighs its greater accuracy. In complex conditions of coronary disease where there may be no normal reference segment such as in multi-vessel disease or coronary microcirculatory dysfunction PET is a preferred tool ( Fig. 7.6) and has been recently proven to be of great value for clinical decision-making and to be cost effective [9]. Quantification may allow the demonstration of endothelial dysfunction before an anatomical stenosis is apparent and it has had a great impact on our understanding of the pathophysiology of coronary disease [10]. In daily clinical routine, however, quantification of absolute myocardial perfusion for CAD assessment has played a minor role, as its added value on top of MPI remains to be elucidated. This may be achieved in the near future due to the fact that availability of PET devices has much increased mostly due to the widespread use in oncology.

An ideal non-invasive technique for the diagnosis of CAD should provide complementary information on coronary artery anatomy and pathophysiologic lesion severity. This is usually achieved by mental superposition of the information from coronary angiography with that from nuclear MPI. However, standardized myocardial distribution territories correspond in only 50–60% to the real anatomy coronary tree. Multislice X-ray computed tomography (MSCT) has emerged as a valuable alternative to conventional angiography with excellent accuracy in selected patients [11].

The combined use of two data sets both equally contributing to image information is the generally preferred definition of hybrid imaging. By contrast, in the setting of MPI with X-ray based attenuation correction the CT part of the imaging does not provide added anatomical or functional information, but is merely used to improve image quality of the other modality (SPECT or PET). If CT is solely used for attenuation correction of MPI acquired in a PET/CT scanner, the term hybrid imaging should probably not be used as attenuation correction with 68Ge sources used in the previous generation of PET scanners provided the same information as parametric maps obtained from low-dose CT, but this was not perceived as hybrid imaging due to the lack of topographic image information. Furthermore, the term ‘hybrid imaging’ does not seem appropriate for side-by-side analysis of MPI and CT images, but it has rather been suggested to be used in describing any combination of structural and functional information beyond that of attenuation correction.

The advent of hybrid scanners which are an integration of SPECT or PET with CT reflect the growing interest for the fusion of image data sets from nuclear and CT. Combined with the advancements in fast-processing software for three-dimensional reconstructions, this has allowed initial promising attempts of purely non-invasive CAD assessment directly relating individual myocardial wall territories to the subtending coronary artery by use of SPECT and CT or PET and CT [12].

The increasing interest in cardiac fusion imaging after establishing its clinical feasibility is currently raising the question of its clinical usefulness, which has been confirmed in early preliminary studies [13]. Such evaluation seems pertinent, as the integration of SPECT or PET devices and high-end CT scanners into hybrid scanners will promote the combined use of both techniques in the same patient. Alternatively, however, software solutions for fusion of SPECT or PET information with CT coronary angiography may allow the combination of image sets obtained on separate non-integrated stand-alone scanners. This can be achieved using commercially available software, which has been recently validated [14] ( Fig. 7.7). It reliably allows superposition of myocardial segments depicted by SPECT onto cardiac CT anatomy, resulting in an easily interpretable panoramic view of the heart, integrating the high-resolution three-dimensional information of the coronary arteries with the functional information of the SPECT perfusion image. This may facilitate a comprehensive non-invasive assessment of CAD yielding complementary information on a coronary lesion and its pathophysiologic relevance.

