Chapter 3. Myocardial Perfusion Single Photon Emission Computed Tomography and Positron Emission Tomography

Myocardial perfusion single photon emission computed tomography (SPECT) is a nuclear medicine imaging technique that uses a radioactive perfusion tracer to detect ischemia. Myocardial SPECT can detect, localize, and quantify the degree of ischemic myocardium. The result of a myocardial perfusion SPECT is related to the prognostic outcome of cardiac events (ie, myocardial infarction and death). Therefore, myocardial perfusion SPECT is used to guide treatment in patients with stable angina and after acute coronary syndrome. Myocardial perfusion SPECT is a mature technique, one of the most commonly used cardiac imaging techniques, and part of the international guidelines for treatment of patients with ischemic heart disease.1-3 There are international guidelines on how to perform,4 how to report,5 and when to use6 myocardial perfusion SPECT. Myocardial positron emission tomography (PET) is a newer nuclear medicine imaging technique that has rapidly become routine in cancer patients. The use in cardiology is also increasing because PET has inherent advantages compared to SPECT (eg, better image resolution and the possibility for absolute quantification of blood flow). However, the limited availability of the technique and especially of the PET perfusion tracers and higher costs are influencing the choice of method, and SPECT has kept its role as the main workhorse for cardiac patients.

Myocardial ischemia is defined as a mismatch between the demand and supply of oxygen to the myocytes, and this is mainly dependent on an adequate myocardial perfusion in various physiologic conditions. The inotropic and chronotropic states determine the myocardial demand, and during physical exercise, the heart rate increases and the force of the myocardium increases to raise the stroke volume and the left ventricular systolic pressure. At normal conditions, the arterioles can dilate during exercise, and perfusion can thus increase from a resting value of 1 mL/min/g to 3 mL/min/g. The maximal vasodilatation of the coronary arteries is even larger; over 5 mL/min/g has been measured, for example, during reactive hyperemia7 or pharmacologic vasodilatation. Coronary artery disease (or ischemic heart disease) affects myocardial perfusion by coronary artery stenoses that increase the resistance of the larger coronary arteries, and thus, the flow is limited. The heart will compensate for this increased resistance by dilatation of the arterioles at rest, thereby preventing a flow decrease. However, this means that part of the flow vasodilator reserve that is supposed to be used at stress is already being used at rest. This explains why patients who have stress-induced ischemia commonly have normal perfusion at rest.

Myocardial perfusion is depicted at rest and exercise using an intravenous injection of a radioactive perfusion tracer in a peripheral vein. The most commonly used perfusion tracers are technetium 99m (99mTc)–labeled tetrofosmin (Myoview; GE Healthcare, Little Chalfont, United Kingdom) or sestamibi (Cardiolite; Bristol-Myers Squibb, New York, NY). The formerly most used tracer, thallium, is an elementary compound with similar physiologic properties as potassium and provides a good linear relationship between blood flow and tracer uptake into the myocardium. The radiation dose is higher using thallium, and the image quality is not as good as that obtained with technetium-labeled tracers. The uptake of the perfusion tracer is related to the perfusion of the tissue, which means that only 3% of the injected radioactive material is distributed to the heart. Photons emitted from the radioactive decay (ie, gamma rays) are registered with a gamma camera. The resulting images are coded according to a gray or color scale from no perfusion (black) to maximum perfusion (white in the gray scale or bright colors in most of the color scales). A normal image is shown in Figure 3–1, and a yellow to white color denotes the highest tracer activity in the heart. Images are often displayed in a two-dimensional summary image of the heart called polar plot or "bull's eye" (Fig. 3–2). A normal finding is an even bright color in all parts of the heart during exercise. When one part of the myocardium shows decreased tracer uptake during stress, a perfusion defect, the images are compared against the images obtained at rest. A perfusion defect that is larger at exercise than at rest is interpreted as ischemia (Fig. 3–3A,B). A perfusion defect that does not change between exercise and rest is considered to represent infarcted myocardium but can also be linked with other causes such as hibernating myocardium (Fig. 3–4).

Uptake of the perfusion tracer to the tissues is related to the perfusion also in extracardiac tissues (eg, skeletal muscle and the gastrointestinal tract). This means that an injection at rest will have higher uptake to the gastrointestinal tract compared with an injection at physical exercise because of the higher blood flow to the gastrointestinal tract at rest. During exercise with high sympathetic activity, blood flow is lowered in the gastrointestinal tract and elevated in skeletal muscles. This will affect the image acquisition because the liver and intestines are close to the heart and the extracardiac tracer uptake in the image field of view can be much higher than the uptake of the heart. High extracardiac tracer uptake close to the heart, as a rule of thumb within the width of the heart wall, may necessitate renewed image acquisition. Therefore, before image acquisition, it is recommended to wait for more than 30 minutes after resting injection or injection during pharmacologic stress. In the waiting period, patients are recommended to take a walk in the corridors and eat or drink something to increase the gastrointestinal motility so that the tracer uptake moves away from the heart. After an exercise stress test, images can be acquired 15 minutes after injection.

Gated SPECT and Electrocardiogram Triggering

Electrocardiogram (ECG) triggering or "gating" is routinely used during the gamma camera acquisition, and the acquisition is called gated SPECT. The recorded data are placed into 8 to 16 bins or gates depending on the time after the R wave of the ECG. Gated SPECT enables studies of global and regional ventricular function. Endocardial contours are automatically generated and provide the left ventricular volumes over the cardiac cycle. By calculation of the end-diastolic volume and end-systolic volume, the ejection fraction can be derived (Fig. 3–3C). Left ventricular function is visualized by displaying the images in a cine loop. The information from gated SPECT has been shown to have incremental prognostic value compared with assessment of perfusion only.8 PET images are often ECG gated in the same manner to obtain functional information.

A perfusion defect that is unchanged between rest and stress can be due to infarction but also due to attenuation artifact, and gated SPECT can be used to differentiate these two (Moving Image 3–1). A myocardial infarct is characterized by reduction in perfusion tracer uptake and decreased regional function (Moving Image 3–2). If normal regional function is seen at gated SPECT in the area of reduced uptake, it is read as an attenuation artifact, and not as an infarct. Decreased ejection fraction and increased left ventricular volumes without any perfusion defects are seen in cases of cardiomyopathy (Fig. 3–5).

