Richard C. Brunken
Nuclear cardiac imaging contributes significantly to the care of patients with known or suspected heart disease. Recent innovations in instrumentation and the use of advanced image reconstruction techniques have permitted reductions in patient radiation doses and shorter imaging times, further enhancing the application of nuclear imaging techniques in the clinical arena. Measurements of absolute myocardial perfusion (in milliliter of blood flow per minute per gram tissue) and perfusion reserve obtained with positron emission tomographic (PET) imaging, once feasible only in research labs, can now be derived from clinical imaging studies using commercially available software. The information provided by noninvasive nuclear imaging studies is valuable for the detection and localization of coronary artery disease, risk stratification, and assessment of myocardial viability. This chapter describes the fundamentals of nuclear imaging and relates them to their specific clinical application.
OVERVIEW OF NUCLEAR IMAGING
A conventional nuclear imaging study employs a gamma camera to record the naturally occurring emissions of a radioactive tracer that is selected to visualize a physiologic process occurring in vivo. The ideal radioactive tracer permits visualization of the physiologic process of interest without disturbing it. The tracer must be nontoxic, emit energies appropriate for imaging, have a high target organ to background ratio, and impart as low a patient radiation dose as possible. It should be readily available and inexpensive to be practical for clinical use. Most radioactive tracers used for cardiac imaging are administered intravenously, but some used for PET imaging (below) may be administered as inhaled gases.
Radioactive tracers spontaneously decay by emitting packets of electromagnetic energy, or photons. The emitted photons are captured by the gamma camera and used to create an image of the physiologic process of interest. However, some of the emitted photons will be attenuated (absorbed) before they leave the body and will not be used for image creation. The probability that a photon will be attenuated increases as the path length to exit the body increases.
A typical gamma camera has a lead collimator, a piece of metal with a “honeycomb” of openings that serves to focus the incident photons (Fig. 12.1). Most collimators used in cardiac imaging have holes that are parallel to each other (parallel-hole collimators). Photons traveling perpendicular to the camera pass through the holes and are used for image creation. Photons approaching the camera at an angle are stopped by the septa and are not used for imaging. Collimator performance depends on the length and diameter of its holes, as well as septal thickness. In general, spatial resolution improves as the ratio of the diameter of the hole to its length decreases. As spatial resolution improves, efficiency (the relative number of incident photons traversing the lead baffle) decreases. Collimator selection therefore reflects a trade-off between spatial resolution and count sensitivity. High-resolution collimators provide better spatial resolution but lower sensitivity than other collimators, because a larger proportion of photons are absorbed and do not reach the crystal. “General-purpose” or high-sensitivity collimators are capable of passing greater numbers of photons per unit of time, with somewhat poorer spatial resolution. “All-purpose” collimators have characteristics that are intermediate between these extremes.
FIGURE 12.1 Schematic representation of a conventional gamma camera. Photons emitted by decay of the radiotracer of interest in the target organ can travel directly to the camera, or be scattered by interposed tissue. Photons that strike the lead collimator are absorbed and do not reach the crystal. The photons that pass through the collimator strike the crystal and interact with it, producing flashes of visible light. The photo-multiplier tube closest to the gamma-ray interaction receives the greatest amount of light, and comparison of the relative response from multiple photomultiplier tubes allows localization of the scintillation event. Light quanta strike the photo-cathodes of the photomultiplier tubes, causing the release of electrons. The electrical signal is amplified by the photomultiplier tubes and transmitted to the energy discrimination and positioning circuitry of the camera.
In a conventional gamma camera, photons exiting the collimator strike a sodium iodide crystal containing a trace amount of nonradioactive thallium. The crystal stops most of the photons and converts their energy into quanta of visible light. Photons that are more energetic produce a greater number of light quanta than less energetic photons. A “light pipe” directs the light quanta to adjacent photomultiplier tubes. The photomultiplier tubes convert the energy of the light quanta into an electrical current and then amplify the strength of the current by several thousandfold. Special electronic circuitry is used to locate the xand y positions of the initial light impulse and to analyze the energy of the incident photons.