Despite many technical advances in invasive coronary angiography, the definition of functionally relevant coronary stenoses by purely morpho-anatomical criteria remains controversial. Although it is generally accepted that a coronary stenosis >50% may start to be haemodynamically relevant, many factors that cannot be fully explored by coronary angiography (including both invasive and CT angiography) will eventually determine whether a given lesion produces stress-induced ischaemia or not. The current approach to mentally match angiographic findings with the SPECT perfusion images faces many difficulties as the planar projections of coronary angiograms and axial slice-by-slice display of SPECT images may lead to inaccurate allocation of the coronary lesion to its subtended myocardial territory ( Fig. 7.8). Although fusion of invasive coronary angiography with SPECT has repeatedly been attempted, the warping and three-dimensional unification to force a planar two-dimensional angiogram into a fusion with a three-dimensional perfusion scan data set proved technically unsatisfying. In addition, such an approach would not allow non-invasive preplanning of the intervention as the information on the coronary anatomy is obtained by invasive coronary angiography, when rapid decision making during an ongoing procedure should not be slowed and delayed by the need of time consuming offline analysis. This may explain why such techniques which do not allow careful non-invasive planning of the elective intervention have not been adopted into daily clinical routine. The continuing rapid evolution of CT angiography suggests that, when combined with perfusion imaging, it has the potential to be implemented into clinical practice. This may further help to reduce the frequency of unnecessary angioplasty and stent placement as it should allow evidence driven intervention targeting relevant lesions only ( Fig. 7.9). First clinical results appear encouraging, supporting that hybrid images offer superior diagnostic information with regard to identification of the culprit vessel with the haemodynamic relevant lesion and increase diagnostic confidence for categorizing intermediate lesions and equivocal perfusion defects. Thus, the greatest added value seems to be firm exclusion of haemodynamic relevance of coronary abnormalities seen on CT angiography. Results from a first multicentre study underline the value of a combined functional and anatomical approach even without hybrid imaging showing that this combination allows improved risk stratification [15]. The clinical usefulness in terms of impact on treatment strategy and subsequently on outcome by hybrid imaging remains, however, to be determined in long-term studies. Similarly, it remains uncertain at this point whether hybrid scanners offer advantages over software fusion of data sets obtained from different scanners, as by either way one can obtain hybrid images ( Fig. 7.10). The discrepancy between emission from SPECT and CT transmission scan times determines that high-end CT facilities constituting the CT component of hybrid cardiac scanners will be blocked by long emission scan time and is therefore forced to operate at low capacity. On the other hand, a combined device may fit into one room, needs one operating team, and does not require positioning of the patient into two different scanners. The development of ultrafast SPECT scanners allowing substantially shorter acquisition time may shift the balance towards hybrid scanners in the future.

The majority of patients presenting to emergency departments with chest pain are admitted because the initial clinical examination, ECG results, and cardiac enzyme levels are insufficient to exclude an acute coronary syndrome, although most patients without obvious ECG changes do not have an acute syndrome. Conversely, a substantial minority of patients who are discharged from the emergency department have undetected acute ischaemia and an adverse outcome. Because the main feature of an acute coronary syndrome is reduced myocardial perfusion, MPS in the emergency department can provide important diagnostic and prognostic information. It has not been used widely because of the logistical problems of providing an acute radionuclide imaging service, but several studies have now shown the effectiveness and cost-effectiveness of MPS in the acute setting, especially when the resting ECG is not diagnostic of myocardial ischaemia. A resting perfusion defect has a high positive predictive value for acute infarction in patients without a history of previous myocardial infarction, particularly if it is associated with a wall motion abnormality on gated imaging, and these patients should be admitted to the coronary care unit. Conversely, a normal perfusion scan excludes acute infarction and exercise ECG or stress MPS can be the next diagnostic steps. If the perfusion tracer can be injected during chest pain, a normal perfusion scan excludes a cardiac cause and allows the patient to be discharged. In patients with symptoms suggestive of an acute coronary syndrome, acute MPS reduces unnecessary hospital admission without reducing appropriate admission of patients with a genuine acute coronary syndrome [16]. The sensitivity of acute rest MPS for the diagnosis of myocardial infarction is high very early after the onset of ischaemia, in contrast to serum enzyme markers, which require several hours to become clearly abnormal. Patients discharged with normal MPS have a very low likelihood of future cardiac events whereas patients with abnormal scans are at higher risk [17].

An intriguing option in patients with acute chest pain that has settled is to perform SPECT with free fatty acids (e.g. ¹²³I-(p-iodophenyl)-3-(R,S)methyl-pentadecanoic acid [BMIPP]) since fatty acid metabolism is reduced for some time after acute ischaemia has resolved. This ‘metabolic memory’ might allow diagnosis for up to 24 hours after ischaemic chest pain and the theory is proven in principle although it has not been widely applied.

Because the prognosis of ST segment elevation myocardial infarction (STEMI) is determined by left ventricular ejection fraction (LVEF), infarct size, and residual viable myocardium, radionuclide techniques provide important information that aids patient management. MPS provides additional prognostic information over clinical factors and LVEF and coronary angiography may not provide prognostic information beyond this. MPS with vasodilator stress allows risk to be assessed safely 2–5 days after infarction and is superior to early sub-maximal exercise testing. Even when used a few weeks after infarction MPS.