The summed images from gated SPECT are used for assessment of perfusion, and therefore no extra scanning is needed; perfusion and function images are obtained simultaneously. Gated SPECT uses information from many heart beats, and if the heart rhythm is not stable (eg, sinus arrhythmia, premature beats, or atrial fibrillation), gated SPECT cannot be performed. One should bear in mind that the perfusion images display the perfusion as it was at the time of injection. The gated SPECT functional images, however, show function at the time of acquisition. This explains why a perfusion defect representing ischemia can be seen with preserved function at gated SPECT. During the time from injection to scanning, function can be restored. A stress perfusion defect with decreased function compared with the resting images is indicative of postischemic myocardial stunning (Moving Image 3–3). Lower ejection fraction and larger volumes after stress compared with rest are signs of severe ischemia. The increased ventricular volumes caused by postischemic stunning can also be seen on the summed perfusion images as transient ischemic dilatation (TID). TID manifests as a larger blood pool at stress compared with rest; in the images, this is manifested as more of the dark lumen at stress compared with rest (see Fig. 3–3A,B).

New Detector Technology for SPECT

Recent developments have introduced ultrafast cardiac SPECT cameras with acquisition times down to a few minutes and increased spatial resolution.9,10 This development also provides the possibility to lower the injected radioactive dose to the patient, resulting in effective doses of only a few millisieverts. This has been made possible by new detector technology composed of cadmium-zinc telluride crystals with higher sensitivity, digital detectors instead of photomultiplier tubes, and improved collimator techniques that enable the acquisition of the entire three-dimensional volume continuously without moving the detectors in the conventional half-circle around the patient. The reconstruction algorithms for myocardial perfusion SPECT have also been refined, shortening the acquisition times by 30%. The cost of the new detector technology is roughly double the price for a conventional dedicated cardiac SPECT Anger gamma camera.

Attenuation and Its Correction

Radiation from different parts of the myocardium travels different distances through tissue with varying densities on the way to the gamma camera detector. If there is tissue of high density between the myocardium and the detector, some photons may not reach the detector. This is called attenuation of the signal and may mimic a perfusion defect.11 Because the body has the same constitution at both stress and rest images, the attenuation will be the same and a similar decrease in counts will be seen at stress and rest. Attenuation causes a decrease in the specificity of the test. Gated SPECT, additional imaging in the prone position, or transmission attenuation correction11 can be used to overcome this problem (Fig. 3–6). Attenuation correction can be performed by acquiring additional information about the body composition and attenuation properties and then mathematically calculating what the signal would be in the case of no attenuation. Traditionally, transmission information (as opposed to the emission perfusion information from the patient) from radioactive rods orbiting the patient and detected by the gamma camera on the opposite side of the patient has been used. Today, hybrid systems with computed tomography (CT) are available, and the CT information can be used for attenuation correction.12 The transmitted x-rays give a map of the attenuation of the body. Indeed, attenuation is the basis for all x-ray–based technology, and Hounsfield units in CT are a measure of attenuation. CT can be used to create an attenuation map of the patient, which can correct for attenuation of the photons by the tissues of the body. The information from a low-dose CT scan is superior to that obtained from radioactive rods and is acquired more rapidly.

A simple and inexpensive method to help in identifying attenuation effects without these technical options is to repeat the image acquisition in the prone position. A defect that is present in one position but moves or disappears in the other image acquisition is not caused by an infarct13,14 (see Fig. 3–6). Prone imaging has therefore been called the "poor man's attenuation correction."

PET has technical advantages over SPECT; the spatial resolution is higher, the acquisition is faster, the motion artifacts are less of a problem, attenuation correction is routine, and absolute quantification of perfusion is possible.15 There are several methodologic differences between SPECT and PET. PET scanners have a different geometry and a different detection principle. Multiple correction algorithms are routinely applied to yield images of absolutely quantitative tracer distribution of the target.

Acquisition

PET scanning aims to produce a three-dimensional image volume, which is an accurate map of the distribution of tracer in the body. To enable quantification, a series of such volumes is normally generated over time to describe the time-activity curves and investigate the kinetics of tracer uptake in different tissues and blood. The isotopes used in PET emit positrons, which are the antimatter of electrons. The positron is annihilated when colliding with an electron, whereupon two photons are emitted (annihilation radiation) and can be detected by the PET camera. The half-lives of the radioactive PET isotopes are much shorter than those of technetium or thallium used in SPECT (eg, 76 seconds for rubidium 82 [82Rb] and 10 minutes for nitrogen-13 ammonia [13NH3]). In addition, compared with radionuclides emitting single gamma-ray photons, the emission of pairs of 511-keV photons gives PET imaging higher detection efficiency, better uniformity of spatial resolution, and easier correction for attenuation. Gated images for assessment of regional and global function are acquired in the same manner as with SPECT (see previous section on SPECT).

Attenuation and Other Corrections

In PET, the photon attenuation correction is routinely used in a robust manner. To exploit the potential of PET to provide quantitative data, a number of other corrections need to be carried out. The geometry of a PET system introduces variation in the detection sensitivity. Normalization corrects this by measuring the count rate for each line of response using a radioactive source. The response time of a detector system is finite. Dead time is the period when a detector is unable to record a new event, and this needs to be corrected also. In addition, to achieve quantitative results, scattered events, random coincidences, and partial volume effects in small targets need to be corrected; these correction methods use mathematical modeling. Furthermore, motion correction is needed when a moving target such as the heart is imaged. Luckily, most of these corrections are performed by the scanners automatically. Partial volume correction for cardiac studies is commonly done after data modeling, and for motion correction, ECG and/or respiratory gating is available.