Some photons originating in the heart may be deflected from their initial path, or scattered, before entering the camera. Photons that are scattered in tissue generally lose energy when they are deflected. Scattered photons degrade the image because they do not travel in a direct path to reach the camera. Some scattered photons are deflected toward the camera at an oblique angle and are prevented by the collimator from reaching the crystal. However, other scattered photons do traverse the collimator and strike the crystal with an energy that is less than that of the photons that are not scattered. The contribution of scattered counts to the image can be minimized by setting an optimal “energy window” on the camera. Photons that strike the crystal with energies that are outside of the “energy window" are rejected, thereby improving image quality.
A nuclear image gradually emerges as increasing numbers of photons (scintillation events) are recorded by the camera. Image quality improves nonlinearly as the number of counts (scintillation events) in the picture increases. Thus, it is important to use a high enough radiotracer dose and/or to image for a long enough period of time to acquire a sufficient number of counts to achieve satisfactory image quality. State-of-the-art gamma cameras are capable of imaging about 200,000 counts per second, and some multiwire cameras are capable of even higher count rates. A camera witn a high count rate capability is needed for high-quality first-pass imaging (imaging performed as the tracer is administered intravenously).
Nuclear images can be obtained using planar or single photon emission computed tomographic (SPECT) acquisition techniques. Planar images are obtained with the gamma camera positioned in a fixed location. Usually, planar images are acquired in anterior, left anterior oblique, and left lateral projections. On occasion, planar images may also be obtained from a right anterior oblique projection. Planar images may be acquired for a specific period of time, for a specific number of counts, or for a given number of cardiac cycles (radionuclide ventriculography).
CARDIAC SPECT IMAGING
In SPECT imaging, pictures are obtained from multiple angles (projections) about the body. Conventional SPECT devices employ between one and four camera heads that rotate about the patient. Images are acquired over 180- or 360-degree arcs, usually by “stepping and shooting” an image every 3 or 6 degrees. Information from the raw projection images is then reconstructed using either filtered back-projectionor iterative reconstruction techniques. The reconstructed images are used to define the three-dimensional (3-D) distribution of the tracer in space. Image sets orthogonal to the long axis of the heart (vertical long axis, horizontal long axis, and short axis) are then generated for review.
SPECT images can be acquired with electrocardiogram (ECG) gating to permit assessment of ventricular function. The R–R interval of the ECG is usually divided into 8 or 16 equal time bins (frames) prior to study acquisition (Fig. 12.2). Each bin depicts a different portion of the cardiac cycle. At each step, the SPECT camera acquires 8 to 10 ECG gated images over multiple cardiac cycles. Gated image sets are reconstructed in the same cardiac axes as the nongated images, providing functional information for each of the “slices” in the image sets. Current software packages can also display the gated information as a 3-D rendering, permitting the study interpreter to visualize a graphical representation of the beating heart. Gated SPECT images provide information about regional wall motion and systolic thickening, as well as global left ventricular size and systolic function.
FIGURE 12.2 Graphic representation of the process of ECG gating. Using the R wave of the patient’s ECG as a timing signal, image count data from multiple cardiac cycles are added to the time bin corresponding to the portion of the R–R interval in which they were acquired. Counts in each bin increase with each succeeding cardiac cycle, improving the quality of the image represented by that time bin. Usually, 8 or 16 images are acquired per R–R interval, each representing a time-averaged look at that portion of the cardiac cycle. Once the images are acquired, they can be displayed in a continuous ciné loop, depicting global and regional ventricular function.
Several advances in instrumentation have served to enhance the quality of SPECT nuclear cardiac studies. Correction for photon attenuation can be performed using a density map of the chest (a set of transmission images) obtained with either an external line source of activity or a low dose x-ray CT scan. The density map is used to correct for the loss of counts resulting from attenuation of myocardial activity by interposed tissue, reducing the likelihood that an attenuation artifact will be mistaken for a true perfusion defect. Solid-state cameras dedicated to cardiac imaging employ detectors made of cadmium–zinc–telluride (CZT) or cesium iodide (CSI) instead of conventional sodium iodide (Tl) crystals to detect emitted photons. Some solid state cameras employ L-shaped detector heads that are large enough to fit around the chest. This permits imaging of the heart without moving the camera. Semiconductors in the solid-state detectors of these cameras directly convert the energy of incident photons into an electrical current. These instruments provide higher count sensitivity, improved energy resolution, and better spatial resolution than conventional gamma cameras. In some camera systems, the patient is imaged while seated in an upright or semiupright position. This improves patient comfort during imaging and helps to reduce movement artifacts. Accompanying the advances in instrumentation is the increasing application of iterative techniques for image reconstruction. Iterative reconstruction methods help to reduce noise and improve image quality in low-count studies. These developments have served to maintain image quality while permitting the use of shorter imaging times and/or lower radiopharmaceutical doses.