Patients with small, fixed perfusion defects have a good prognosis, and are unlikely to benefit from invasive investigation and revascularization. Conversely, patients with MPS markers of high risk can be referred for coronary angiography and possible revascularization, although the superiority of revascularization over medical therapy has not been established in this setting. Although primary percutaneous coronary intervention is the treatment of choice in STEMI, it is not currently available in all centres and some patients present too late for alternative thrombolysis. When this is the case MPS is very helpful for risk stratification and a large prospective randomized trial (INSPIRE—AdenosINE Sestamibi SPECT Post InfaRction Evaluation) using gated SPECT MPS and adenosine stress has determined the value of MPS to assess risk [18] and to guide subsequent therapeutic decision making in clinically stable patients early after acute myocardial infarction [19]. The INSPIRE results establish that the perfusion variables obtained by MPS, i.e. total and ischaemic perfusion defect size, improve the precision for assessing risk beyond that provided by the Thrombolysis In Myocardial Infarction (TIMI) risk score alone or when combined with LVEF. The cornerstone of risk stratification in stable survivors of an acute myocardial infarction is rapid discrimination of patients at high risk who might benefit from coronary revascularization from low-risk patients for whom medical therapy and early hospital discharge is appropriate. The INSPIRE trial proved that this goal can be achieved using MPS. In the interventional part of the trial confined to high-risk patients, both intensive medical therapy and revascularization produced comparable reductions (and frequently elimination) of both total and ischaemic perfusion defect sizes on follow-up scans. This resulted in no difference between the two groups with regards to total cardiac events, cardiac death, and reinfarction. In hospitals without cardiac catheterization facilities, MPS can identify those patients who do not require transfer to another facility for cardiac catheterization and can be discharged safely at an early date. In unstable angina and non-STEMI an early invasive strategy is recommended for patients with indicators of high risk and no serious comorbidities and this can be assessed by exercise ECG and by MPS. MPS is particularly useful for risk assessment of unstable angina once stabilized.

The exercise ECG has moderate specificity for the detection of CAD in the absence of resting repolarization abnormalities, left ventricular hypertrophy, and if patients are not treated with digoxin. Thus, when the resting ECG is normal and the likelihood of CAD from clinical assessment is low (for instance <25%) a stepwise strategy is appropriate with an exercise ECG as the initial diagnostic test. When the likelihood of CAD is very low (for instance <10%) then the best strategy will be reassurance without any provocative testing. If the resting ECG is abnormal or the likelihood of CAD is >25% then MPS may be the better initial test on grounds of cost-effectiveness [2].

The exercise ECG has lower specificity for the detection of CAD in women than in men and so MPS is a better diagnostic test even at lower likelihoods of disease. Pharmacological stress MPS is particularly valuable in women who are unable to exercise maximally. Although perfusion images are susceptible to breast attenuation artefacts, specificity can be maintained with awareness of the potential for artefacts, by using of ^99mTc perfusion tracers rather than ²⁰¹Tl, and employing ECG-gating and attenuation correction. Sensitivity of MPS is similar in men and women. PET perfusion imaging, when available, may be an additional way of avoiding attenuation artefacts.

Patients unable to exercise because of physical limitations such as arthritis, amputations, peripheral vascular disease, or pulmonary disease should undergo MPS with pharmacological stress as the initial diagnostic test. Inability to exercise is itself an adverse prognostic indicator, presumably because of the increased incidence of CAD, and this should be borne in mind when interpreting MPS in these subjects.

Patients with conduction abnormalities such as left bundle branch block, bifascicular block, and paced rhythms may have inducible and fixed perfusion abnormalities on MPS even in the absence of underlying CAD, particularly when imaged during exercise or dobutamine stress ( Fig. 7.11). Similar defects are much less common in patients with right bundle branch block although they can occur. These defects most commonly are confined to the septum although they can be more extensive. The causes of these defects in patients with conduction abnormalities are still uncertain and likely to be multifactorial, but they generally reflect true perfusion heterogeneities related to delayed septal relaxation and shorter diastolic perfusion time, or possibly to reduced regional afterload and hence reduced myocardial oxygen demand. Fixed defects may be due to reduced myocardial thickening or they may result from an underlying myocardial abnormality such as cardiomyopathy.

The specificity of MPS for the detection of CAD is therefore reduced in these patients when dynamic exercise is used, but specificity is maintained using vasodilator stress if the heart rate does not increase significantly. In practice, when there is a diagnostic problem in a patient with left bundle branch block or paced rhythm, MPS should be performed with adenosine or dipyridamole without additional exercise. A normal study excludes underlying coronary obstruction but an abnormal study may be less helpful diagnostically.