Exercise Stress Test

An exercise stress test is recommended as the method of choice when performing myocardial perfusion SPECT, and the perfusion tracer is injected during maximum cardiac exertion. To avoid injecting the perfusion tracer at submaximum exertion and thereby risking a false-negative test, injection should be done when the heart rate is above 85% of the age-predicted maximum. This can be calculated by the following formula: (220 – age) × 0.85. After injection, the patient should continue exercise for 1 minute to make certain that the perfusion tracer has entered the cells and left the blood pool. The workload can be lowered somewhat to enable the patient to continue exercise for 1 whole minute. Contraindications for an exercise stress test and criteria for termination of the test before reaching the age-predicted heart rate are the same as for a standard exercise stress test according to the international recommendations.16 Contraindications include any acute or ongoing severe cardiopulmonary disease, such as acute coronary syndrome, aortic dissection, cardiac failure or complex arrhythmias with hemodynamic instability, pulmonary embolism, severe hypertension (blood pressure >200/110 mm Hg), and severe aortic stenosis. Absolute indications for early termination of the exercise test include significant decrease in blood pressure, severe angina, sustained ventricular tachycardia, pathologic ST-segment elevation (>1 mm in leads with no Q waves other than V1 or aVR), increasing nervous system symptoms, or signs of poor perfusion. If there are technical difficulties affecting the ability to record ECG and blood pressure, the test should be terminated, and the problems should be fixed before continuing. Relative contraindications include marked ST-segment depression (>3 mm), arrhythmias such as triplets of premature ventricular contractions or supraventricular tachycardia, development of left bundle branch block, and severe hypertension (>250/130 mm Hg).

Pharmacologic Stress Test

Pharmacologic stress with adenosine, dipyridamole, regadenoson, or dobutamine can be used to cause coronary vasodilatation in patients who cannot perform a maximum exercise stress test, including patients with orthopedic problems and patients who cannot reach adequate heart rate because of β-blockers or a concomitant disease such as chronic obstructive pulmonary disease or kidney failure. A vasodilator drug is also used in patients with left bundle branch block or a pacemaker with ventricular stimulation because of potential false-positive perfusion images when using the exercise stress test (see "Pitfalls" later). Pharmacologic stress is used most commonly with PET imaging. Direct vasodilatation is caused by adenosine, dipyridamole, and regadenoson. Dobutamine is a sympathomimetic drug with positive inotropic and chronotropic properties that causes a secondary vasodilatation. Pharmacologic stress testing is safe under correct monitoring and patient supervision. ECG, blood pressure, and patient status should be assessed and supervised by a physician, and cessation of the drug and use of antidotes may be necessary.

Adenosine

Adenosine is the most commonly used pharmacologic stress agent and acts as a direct vasodilator on the coronary vessels by binding to the A2A receptor on the arterioles. Adenosine is given as a continuous infusion in a peripheral vein for 5 to 6 minutes. The half-life of adenosine is very short (<10 seconds), and therefore, no antidote is needed. After cessation of the infusion, the effect stops. Adenosine is given at a dose of 140 μg/kg/min, but the dose can be lowered if adverse effects occur. Common adverse effects to adenosine are flushing, headache, chest discomfort, and dyspnea. These adverse effects are unspecific and not related to ischemia; therefore, patients are informed of their benign nature, and most patients can tolerate them. Criteria for early termination of adenosine infusion are persistent high-grade atrioventricular (AV) block or sinoatrial block, severe hypotension, bronchoconstriction, and severe angina. Adenosine causes AV block in high doses, and especially in older patients, transient AV block is seen. Lowering the adenosine dose is often sufficient to avoid high-grade AV block. Blood pressure drops in many patients because of the vasodilatation, but there is rarely a need to stop the infusion. A profound decrease of blood pressure is related to a worse patient status and ischemia, and in some patients, adenosine infusion has to be terminated. Healthy subjects given adenosine do not experience a drop in blood pressure but do experience an increase in the heart rate and stroke volume to maintain the blood pressure. Bronchoconstriction is a rare but potentially severe adverse effect of adenosine and is the reason for the contraindication of adenosine in patients with active asthma or severe chronic obstructive pulmonary disease. In patients with mild chronic obstructive pulmonary disease, premedication with inhalation of β2 stimulants can be used. Auscultation of the lungs during adenosine infusion can be used to detect bronchoconstriction. Of note is that the dyspnea experienced by most patients is not related to bronchoconstriction. Dyspnea is believed to be caused by stimulation of the peripheral oxygen receptors in the carotid sinus. The safety of adenosine has been assessed in a large study of over 9000 patients; 0.1% of patients experienced bronchoconstrictions, although patients with asthma and severe chronic obstructive pulmonary disease were excluded from receiving adenosine. Two patients experienced severe complications with pulmonary edema and myocardial infarction, respectively. The combination of the adenosine infusion with physical exercise on a bicycle ergometer or treadmill decreases adverse effects. The increase in sympathetic activity during exercise increases blood pressure and decreases the risk of AV block and possibly bronchoconstriction. Furthermore, the blood supply to the gastrointestinal tract decreases and hence tracer uptake in that region, which is advantageous for the subsequent SPECT acquisition. In patients with left bundle branch block and pacemakers with ventricular stimulation, adenosine infusion should not be combined with physical exercise (see later section "Pitfalls").

Dipyridamole

Dipyridamole acts via the endogenous adenosine of the body by inhibiting the reuptake of adenosine and thereby raising the concentration around the receptor. Because of its long half-life, an antidote is needed, and theophylline is most commonly used. Adverse effects to dipyridamole are similar to those of adenosine.

Regadenoson

Regadenoson is a specific A2A receptor agonist recently made available with similar diagnostic results as adenosine.17 The effect of regadenoson is sustained, with a half-life of 2 to 3 minutes. Therefore, an antidote such as theophylline is often used. Because of its specific A2A receptor characteristics, regadenoson can be used in chronic obstructive pulmonary disease and asthma patients. Regadenoson was approved in the United States in 2008.

Dobutamine

Dobutamine is a sympathomimetic drug with positive inotropic and chronotropic properties that causes secondary vasodilatation. The infusion rate is increased in steps, starting at 5 or 10 μg/kg/min and increasing every third minute to 20, 30, and 40 μg/kg/min. Blood pressure and ECG are monitored continuously, and criteria for termination of infusion and contraindications are the same as for an exercise stress test. Injection of the radioactive tracer is performed at 85% of the age-predicted heart rate, which often occurs only after adding atropine starting at a dose of 0.25 mg. It is important that β-blockers are discontinued before the test in order to reach the age-predicted heart rate. Common adverse effects of dobutamine are headache, nausea, anxiety, and atypical chest pain. Angina and ventricular arrhythmias need to be monitored, and paroxysmal atrial fibrillation can be triggered by dobutamine. If the termination criteria are followed, severe complications such as myocardial infarction and death are extremely rare. When discontinuing the infusion, the effect of dobutamine lasts longer compared with adenosine, and sometimes an intravenous injection of a β2-blocker is needed. There are additional contraindications for the use of atropine (eg, narrow-angle glaucoma).