CARDIAC PET IMAGING
PET cardiac perfusion imaging is more sensitive and specific for the detection of coronary artery disease than conventional SPECT imaging. When PET imaging is performed with the metabolic tracer 18F-2-fluoro-2-deoxyglucose (FDG), it provides important information about regional myocardial viability. Tracers used for PET imaging decay by ejecting a positron (a β+ particle) from a proton-rich nucleus. Once ejected from the nucleus, the positron interacts with atoms in the surrounding medium, producing excitations and ionizations that slow its travel. As it slows, the positron eventually comes into close proximity with an electron in the surrounding medium. The electron and the positron are “anti-particles" and they mutually annihilate, liberating energy in the form of two 511-kiloelectron volt (keV) annihilation photons, which exit the annihilation site in opposite directions.
PET cameras employ circular banks of gamma-ray detectors to identify paired scintillation events occurring about 180 degrees apart (annihilation coincidence detection). If two detectors opposite each other simultaneously register photon hits in coincidence, then the annihilation event is localized to the volume between the two detectors (Fig. 12.3). If only a single photon is detected, the requirement for coincidence detection is not satisfied and the scintillation event is not used for image creation. Data from millions of annihilation events are reconstructed using standard filtered back-projection or iterative reconstruction techniques to create an image.
FIGURE 12.3 Schematic representation of PET imaging. When a PET radiotracer decays, it emits a positron, the antiparticle of the electron. The positron travels a short distance before interacting with an electron, resulting in their mutual annihilation. The annihilation event produces two 511-keV photons traveling in diametrically opposed directions. PET imaging relies on the detection of two photons “in coincidence” (arriving at opposite detectors at about the same time) to identify true annihilation events and localize them to the volume between the two detectors.
When a PET scanner operates in the two-dimensional (2-D) mode, each ring of detectors is separated from adjacent rings by lead or tungsten septa. An image is generated in the plane of each ring by using coincidence events occurring between detectors in that ring. A detector in one plane can also be in the “line of sight” and in coincidence with some of the detectors in the next higher or lower adjacent ring. By using interplane coincidences, an additional set of interpolated images is generated midway between the direct planes defined by the detector rings. The 3-D spatial distribution of the tracer is achieved by “stacking” images from multiple 2-D planes. When a PET camera operates in the 3-D mode, it operates without septa between the detector rings. Coincidences are identified between detectors lying in any ring combination. 3-D image reconstruction techniques are then used to define the tracer distribution over the entire volume that is imaged. The reconstruction techniques employed for a 3-D camera are more computationally demanding than those used for a 2-D camera. However, a PET camera operating in the 3-D mode is more sensitive for detecting annihilation events. This means that comparable image quality can be achieved using lower doses of radioactive tracers than those used for 2-D imaging.
PET images can be acquired at the moment of peak myocardial uptake of the tracer (static image acquisition), or a series of images can rapidly be acquired over time (dynamic list mode image acquisition). The images first acquired, the transverse images, are orthogonal to the body. Transverse images are reoriented into standard short-axis, and vertical and horizontal long-axis image sets analogous to those used in SPECT imaging. ECG-gated PET images can be obtained to assess segmental wall motion and thickening, ventricular volumes and ejection fractions.
PET has several advantages relative to SPECT for cardiac imaging, including better temporal and spatial image resolution and more accurate correction for attenuation of emitted photons. Current PET instruments can acquire multiple cross-sectional images of the entire heart as rapidly as every 5 seconds. PET cameras have spatial resolutions on the order of 6 to 8 mm full-width, half-maximum (FWHM) in the center of the field of view, as compared to 12 to 15 mm FWHM for conventional SPECT cameras. In addition, the radiation dose imparted to the patient from rubidium-82 or nitrogen-13 ammonia is less than half that of a conventional rest-stress SPECT perfusion study performed with a technetium-99m labeled tracer.