Myocardial ischaemia in patients with type 2 diabetes is often asymptomatic and frequently in an advanced stage when it becomes clinically manifest. Once coronary artery disease is symptomatic in diabetes, it is associated with a morbidity and mortality significantly higher than in patients without diabetes. In a large study in patients with type 2 diabetes (DIAD—Detection of Ischemia in Asymptomatic Diabetics trial) silent ischaemia was found in more than one in five asymptomatic patients [20]. In the follow-up a resolution of ischaemia occurred in the majority of these patients [21]. These unexpected results highlight the importance of studying whether a strategy of screening for inducible ischaemia impacts on clinical outcomes in such patients. So far it remains unclear whether the observed resolution of ischaemia which was associated with an intensification of the medical treatment was causally related to initial screening.

Beyond diagnosis, the most valuable contribution that MPS can make to the management of known or suspected CAD is to assess the likelihood of a future coronary event such as myocardial infarction or coronary death. Prognosis is strongly influenced by the extent and severity of inducible perfusion defects and this can guide the need for further invasive investigation and revascularization. MPS is a more powerful prognostic indicator than clinical assessment, the exercise ECG, and coronary angiography, and provides incremental prognostic value even once the other tests have been performed.

The most important variables that predict the likelihood of future events are the extent and depth of the inducible perfusion abnormality. The relative value of the fixed component of a stress defect is unclear, but it is likely that left ventricular function is the best indicator of prognosis in patients with predominantly fixed defects. Thus, the patient with extensive ischaemia is at high risk of a coronary event and sudden death irrespective of the presence of infarction, and the patient without ischaemia, but with a fixed defect is only at risk if the defect leads to significantly impaired function. Additional markers of risk are increased lung uptake on stress thallium images, since this indicates raised pulmonary capillary pressure either at rest or in response to stress, and ventricular dilation that is greater in stress thallium images than at rest. Transient ischaemic dilation can also be seen with technetium imaging and it may be the result of extensive sub-endocardial ischaemia giving the impression of cavity dilation.

In patients with known or suspected CAD, a normal perfusion scan is very valuable because it indicates a likelihood of coronary events of <1% per year [22], a rate that is lower than that in an asymptomatic population ( Fig. 7.12). Thus whether non-obstructive CAD is present or not, further investigation can be avoided. This negative predictive value is independent of the imaging agent and technique, the method of stress, the population studied, and the clinical setting. Exercise radionuclide ventriculography has also been used to assess prognosis because abnormal regional contraction is an early manifestation of inducible ischaemia. Stress LVEF provides more information than resting LVEF since it reflects the extent of both infarction and transient ischaemia. However, the comparative prognostic value of perfusion imaging and exercise ventriculography has not been fully assessed, although it has been suggested that knowledge of rest and stress LVEF from resting gated MPS and stress first-pass ventriculography provides additional prognostic information compared with the perfusion information alone.

A common clinical problem is that of assessing cardiac risk in patients that require non-cardiac surgery. In this as in other clinical settings, MPS provides useful information although these patients are generally at low cardiac risk and the predictive value of a normal perfusion study is greater than that of an abnormal study. Whether investigation beyond simple clinical assessment is required should be based upon the urgency of the surgery and its own cardiac risk, the risk factors of the individual, and his or her exercise tolerance. Patients with only minor clinical predictors (age >70 years, abnormal resting ECG, history of stroke or hypertension) who require low- (endoscopic or superficial procedures, cataract surgery, and breast surgery) to moderate-risk surgery (carotid endarterectomy, head and neck surgery, intraperitoneal and intrathoracic surgery, orthopaedic surgery, and prostate surgery), are not at high risk and do not require further investigation. Patients with intermediate clinical predictors (stable angina, prior infarction, treated heart failure or diabetes) or with minor predictors and impaired exercise tolerance need further assessment if they are to undergo moderate- or high-risk surgery. Patients at high clinical risk (recent infarction or unstable angina, decompensated heart failure, or significant arrhythmias) require investigation even for low-risk surgery.