Medication and Caffeine before Myocardial Perfusion Imaging

Oral nitrates should be avoided before stress because of their anti-ischemic effect. If an exercise test or dobutamine infusion is used, β-blockers and heart rate–lowering calcium blockers should be discontinued 48 hours before the injection. A common indication for vasodilator perfusion imaging is inappropriate heart rate response in patients who cannot or who have forgotten to discontinue their medication. The amount of ischemia will be less if anti-ischemic medication is maintained even with vasodilator stress tests,4 however, with β-blockers, not to the extent that it affects the sensitivity and specificity.18 Caffeine needs to be discontinued before adenosine, dipyridamole, and regadenoson because of a competitive blockage of the adenosine receptor. Abstinence from caffeine is recommended for preferably 24 hours.

The characteristics of an ideal perfusion tracer would be uptake to the myocardium in direct proportion to blood flow; high extraction from blood to the myocardium during the first passage; medium half-life to enable time for acquisition but not so long that the radiation dose becomes high; radiation properties that suit the acquisition with regard to energy emission, resolution, and attenuation; and of course, a stable, cheap, and ready supply of the tracer. Unfortunately, there are no tracers today that meet all of these criteria, but three SPECT tracers fulfill several and are therefore used clinically. For PET imaging, three tracers are being used. The radiopharmaceuticals used are injected as a bolus injection in a peripheral vein.

SPECT Tracers

Thallium 201

In the case of thallium, the perfusion tracer and the radioactive substance are one and the same. Thallium 201 (201Tl) can be found in the same group as potassium in the periodic table and is distributed in the body as potassium in proportion to blood flow with a high extraction from the blood. The half-life is rather long, the radiation dose to the patient is high (~18 mSv), and the characteristics for SPECT acquisition are not ideal. After an injection during stress, imaging must be started within 10 minutes because of the so-called redistribution of thallium. Thallium does not stay fixed in the cell after uptake but reenters the blood stream and is continuously redistributed to the body. Therefore, a perfusion defect will disappear at rest because of uptake from thallium in the blood. This means that the same injected dose can be used for both stress and rest images, and the redistribution images at rest are performed 3 to 4 hours after injection. Thallium can also be used for viability testing in the case of a fixed perfusion defect. A late redistribution image the following day is used, and in the case of an infarct, the perfusion defect is still present. A fixed perfusion defect that disappears is an indication of viable myocardium.

99mTc-Labeled Tetrofosmin or Sestamibi

99mTc is a radioactive substance and must be linked to a perfusion tracer (eg, tetrofosmin or sestamibi). 99mTc has good characteristics for SPECT acquisitions and a medium to long half-life of 6 hours. Both perfusion tracers have lower extraction compared with thallium but thereafter remain stable in the cells binding to the mitochondria. This means that injection does not have to be adjacent to the SPECT camera, and the patient can be transported to the acquisition without haste. It also means that a separate resting injection must be performed if the stress image is abnormal. Tetrofosmin and sestamibi have a rather high gastrointestinal uptake, which may cause problems with image acquisition at rest or after pharmacologic stress (see "Pitfalls" later).

PET Tracers

All PET perfusion tracers allow for short imaging protocols and repeated studies due to their short half-lives. A PET perfusion study can thus be performed in a few minutes. Currently, three tracers are mainly used with PET.

13NH3 has high first-pass extraction, and uptake is linear over a wide range of myocardial blood flow except at very high flow rates. Imaging with 13NH3 requires either an on-site cyclotron or proximity to a regional positron radiopharmaceutical source center. Images are of high quality and resolution.

Oxygen 15–labeled water (H215O) is potentially superior to other tracers because it is metabolically inert and freely diffusible across capillary and cell membranes. However, a drawback is that the tracer is not accumulated in myocardium, but reaches equilibrium between tissue compartments. Thus, images of regional myocardial perfusion distribution are not readily produced by standard image visualization, and image processing for blood pool subtraction is needed; however, the image processing is simple process with current software.

82Rb is a potassium analog that has a first-pass extraction of 65%, and its uptake requires active transport via Na/K-ATPase, which is dependent on coronary flow. Also, with 82Rb, the extraction fraction decreases in a nonlinear manner with increasing blood flow, and this effect is more pronounced when compared with ammonia, although still superior when compared with 99mTc-labeled SPECT compounds. Image resolution and quality are somewhat compromised due to the high energy of positrons emitted during the decay of 82Rb and due to lower count rates as a result of the ultrashort half-life. An advantage of 82Rb over 13NH3 and H215O is that it is produced by a strontium 82/82Rb generator without the need for a costly cyclotron.19

18F-2-deoxy-2-fluoro-D-glucose (18F-FDG) is widely available due to its success as a metabolic imaging tracer in clinical oncology. It has been known for decades that the tracer is of high value to determine myocardial glucose utilization as an indicator of myocardial viability. Increased 18F-FDG uptake can be observed in ischemic tissue, whereas markedly reduced or absent uptake indicates scar formation. 18F-FDG uptake is heterogeneous in normal myocardium in the fasting state, and therefore, oral glucose loading, nicotinic acid derivatives, and infusion of insulin and glucose have been used to enhance myocardial 18F-FDG uptake.20

SPECT Protocols

Several different SPECT imaging protocols are being used depending on the local demands and circumstances.

201Tl Protocols

Imaging is performed directly after stress and after 3 to 4 hours at rest using the same isotope injection.

99mTc Protocols

One-day or 2-day protocols can be used. In the 1-day protocol, the first injection can be at rest or stress, and the injected dose is kept as low as possible. After image acquisition, the second dose is injected, which must be at least 2.5 to 3 times as high as the first dose to override the radioactivity already present in the myocardium. The advantages with a 1-day protocol are that the patients do not need to come back another day and the referring physician can get a quicker reply. The disadvantages are that the radiation dose will be somewhat higher and a perfusion defect can be underestimated if the second dose is too low to override the activity of the first dose. To keep the radiation dose as low as possible, stress imaging can be performed first, and if there is no perfusion abnormality, no rest imaging is needed.

Dual-Isotope Protocols

Resting images are performed with 201Tl, and stress images are performed with a 99mTc tracer. The advantage is the speed of the protocol; because of different energy peaks, both tracers can be injected shortly after each other without interference. The main disadvantage is that the comparison at rest and stress will be with two different tracers, with differences in, for example, attenuation.