Cardiac PET images are corrected for attenuation using transmission images depicting the density of the thorax. In conventional PET systems, transmission images are acquired using an external ring source of activity. In newer PET/CT systems, which are the instruments used most frequently in clinical practice for imaging, transmission images are obtained using a low dose x-ray CT of the thorax. The transmission images are acquired immediately before or after the nuclear images. If desired, the CT portion of the camera can be used for coronary CT angiography and/or calcium scoring at the same session.
On a PET camera that has been appropriately calibrated, counts on the images accurately reflect true tissue activity concentrations. Rates of myocardial perfusion and/or metabolism can be measured by examining the change in tissue tracer concentrations over time on dynamic PET images. Because the acquisition time of each image frame is operator-defined, it is possible to generate time-activity curvesdepicting changing tissue and vascular tracer activity concentrations over time. Cardiac count data from the time-activity curves are then fit using a mathematical model that describes the biologic behavior of the radioactive label. The parameter of interest (e.g., myocardial blood flow in milliliters per minute per gram of tissue or glucose consumption in micromoles per minute per gram of tissue) is then determined by the equation best fitting the patient’s myocardial time-activity data. Commercially available software now permits noninvasive measurements of rest and hyperemic perfusion and myocardial perfusion reserves from dynamic rubidium-82 PET perfusion images.
INTERPRETATION OF NUCLEAR IMAGES
Myocardial perfusion images should be interpreted by an experienced physician. Care is taken to identify significant extracardiac activity that might impact image interpretation, to define cardiac chamber size, assess relative myocardial perfusion, and, where appropriate, examine regional and global myocardial thickening and wall motion. As a quality-control measure for the SPECT studies, the raw projection images are reviewed in a cinematic display. This enables the physician to look for motion and displacement of the heart during imaging (which may create image artifacts), to identify interposed tissue that might attenuate the images, and to determine if the extracardiac distribution of the radioactive tracer is normal.
Once the projection images have been examined, the short axis, vertical long axis, and horizontal long axis images are reviewed (Fig. 12.4). The standard 17-segment left ventricular model is typically used for study interpretation. In this model each segment represents a near-equal proportion of the ventricular mass (0% for each segment except for the apical cap, which constitutes 4% of left ventricular mass) (Fig. 12.5).
FIGURE 12.4 Cardiac plane definition. (From the Committee on Advanced Cardiac Imaging and Technology, Council on Clinical Cardiology, American Heart Association; Cardiovascular Imaging Committee, American College of Cardiology; and Board of Directors, Cardiovascular Council, Society of Nuclear Medicine. Standardization of cardiac tomographic imaging. Circulation. 1992;86:338–339, with permission from Wolters Kluwer Health.)
FIGURE 12.5 Regional wall segments. This diagram demonstrates how the left ventricle can be divided into standardized segments for cardiac imaging. Short-axis, horizontal long-axis, and vertical long- axis views are depicted. (From American Heart Association Writing Group on Myocardial Segmentation and Registration for Cardiac Imaging. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation.2002;105:539–542, with permission from Wolters Kluwer Health.)
Segmental perfusion and function are scored based on a visual analysis of the images. One system assigns perfusion scores from 0 through 4; in this system, normal tracer uptake is scored as 0, a mild (equivocal) reduction in activity is scored as 1, a moderate definitive reduction in activity is given a score of 2, a severe reduction in activity is considered a 3, and a complete absence of activity is given a 4. Summing the scores generates an index that incorporates both the severity and the anatomic extent of a perfusion abnormality. For example, if the stress images demonstrate a perfusion defect involving three segments, with visual scores of 4, 3, and 2, the summed stress scoreis 9 (= [1 × 4] + [1 × 3] + [1 × 2]). Defects on the rest images are scored in similar fashion and added together to calculate the summed rest score. Rest and stress summed scores may be considered of low (0 to 4), intermediate (5 to 8) or high (>8) severity. The summed difference score is the sum of the differences between the stress and rest scores for each segment. The degree (or amount) of defect reversibility is considered low (summed difference score of 0 to 2), intermediate (summed difference score of 3 to 7) or high (summed difference score > 7).