For patients who are able to exercise, further investigation normally means exercise ECG, but if the resting ECG is abnormal or in patients who are unable to exercise, MPS should be used instead. If further testing suggests a low risk, then surgery can proceed as planned. If it suggests a high risk then the need for coronary angiography and revascularization is determined by the clinical setting. In general terms, revascularization should not be performed if it would not have been performed in the absence of surgery since the risk of revascularization may still exceed the risk of non-cardiac surgery. In patients at intermediate risk after further testing, the best strategy is uncertain but aggressive medical management at the time of surgery rather than revascularization is preferred. This medical management involves rigorous control of pain, fluid balance, and coagulation state after surgery, as well as preoperative beta-blockade and possibly perioperative nitrate infusion.

MPS can be valuable both before and after myocardial revascularization, either by angioplasty or bypass surgery. Neither procedure should be undertaken without objective evidence of ischaemia, and perfusion imaging is often the most reliable way of obtaining this information and of ensuring that angioplasty is targeted at the culprit lesion [23]. It has an excellent negative predictive value for predicting restenosis and clinical events after angioplasty, and this can be particularly helpful in patients with recurrent, but atypical, symptoms. Routine MPS after angioplasty in the absence of symptoms is not common, although it can sometimes be useful as a new baseline in case symptoms recur. It can however be justified routinely in patients with impaired left ventricular function, proximal disease of the left anterior descending coronary artery and multivessel disease, suboptimal results of angioplasty, diabetes, and those with occupations requiring low coronary risk. If MPS is performed after angioplasty, then it should ideally be performed at least 6 weeks after the procedure since perfusion abnormalities can persist for some time even with a good anatomical result. Possible exceptions to this are patients with high-risk anatomy who can benefit from earlier imaging.

As with angioplasty, patients who are asymptomatic after bypass surgery do not routinely undergo perfusion imaging, although it can be helpful as a baseline for future management since revascularization is not infrequently incomplete. More commonly it is used for follow-up and it can be used roughly 5 years after surgery to guard against silent progression of prognostically important disease. Patients with symptoms after surgery may certainly benefit from MPS and the algorithms to be used are very similar to those in the diagnostic setting.

The diagnosis of acute myocardial infarction is normally made from the clinical history, the ECG, and from cardiac enzymes. In most cases these provide a conclusive answer but the diagnosis can be unclear in those seen late after the onset of chest pain, those with a conduction abnormality or pacemaker, those with perioperative infarction, and those in whom right ventricular infarction is a possibility. Nuclear techniques may then be helpful.

A number of radiopharmaceuticals have an affinity for acutely necrotic myocardium. Imaging of ^99mTc-pyrophosphate has a sensitivity of at least 85% for the detection of acute infarction when performed 1–3 days after the event. Specificity is lower because uptake may occur in areas of old infarction or aneurysm, and also in areas of subclinical myocardial damage after unstable angina. Persistent blood pool activity or activity in bone and skeletal muscles can also cause difficulties, although these may be overcome by tomographic acquisition. In clinical practice, the technique is not used commonly, but it can be helpful in cases of doubt.

Imaging with antimyosin antibodies labelled with indium has also been used and it has both high sensitivity and specificity. A multicentre trial of 492 patients showed sensitivities of 94% in Q-wave infarction and 84% in non-Q-wave infarction. Specificity was 93% in patients with chest pain but no infarction and there was focal uptake in 48% of patients with unstable angina suggesting sub-clinical infarction [24]. Despite this, the long time that is required after injection to obtain images limits its use for the early detection of infarction. This is also a drawback when using this compound for the detection of myocarditis and transplant rejection.

Because ^99mTc labelled perfusion agents (MIBI and tetrofosmin) do not redistribute, they can be used in acute infarction to demonstrate the territory at risk before thrombolysis or acute angioplasty, and to assess the amount of myocardial salvage. The tracer is injected immediately before intervention and imaging can be performed several hours later once the intervention is complete and the clinical situation is stable. The defect will then correspond to the territory at risk and repeat injection and imaging several days later will show the actual extent of infarction. This is not a technique that can guide the need for intervention but it has been used in a number of clinical trials to assess the effect of acute intervention and to compare different regimens of thrombolysis on infarct size.

An important aspect of clinical management after infarction is to identify patients at high risk of further events such as re-infarction or death, and hopefully to intervene in order to prevent these events. Clinical indicators of high risk in the acute phase include hypotension, left ventricular failure, and malignant arrhythmias and these patients are candidates for early coronary angiography. After the acute phase however, prognosis is related to the degree of left ventricular dysfunction and the extent and severity of residual ischaemia, and radionuclide imaging can assess both objectively. LVEF at the time of discharge or 10–14 days after infarction is a strong predictor of mortality, and patients with impaired function in particular can benefit from MPS to assess whether viable but jeopardized myocardium remains in the infarct zone and whether remote territories may also be jeopardized by ischaemia.