PET Protocols and Data Analysis

The patient preparation for PET study is practically the same as for SPECT scans. Drinks containing caffeine need to be avoided during the preceding 12 hours because pharmacologic stressors are commonly used in PET imaging.

The comfortable positioning of the patient on the scanner bed is critical to prevent any motion artifacts during and between the scans. It is strongly recommended that hands are supported upright and not within the field of view. The perfusion imaging protocol depends on which tracer is used. With tracers such as 82Rb and H215O, the stress study can be performed practically without delay after the rest study. With 13NH3, stress testing is delayed for approximately 30 minutes to allow tracer decay. In all studies, a quality control process is needed to ensure optimal alignment of the CT attenuation and PET emission scans, and if necessary, misalignment needs to be corrected. If the system is capable of list mode acquisition, the data can be collected in ECG-gated mode that allows the simultaneous assessment of regional and global left ventricular wall motion from the same scan data. This is particularly practical when 82Rb is used as tracer. Because the stress images are acquired during stress (and not, as in SPECT, up to an hour after stress), the gated images in PET have higher probability to detect ischemia. The total time required for a whole study session depends on the tracer used. With 82Rb and H215O, the whole session can be finished in 20 to 30 minutes, and with 13NH3, the session can be finished in 60 to 80 minutes. The protocols can be further shortened significantly because only single stress perfusion imaging may be needed, especially when using quantification, and then the protocol with all tracers can be as short as 10 to 15 minutes.

The PET images can be displayed and analyzed using semiquantitative or visual assessment as with SPECT. The interpretation is also similar, including the gated images analysis. However, for full quantification, the dynamic image sets need to be processed using dedicated software that is able to perform image segmentation and tracer-specific kinetic modeling (Fig. 3–7). This process typically produces a parametric image in which the original radioactivity values are replaced by physiologic parameters such as perfusion in mL/g/min. Alternatively, numerical values with relative distribution within the left ventricle are reported along with standard image views.

Currently, all commercially available PET scanners are hybrid devices with multidetector CT in the same gantry as PET. Also with SPECT, combination with a CT is becoming popular. In addition, when using CT for attenuation correction (see sections on SPECT and PET technique earlier), many hybrid systems provide the ability to perform a full diagnostic CT coronary angiography, which can be readily combined with the perfusion or viability images from SPECT or PET.

The efficiency of hybrid imaging with PET/CT is very high because PET imaging protocols are short, allowing both CT angiography and perfusion imaging to be performed in a single session clearly below 30 minutes of total scan duration. The radiation dose from PET perfusion studies is very small; for example, the radiation dose from a single PET perfusion study is 0.8 mSv for H215O and 1 mSv for 3NH3. Recently, techniques that reduce patient dose in CT have been developed, and the doses have been reduced to as low as 1 to 7 mSv. Therefore, although the use of hybrid imaging causes an increased radiation dose for the patient compared with single imaging, currently comprehensive cardiac PET/CT imaging can be performed with a radiation dose of less than 10 mSv.21 The new detector technology for SPECT described earlier has decreased imaging time and lowered the dose needed for perfusion imaging. Using these technical developments, a hybrid coronary CT and perfusion stress only SPECT image can be obtained with radiation doses as low as 3 mSv.12

The rationale of using hybrid imaging is based on the possibility of simultaneous combination of anatomic and functional information. Because imaging procedures are a significant financial burden to the overall cost of health care, it is important to carefully consider and justify the use of all imaging tests. Coronary artery calcium (CAC) according to the Agatston score has been used a marker of plaque burden, and increased scores are associated with higher cardiovascular risk. There is some evidence that CAC may add incremental value to the perfusion information.22 For example, the presence of extensive CAC despite a normal relative perfusion distribution increased the sensitivity of detecting coronary stenosis without decreasing specificity, presumably by not missing balanced ischemia.23 CT coronary angiography is valuable for the evaluation of many subgroups of patients with known or suspected coronary artery disease (CAD), but it provides purely morphologic information that has well-known limitations. It has been shown that only approximately half of the lesions classified as significant in CT are linked with abnormal perfusion.24,25 The vasomotor tone and coronary collateral flow cannot be estimated because the percent diameter stenosis is only a weak descriptor of coronary resistance. However, myocardial perfusion imaging provides a simple and accurate integrated measure of the effect of all the parameters on coronary resistance and tissue perfusion, thereby optimizing selection of patients who may ultimately benefit from revascularization. However, SPECT and PET imaging do not provide morphologic information. The perfusion imaging only provides information about the existence and severity of perfusion abnormalities, but not about the mechanism; for example, patients with diffuse balanced coronary heart disease and patients with microvascular dysfunction leading to globally compromised perfusion cannot be separated by perfusion imaging.

The analysis of hybrid studies follows the same standard procedures that have been explained earlier. In addition, the analysis of CT angiography includes the standard processes and techniques used in stand-alone situations, such as visual assessment of original transaxial slices, multiplanar reconstructions, and utilization of the quantitative tools available. To make use of the true power of hybrid imaging, the fused images and data should also be used. In fused images, the individual coronary anatomy can be linked with perfusion information (Fig. 3–8).

Hybrid images provide a comprehensive view of the myocardium, the regional myocardial perfusion, and the coronary artery tree, thus eliminating uncertainties in the relationship of perfusion defects and the diseased coronary arteries. This may be particularly helpful in patients with multiple perfusion abnormalities. Combining anatomic information with perfusion also helps to identify and correctly register the subtle irregularities in myocardial perfusion. In patients with suspected CAD, hybrid imaging provided a sensitivity and specificity of 90% and 98%, respectively.26 Further studies are warranted to ensure the accuracy of hybrid imaging in different clinical populations. Furthermore, the incremental value of hybrid imaging over the stand-alone individual scan needs to be defined, and studies showing the potential effect on treatment, outcome, and cost effectiveness are needed.12

Primary Diagnosis of Ischemic Heart Disease

Myocardial perfusion imaging has a key role in the evaluation of patients with symptoms but no history of ischemic heart disease. The patient typically presents with nonacute chest pain or dyspnea of unknown origin (Figs. 3–9 and 3–10). Studies have shown that patients benefitting from myocardial perfusion imaging are those with intermediate risk of having ischemic heart disease (eg, an equivocal exercise stress test). Patients with low likelihood (eg, no risk factors and a negative exercise stress test) do not benefit from myocardial perfusion imaging. Patients with high likelihood of ischemic heart disease (eg, risk factors, typical chest pain at physical exertion, and a positive exercise stress test) should not undergo myocardial perfusion imaging but should be treated medically or invasively according to symptoms and risk.