Computer-generated analyses of the images are also used to aid in study interpretation. Relative tracer concentrations on the patient’s images are compared to those of a database of gender-matched normal subjects, permitting rapid identification of image voxels with abnormal tracer concentrations. This analysis of the relative 3-D tracer distribution in the heart is usually displayed as a 2-D “polar map." In the polar mapping technique, the ventricle is considered to be composed of a group of short-axis slices. The apical short-axis slice is smallest in diameter, and as the slices get closer to the base of the heart the diameter of the slices gets larger. Imagine that the smallest short-axis slice is put inside the center of the next larger slice. Both slices are oriented as if the observer is looking up the ventricle toward the base: the septum is on the observer’s left, the anterior wall is above, the lateral wall is to the observer’s right, and the inferior wall is below (Fig. 12.6). If those two slices are put inside the center of the third slice, and these three slices are put inside the center of the fourth slice, and so on, the process can be continued until all of the short-axis slices are stacked inside the largest basal short-axis slice, each having the same orientation. Basically, the image slices form a “target” with concentric rings and the 3-D information has been mapped into a 2-D display. The parameter of interest (e.g., relative technetium-99m sestamibi activity) in each short-axis slice is expressed as a percentage of the maximal value over the entire heart ("normalized" to peak myocardial values). The normalized activity in each picture element (voxel) can then be displayed using either a gray or color scale to provide the viewer a map of the activity distribution over the entire left ventricle in a single image. Normalized patient data are referenced to gender-matched databases, and a second polar map is generated that displays the extent and the severity of the patient’s abnormalities (e.g., two or three standard deviations from normal) on a voxel-by-voxel basis. A variety of parameters can be displayed using the polar mapping technique, including relative and absolute myocardial perfusion, relative systolic thickening (systolic change in counts on gated perfusion or metabolic images), segmental wall motion, thallium-201 redistribution, and PET perfusion-metabolism mismatches. These measurements provide an “independent observer” that assists in the interpretation of the images.
FIGURE 12.6 Left ventricular segmentation. A polar plot demonstrating the 17 myocardial segments and the recommended nomenclature. (From Port SC. Imaging guidelines for nuclear cardiology procedures, part II. J Nucl Cardiol 1999;6:G48–84, with permission of the American Society of Nuclear Cardiology.)
MYOCARDIAL PERFUSION IMAGING
Perfusion imaging is used to show relative regional myocardial blood flow during physiologic states of interest, usually during stress (exercise, pharmacologic) and at rest. The ideal perfusion tracer localizes to the myocardium in direct proportion to blood flow, such that tissue counts increase linearly as blood flow increases. In practice, only two perfusion tracers come close to exhibiting this linear relationship: technetium-99m teboroxime (SPECT imaging) and oxygen-15 water (PET imaging). The other tracers used for clinical SPECT and PET perfusion imaging exhibit progressively smaller increases in tissue uptake as blood flows increase above about 2 mL/min/g tissue.
A summary of the various myocardial perfusion imaging protocols is given in Figure 12.7. Recent clinical studies indicate that use of a “stress only” image acquisition might be utilized in appropriately selected individuals (patients in whom the likelihood of a stress defect is considered low by their clinical presentation). Use of a “stress only" image acquisition protocol in appropriately selected patients reduces the time required for testing and patient radiation exposure. In individuals in whom exercise stress is not feasible, pharmacologic stress perfusion imaging is an acceptable alternative. As the severity of a coronary stenosis becomes more pronounced, the vessel loses its ability to increase blood flow in parallel with increases in tissue oxygen demand. In vascular territories supplied by a diseased artery, coronary flow reserve is impaired and a defect will be observed if images depicting myocardial perfusion during stress (exercise, hyperemia) are obtained. Unless the coronary stenosis is very severe, diseased vessels usually have a sufficient reserve to sustain blood flow under resting conditions, and perfusion images obtained in the basal state will not demonstrate a defect. A reversible defect (Table 12.1) is one that is present on stress images and not present on rest or redistribution images; it is the hallmark of stress-induced ischemia. Sometimes a coronary stenosis is severe enough to result in a perfusion defect if imaging is performed in a resting state. If a rest defect improves on redistribution or reinjection images, the segment is viable and likely to benefit from revascularization.