The term’viable’ is an umbrella term that includes several different subtypes of myocardium. One of these is hibernating myocardium, which is chronically dysfunctional but still viable myocardium that recovers function after coronary revascularization. For many years the functional sequelae of chronic CAD were considered to be irreversible and amenable only to palliative therapy. For example, akinesis on the left ventriculogram implied infarcted myocardium or scar. We now know that chronic left ventricular dysfunction in patients with CAD is not necessarily irreversible and areas of akinetic myocardium have frequently been observed to improve in function after revascularisation.

In 1978 Diamond et al. [25] suggested the possibility that ‘ischaemic non-infarcted myocardium can exist in a state of function hibernation’. Several years later Rahimtoola [26] popularized the concept of hibernating myocardium and noted ‘there is a prolonged subacute or chronic stage of myocardial ischaemia that is frequently not accompanied by pain and in which myocardial contractility and metabolism and ventricular function are reduced to match the reduced blood supply’. It is now known that perfusion is not always significantly reduced at rest in myocardial hibernation, but the debate on whether resting myocardial blood flow to hibernating myocardium is reduced or not has attracted a lot of interest and, undoubtedly, has contributed significantly to stimulate new research on heart failure patients with coronary artery disease. Although the debate is not over yet, some of the initial paradigms have been shown to be incorrect while new pathophysiological concepts have emerged. Clinically, the concept of hibernation has made a significant contribution to our understanding and management of patients with advanced ischaemic left ventricular dysfunction. Failure to identify and rescue hibernating myocardium may lead to loss of viable myocytes, progressive deterioration of heart failure, and death. A number of imaging techniques have been used to detect viable myocardium and to characterize it as hibernating.

Initial studies indicated that myocardial hibernation and infarction could be distinguished by a combination of PET perfusion imaging using ¹³N-ammonia and metabolic imaging using the glucose analogue ¹⁸F-fluorodeoxy-glucose (FDG) after an oral glucose load. Regions with a concordant reduction in both ¹³N-ammonia and FDG uptake (‘perfusion-metabolism match’, Fig. 7.13) were predominantly infarcted, whereas regions with reduced ¹³N-ammonia uptake but preserved or increased FDG uptake (‘perfusion-metabolism mismatch’, Fig. 7.13) were hibernating [27]. Myocardial FDG uptake, however, depends on many factors such as dietary state, cardiac workload, insulin sensitivity, sympathetic drive, and the presence and severity of ischaemia. These factors lead to variable FDG uptake in the fasted or glucose-loaded state and complicate image interpretation.

Semi-quantitative and quantitative analyses of FDG uptake improve the detection of viable myocardium but require standardization of imaging conditions particularly with regard to myocardial glucose uptake. Many patients with CAD are insulin resistant and have poor FDG image quality after an oral glucose load. Myocardial glucose metabolism can therefore be standardized using the hyperinsulinaemic euglycaemic clamp, essentially the simultaneous infusion of insulin and glucose acting on the tissue as a metabolic challenge and stimulating maximal FDG uptake [28]. This allows absolute values of glucose metabolism to be measured (µmol/g/min) and aids comparisons between different subjects and centres ( Fig. 7.14). To determine the threshold value above which the best prediction of improvement in functional class of at least one grade could be obtained, in a prospective study in 24 patients undergoing coronary revascularization, a receiver-operator-characteristic curve (ROC) was constructed. According to this analysis the optimal operating point on the curve (point of best compromise between sensitivity and specificity) was at the absolute threshold of FDG uptake of 0.25µmol/g/min (where the gold standard was the evidence of functional recovery at follow up) [29]. By comparing FDG images obtained under these conditions with regional wall motion from another imaging technique, the need for a simultaneous perfusion tracer is avoided.

In summary, clinically there is now wide consensus on the importance of identifying and treating hibernating myocardium in patients with coronary artery disease and heart failure. Although randomized studies are needed before a definitive influence on clinical practice is achieved, the contribution of the existing experimental studies is compelling.