Myocardial perfusion imaging can be the principal diagnostic test if the patient cannot perform an exercise stress test or if ST-segment changes in the resting ECG make the ST-segment response to exercise not interpretable. The former case applies to patients with orthopedic problems that prohibit maximum exercise, and the latter case applies to patients with left bundle branch block or ST-segment depression due to left ventricular hypertrophy. Patients with an equivocal exercise stress test include those who have either significant ST-segment depression or chest pain during exercise but not both. Also, patients who do not reach a sufficient end-exercise heart rate due to rate-limiting medical therapy that cannot be discontinued may undergo a pharmacologic myocardial perfusion imaging study to obtain a diagnosis. Patients with a high CAC score (Agatston >400) with or without chest pain are another group of patients for whom myocardial perfusion SPECT may be appropriate.

Patients with "Low Risk" Acute Chest Pain

Myocardial perfusion imaging can be used to rule out acute coronary syndrome in patients with acute chest pain of unclear origin in whom initial testing with ECG and blood tests was not conclusive. This includes patients with resting images only and injection during or shortly after the episode of acute chest pain (Fig. 3–11), as well as patients with stress and rest imaging admitted for suspected acute coronary syndrome. In this setting, only SPECT data are available.

Evaluation of the Physiologic Significance of a Coronary Stenosis

Stenoses of unclear significance are quite often found at coronary angiography both in patients with stable angina and patients with acute coronary syndrome. Myocardial perfusion imaging can be used to evaluate the flow-limiting effects of the stenosis and to guide therapy. If no or minor ischemia is found, the patient can be treated medically (Fig. 3–12). If a significant amount of the myocardium is jeopardized by ischemia, invasive treatment is indicated. The amount of ischemia at myocardial perfusion imaging correlates with the risk of major adverse cardiac events.

Assessment of Ischemia in Patients with Prior Revascularization

Instead of relying on a conventional exercise stress test, a noninvasive imaging study is recommended by guidelines to assess stress-induced ischemia in patients who have undergone revascularization either by percutaneous coronary intervention (PCI) or bypass surgery (Fig. 3–13). The rationale is that because there is already proven CAD, the likelihood of ischemia is higher, and the ST-segment response is often compromised due to previous myocardial infarction in this patient group. Also, myocardial perfusion imaging provides the location and extent of ischemia, which is important in guiding therapy. Another group includes patients who have had an episode of an acute coronary syndrome where coronary angiography has not been performed and where myocardial perfusion imaging is used for risk assessment.

Risk Assessment Before Major Surgery

Myocardial perfusion imaging can be used to evaluate ischemia in patients with high probability of or known heart disease planned for major surgery. An exercise stress test is often sufficient, but myocardial perfusion imaging can be used in high-risk cases (eg, diabetic and renal failure patients) and especially in cases of vascular surgery.

Viability Testing

Viability testing is important in patients with heart failure to detect the functionally compromised myocardium that is able to recover function by revascularization. With SPECT, an isotope uptake defect seen at rest on myocardial perfusion imaging can be caused by an infarction or decreased perfusion due to a severe stenosis causing resting ischemia. Revascularization is considered useful if a significant part of the dysfunctional myocardium is viable. Late redistribution imaging with thallium can be used, particularly in institutions where other techniques such as PET or magnetic resonance imaging (MRI) are not available (see "Viability Assessment" later). Low-resolution and bad images (too little activity left) at late redistribution hamper the thallium technique, and a reinjection of thallium is often needed. Technetium-labeled tracers can be used for viability testing, and preserved uptake in a myocardial segment implies viable myocytes. Oral nitrate is usually administered before the resting injection when using technetium-labeled tracers for viability testing to minimize resting ischemia. It is important to note, however, that inducible ischemia itself is an indicator of viability and a resting perfusion defect unchanged during stress in a patient with decreased left ventricular function indicates the need for specific viability testing.

For viability testing with PET, 18F-FDG is used to differentiate between viable and nonviable myocardium in a perfusion defect.27 18F-FDG is a glucose analog, and hibernating myocardium maintains metabolic activity; thus, 18F-FDG uptake is enhanced even when perfusion is decreased. The sign of viable hibernating myocardium on PET is typically detected as a decreased perfusion tracer uptake with a maintained 18F-FDG uptake.

What Determines Whether a Perfusion Imaging Test Is Useful?

Coronary angiography has been commonly used as reference standard when assessing diagnostic accuracy for noninvasive cardiac imaging tests. However, coronary angiography depicts coronary anatomy, and myocardial perfusion SPECT assesses myocardial perfusion, and the relationship between coronary stenosis and perfusion decrease is not straightforward. Early invasive coronary angiography studies showed that there is no correlation between the reactive hyperemic response and the degree of coronary stenosis.7 Therefore, one should bear in mind that sensitivity and specificity when using coronary anatomy as the reference method may not be the best markers for the accuracy of a perfusion assessment. Recent studies have demonstrated sensitivity in the range of 85% to 90% and specificity in the range of 80% to 90% for myocardial perfusion SPECT versus coronary angiography.15,28-33 Older studies showed lower specificity, but the use of attenuation correction, prone imaging, and gated SPECT has resulted in higher specificity, especially in obese and female patients.13,34 The sensitivity and specificity of PET have been reported as 93% and 92%, respectively.19

Research in noninvasive cardiology testing has moved toward evaluating the relationship between the findings of the test and the risk for a major adverse cardiovascular event (MACE)15 and cost-effectiveness analysis partly because of the problems of diagnostic accuracy mentioned earlier. There is a wealth of scientific evidence on the prognostic value and cost effectiveness of myocardial perfusion SPECT15,35-38 and, recently, of PET perfusion imaging.39,40

Amount of Ischemic Myocardium

There seems to be a nearly direct relationship between the amount of ischemic myocardium and the risk for a MACE,41 and when assessing the ischemic myocardium, one needs to take both the extent and the severity of the perfusion defect into account. The myocardium is divided into 17 segments according to the American Heart Association model,42 and the perfusion is scored for each segment. Zero points are given for no perfusion defect, 1 point is given for mildly reduced perfusion, 2 points are given for moderately reduced perfusion, 3 points are given for severely reduced perfusion, and 4 points are given for absent perfusion. The summed stress score (SSS) is the result of adding the scores for all segments at stress, and the summed rest score (SRS) is the corresponding result at rest. The summed difference score (SDS) is the difference between SSS and SRS. Note that only segments with an SSS can get an SRS; some irregularities in perfusion can be expected at rest in healthy subjects, but at stress, these may disappear. One of the reasons is that the perfusion map is normalized to the maximum perfusion of the heart at stress and rest. At rest, this means that a region with perfusion of 70% of maximum will be the result of a difference between approximately 0.7 and 1.0 mL/min/g or an absolute difference of 0.3 mL/min/g. At stress with a normal maximum perfusion of 3 mL/min/g, the same absolute difference of 0.3 mL/min/g would be 90% of maximum perfusion and would be read as normal. An SSS of 0 to 3 is interpreted as normal.43,44 With PET, the possibility of quantifying myocardial perfusion in absolute terms is possible and likely provides clinically useful information and makes the technique more accurate, especially in the assessment of severity of disease in patients with multivessel disease45 (Fig. 3–14). However, more clinical trials are warranted to elucidate the clinical value of quantification of perfusion.

Studies Showing Usefulness of Myocardial Perfusion Imaging in Different Clinical Scenarios

Stable Angina

In patients with stable angina, a normal myocardial perfusion imaging study corresponds to an annual risk for a MACE of 0.6% or, in other words, lower risk than the average population.41 With increasing SSS, the risk increases,46 and an SSS of 7 to 8 or more representing 10% of the myocardium is associated with an annual risk of MACE of 5%.47 Studies with myocardial perfusion imaging have showed that this patient population is a risk group in which medical treatment and risk factor management are associated with higher cardiac death rates compared with treatment with coronary revascularization.48 However, patients with less extensive ischemia benefit from medical management only compared with coronary revascularization.49 These results have been reproduced in a substudy of the Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation (COURAGE) trial,50 where increasing degree of ischemia was associated with increasing rates of MACE and patients with more severe ischemia seemed to experience the greatest benefit from PCI.15 Several studies in patients with stable angina have shown that using myocardial perfusion SPECT as the initial test before deciding on a coronary angiography saves money compared with an initial coronary angiography especially in patients with intermediate risk of CAD.51 More frequent coronary revascularizations are performed in patients with initial coronary angiography without a benefit in MACE.51 These findings have led to the use of myocardial perfusion SPECT and other noninvasive imaging tests as gatekeepers for coronary angiography in patients with intermediate risk for CAD. Patients with low likelihood after initial testing should not undergo myocardial perfusion imaging, and patients with high likelihood can be sent directly to invasive evaluation. Another indication for myocardial perfusion imaging in stable angina is patients who have undergone a coronary angiography in whom stenoses of unclear significance have been found.

Suspected Acute Coronary Syndrome

Myocardial SPECT has two roles in the setting of suspected acute coronary syndrome. It can be used in the emergency department in patients with intermediate risk of acute coronary syndrome with tracer injection during or shortly after chest pain.6,52-54 The negative predictive value in this patient population is very high (95%-99%), and therefore, patients with a normal scan may be discharged instead of admitted to the hospital. This has been shown to be cost effective.53,55 In patients who have been admitted to the hospital for suspected acute coronary syndrome with unclear diagnosis after repeated ECG and cardiac biomarkers, a stress test (exercise or pharmacologic) is recommended.1 Myocardial perfusion SPECT is useful in these patients, especially when abnormalities of the resting ECG hamper ST-segment analysis or when maximum heart rate cannot be reached.

After A Myocardial Infarction

International guidelines include myocardial perfusion imaging as an appropriate test in patients after myocardial infarction for risk assessment if coronary angiography has not been performed or if a coronary stenosis of unclear significance has been found on the primary PCI. However, less data are available in these patient populations on the prognostic value of myocardial perfusion imaging. A systematic review found four observational studies with a total of 2106 patients and concluded that there was independent and incremental prognostic value of myocardial perfusion SPECT after myocardial infarction.56 More recently, a multinational study showed that early myocardial perfusion SPECT can identify patients who can be discharged early after myocardial infarction.57

Overreporting

One of the most common mistakes in myocardial perfusion imaging is overreporting minor defects as ischemia. Small irregularities in the perfusion images are often seen and are not associated with a worse prognosis. Rather, overreporting may lead to unnecessary coronary catheterizations without any benefit to the patient. The parametric maps of perfusion are helpful in locating the ischemia but cannot be the basis for ischemia detection because small areas of irregularities tend to be overestimated on these images. Therefore, the full set of short and long axis images should be evaluated before looking at the parametric images (see Fig. 3–3B). This problem can also be solved by absolute quantification when PET is used (see Fig. 3–14).

Inadequate Stress Test

The perfusion distribution during stress is visualized and compared with rest when assessing ischemia with myocardial perfusion imaging. A stenosis will cause an inability to increase perfusion in the perfusion territory of the corresponding coronary artery during stress, and therefore, adequate stress must be present during injection of the perfusion tracer. In the case of an inadequate stress test, there is the risk of a false-negative test. If an exercise stress test is performed, the limit of at least 85% of predicted maximum heart rate is used. For example, patients on β-blockers often do not reach this level, and in these cases, the perfusion tracer is not injected; instead, a pharmacologic stress test should be performed. In the case of adenosine, dipyridamole, and regadenoson, caffeine abstinence is important before the test because of the competitive inhibition of the binding to the adenosine receptor by caffeine. One recent study used an increased dose of 210 μg/kg/min in patients with recent caffeine intake and showed increased detection of ischemia compared with 140 μg/kg/min and recent caffeine intake.58 Many patients with chest pain have prescriptions for oral nitrates, and it is important that patients have a 12-hour cessation of long-acting nitrates and at least a 1-hour cessation of short-acting nitrates before the stress injection. Resting injection should likewise be a true rest, and checking the patient's heart rate before injection is an easy way to avoid "false" resting images. Patients exhibiting chest discomfort are often administered short-acting nitrates before resting injection to ensure a true resting perfusion image.