The disadvantage of PET is that it is not widely available. Thallium provides information on both myocardial perfusion and viability and has been widely used for identifying myocardial hibernation. Because redistribution can be slow or incomplete in regions of reduced perfusion, the usual stress/redistribution protocol can underestimate myocardial viability and additional steps to ensure complete assessment of viability are required. These include late redistribution imaging at 8–24 hours after stress injection, re-injection of tracer at rest after redistribution imaging, and a resting injection on a separate day with both early and delayed imaging.

In any of these viability images, the amount of viable myocardium is proportional to the amount of tracer uptake relative to a normal area. A common threshold for defining clinically significant viability is 50% of maximal uptake although the best threshold may be higher. In addition to detecting viable myocardium in an area of akinesis it is important to demonstrate inducible ischaemia before diagnosing hibernation since hibernation is an ischaemic syndrome.

MIBI and tetrofosmin have also been used for the detection of viable and hibernating myocardium. In theory these tracers may underestimate viability in areas with reduced resting perfusion because they are combined tracers of viability and perfusion without the property of redistribution that allows viability to be assessed independently. This results in a high positive but low negative predictive value. Some studies have therefore found thallium to be better for the assessment of viability but others have found them to provide comparable information. It does appear though that if the tracers are given under the cover of intravenous or sublingual nitrates, then resting perfusion is improved and the technetium tracers may be used as markers of viability.

An important problem in studies of hibernation is that viability and function are often assessed from different techniques, and it can be difficult to be sure that the same myocardial segments are being compared. Thus, the ideal technique should combine information on viability, perfusion, and function in a single image, and ECG-gated technetium MPS is very helpful. In regions of previous infarction with reduced tracer uptake, the assessment is more difficult, but this is not a major limitation since these areas contain little viable myocardium and may not benefit from revascularization.

Until recently, ischaemic heart disease was primarily thought to be caused by disease of large vessels, particularly the conduit coronary arteries. However, it is now clear that abnormalities of the coronary microcirculation may contribute to the generation of ischaemia even in the absence of demonstrable disease of the large epicardial arteries. Microvascular disease often precedes epicardial coronary disease and its extent may have independent prognostic value.

Myocardial perfusion reserve is the ratio of myocardial perfusion during maximal coronary vasodilation and at baseline. It is an integrated measure of flow through the epicardial coronary arteries and perfusion through the microcirculation and it can be used to assess the function of the coronary circulation as a whole. An abnormal perfusion reserve can be due to narrowing of the epicardial coronary arteries or, in the absence of angiographically demonstrable atherosclerotic disease, may reflect dysfunction of the coronary microcirculation. The latter can be caused by structural (e.g. vascular remodelling with reduced lumen to wall ratio) or functional (e.g. vasoconstriction) changes, which may involve neurohumoral factors and/or endothelial dysfunction. Furthermore, an abnormal perfusion reserve may also reflect changes in coronary and/or systemic haemodynamics as well as changes in extravascular coronary resistance (e.g. increased intramyocardial pressure).

The coronary microcirculation cannot be imaged directly in man in vivo. The resistance vessels in the coronary circulation are not generally visible on angiography and are too small to be catheterized selectively. Instead, indirect parameters such as myocardial perfusion and perfusion reserve can be used since, in the absence of coronary stenoses, they provide an index of microvascular function.

PET can be used to measure both absolute myocardial perfusion and perfusion reserve and microvascular dysfunction has been demonstrated in patients with hypercholesterolaemia, hypertension, diabetes, and smoking. The measurements can also be used as surrogate endpoints to assess the effectiveness of therapeutic interventions such as alpha- and beta-adrenoreceptor blockade [30], lipid-lowering, antioxidants [31], cardiovascular conditioning, and coronary angioplasty. They also provide prognostic information [32, 33] and microvascular dysfunction assessed by PET is an independent predictor of long-term outcome and cardiovascular death in patients with hypertrophic and dilated cardiomyopathies ( Fig. 7.15). The best established flow tracers are ¹⁵O-water and ¹³N-ammonia. The short half-lives (2min and 9.8min, respectively) require an on-site cyclotron for the use of these isotopes. As an alternative, ⁶⁸Rb has been introduced, which is a generator product and does therefore not need a cyclotron. For quantifying myocardial perfusion, however, ⁶⁸Rb is much less well established than ¹⁵O-water and ¹³N-ammonia.

For full references and multimedia materials please visit the online version of the book (http://esctextbook.oxfordonline.com).