Left Bundle Branch Block

Special consideration must be paid to myocardial perfusion SPECT for patients with left bundle branch block (LBBB) on the ECG as well as patients with ventricular stimulation pacemakers causing the ECG to have the appearance of an LBBB. The problem with LBBB in myocardial perfusion SPECT is that many patients seem to have a perfusion defect in the septal half of the heart. The cause of this apparent perfusion deficit is not entirely known and is not seen in all patients with LBBB. What is known empirically is that a stress test where the heart rate increases often causes the perfusion defect to worsen, although no stenosis is found on a subsequent coronary angiography. This is demonstrated in Fig. 3–15 showing a patient with stress-induced LBBB. The stress image shows a perfusion defect in the entire septum but most prominent in the anteroseptal region. At rest, the ECG returns to normal, and the SPECT shows a normal perfusion distribution. Therefore, a pharmacologic vasodilator stress test (eg, adenosine, dipyridamole, or regadenoson) without exercise is used in cases of LBBB to avoid a heart rate increase.

Triple-Vessel Disease

Detection of ischemia on standard analysis myocardial perfusion imaging depends on one region having normal perfusion; this region is set to 100%, and the rest of the myocardium is normalized to this normal region. In the case of multivessel disease with physiologically significant stenoses in all coronary vessels, the perfusion will be globally decreased, the region with a 100% signal will be falsely read as normal, and occasionally, so-called balanced triple-vessel disease is missed in standard relative perfusion analysis. There are also cases where the perfusion SPECT shows ischemia in one territory but on coronary angiography multivessel disease is found. In these cases, SPECT has detected only the tip of the ischemia iceberg. Therefore, it is important to look for other signs of multivessel coronary disease (eg, the blood pressure reaction and left ventricular volumes and function).59,60 During exercise stress test, the blood pressure reaction is often pathologic, with a decrease in the blood pressure during the final stages of exercise. A dilatation in the stress perfusion images compared with the rest images (ie, a TID) is also a sign of a pathologic test (Figs. 3–16 and 3–17). This is caused by postischemic stunning where the cardiac contractile function still is depressed after the ischemic event. In these cases, gated SPECT also shows dilated ventricular volumes and decreased ejection fraction after stress compared with rest.

Most of these problems can be avoided by using absolute quantification of myocardial perfusion (see Fig. 3–14).45 Unfortunately, SPECT does not offer absolute quantification of tissue perfusion, and most clinical routine use of PET imaging is still based on relative analysis of perfusion distribution. Even without quantification, PET has been considered better in complex revascularized patients and obese patients (Fig. 3–18). However, in most patients with bypass surgery, myocardial perfusion SPECT can still be of use (Fig. 3–19).

High Gastrointestinal Tracer Activity

Perfusion tracer activity in the gastrointestinal tract adjacent to the heart may cause a false-positive test if present during rest and a false-negative test if present during stress. High uptake is almost always close to the inferior wall and is seldom a problem during exercise. The image acquisition can be repeated in the case of high uptake of perfusion tracer in the regions adjacent to the heart if this leads to problems in interpretation (see Fig. 3–13).

There are three roles of myocardial perfusion SPECT and PET that can be compared with other available imaging modalities; these are ischemia detection, assessment of viability/hibernation, and assessment of left ventricular function.

Ischemia Detection

Myocardial perfusion SPECT is by far the most commonly used cardiac imaging technique to detect ischemia. The use of PET is growing. Other common cardiac imaging techniques used are stress echocardiography and cardiac MRI. Stress echocardiography detects regional function during exercise or dobutamine as an indicator of ischemia. Several studies have shown that stress echocardiography myocardial perfusion SPECT have similar diagnostic and prognostic performance. Stress echocardiography, however is more operator dependent than myocardial perfusion SPECT. Perfusion contrast echocardiography has also been recently used for detection of CAD, but its role has not yet been established. Cardiac MRI in patients with suspected CAD is based on either wall motion analysis or first-pass perfusion detection. The accuracy of cardiac MRI is in a similar range as other noninvasive imaging tests in skillful hands. Currently, the availability of MRI scanners for cardiac use is limited. In addition, less evidence about the prognostic value and cost effectiveness of the newer techniques exists.

Viability Assessment

Multiple imaging techniques have been developed to assess viable and nonviable myocardium by evaluating perfusion, cell membrane integrity, mitochondrial function, glucose metabolism, scar tissue, and contractile reserve. PET, SPECT, and dobutamine stress echocardiography have been extensively evaluated for assessment of viability and prediction of clinical outcome after coronary revascularization. Low-dose dobutamine echocardiography has been successfully used, and an increase in contraction from rest to 5 to 10 μg/kg/min dobutamine is interpreted as viable myocardium. On MRI, late gadolinium enhanced imaging 10 to 20 minutes after contrast injection can be used to visualize the infarct.61

In general, nuclear imaging techniques have a high sensitivity for the detection of viability, whereas techniques evaluating contractile reserve have somewhat lower sensitivity but higher specificity. MRI has a high diagnostic accuracy for assessment of the transmural extent of myocardial scar tissue. With this technique, viability detection is based on indirect estimation of nonscarred myocardium. The fraction of the scar is inversely related to recovery of wall motion after revascularization, but there is no sharp cutoff for accurate prediction of wall motion recovery after revascularization. Because this technique cannot characterize the viable tissue adjacent to scar, low-dose dobutamine infusions have been used with cardiac MRI to improve the accuracy. Thus, the performance of cardiac MRI in detection of viability and predicting recovery of wall motion is similar to other imaging techniques.62 Of the imaging techniques, 18F-FDG PET is very sensitive; this technique has been used for a long time for viability detection and is still the most documented technique. The evidence is currently mostly based on observational studies or meta-analyses, and there are only two randomized multicenter studies on viability, both of which used PET imaging.63,64 Patients with a substantial amount of dysfunctional but viable myocardium are likely to benefit from coronary revascularization and may show improvements in regional and global contractile function, symptoms, exercise capacity, and long-term prognosis. In summary, the differences in performance of the various imaging techniques are small, and commonly, the experience and availability determine which technique is used.

Left Ventricular Function

Gated SPECT and gated PET provide automated, mainly user-independent, three-dimensional assessment of left ventricular function. However, the method of choice for left ventricular function is still echocardiography because of its lack of radiation and concomitant assessment of valvular function, which cannot be done by the nuclear imaging methods. Gated SPECT has been extensively validated against other methods, for example MRI.65