FIGURE 12.7 Schematic summary of various myocardial perfusion imaging protocols. (Adapted from DePuey EG. Updated Imaging Guidelines for Nuclear Cardiology Procedures, Part 1. J Nucl Cardiol. 2001;8:G1–G58. Additional in-depth information regarding myocardial perfusion imaging protocols can be found at www.asnc.org under Guidelines and Standards. Stress protocols and tracers.)
TABLE
12.1 Defect interpretation
A perfusion defect that persists on both stress and rest or redistribution/reinjection images, a fixed defect, may reflect myocardial scar or myocardial hibernation. Partially reversible defects are stress defects that improve but do not normalize completely on rest or redistribution/reinjection images; these likely reflect nontransmural scar with superimposable ischemia. A defect is said to exhibit reverse redistributionif it is present on rest or redistribution images and is absent or much less prominent on stress images. Reverse redistribution has been identified in patients with multivessel coronary artery disease and in patients with acute myocardial infarction; it may reflect a differential washout of the perfusion tracer. Reverse redistribution may also reflect a technical artifact (oversubtraction of background activity on the rest/redistribution images).
The specific coronary artery affected by a stenosis is inferred by the anatomic location of a perfusion defect on the images. Typically, the left anterior descending (LAD) artery supplies the anterior wall, anterior septum, and apex. When it is dominant, the right coronary artery (RCA) usually supplies the inferior wall and the basal inferoseptum. The circumflex artery typically supplies the lateral wall when it is nondominant, and will additionally supply the inferior wall and basal inferoseptum when it is dominant (Fig. 12.8). Individual patient variations in the distribution of the coronary arteries do exist and may affect the patterns of myocardial perfusion identified on the nuclear images. In addition, observed perfusion patterns may be affected by the presence and adequacy of coronary collaterals.
FIGURE 12.8 Coronary artery territories. Assignment of the 17 myocardial segments to the territories of the left anterior descending (LAD), right coronary artery (RCA), and left circumflex coronary artery (LCX). (From Port SC. Imaging guidelines for nuclear cardiology procedures, part II. J Nucl Cardiol1999;6:G48–84, with permission of the American Society of Nuclear Cardiology.)
Myocardial perfusion imaging is utilized in clinical practice to identify coronary artery disease and to ascertain the physiologic significance of lesions of uncertain severity. It is used to stratify risk in patients following acute myocardial infarction and in patients with chronic coronary artery disease. The assessment of ventricular function from gated SPECT imaging provides incremental prognostic information beyond that provided by the pattern of myocardial perfusion alone. Myocardial perfusion imaging is also useful for risk stratification in patients prior to noncardiac surgical procedures, and in the follow-up of symptomatic patients with prior percutaneous transluminal coronary angio-plasty (PTCA) or coronary artery bypass grafts (CABG). It may also be useful for the differentiation of ischemic from nonischemic cardiomyopathies. In patients with ischemic cardiomyopathy, some have advocated late redistribution or reinjection imaging with thallium-201 to distinguish viable myocardium from scar.
PET METABOLIC IMAGING FOR MYOCARDIAL VIABILITY
In patients with ischemic heart disease, the presence of tissue glucose metabolism in hypoperfused ventricular segments on PET metabolic imaging with FDG is a reliable marker of clinically important myocardial viability. This is manifest clinically by an improvement in regional contractile function in metabolically active myocardial segments following interventional restoration of blood flow. Ultimately, improvement in left ventricular ejection fraction (LVEF) and functional capacity are related to the anatomic extent and severity of the mismatch between perfusion and glucose metabolism on PET images obtained prior to coronary revascularization. Prior clinical studies indicate that individuals with the most extensive perfusion-metabolism mismatches derive the greatest functional benefit from revascularization and are most likely to exhibit an increase in global LVEF after the procedure.
RADIONUCLIDE VENTRICULOGRAPHY
Two types of nuclear imaging are used to assess ventricular function. In first-pass imaging, high-temporal-resolution sequential (or list mode) images of the central circulation are obtained as the radioactive tracer is administered intravenously. A camera with a high count rate capability and a “tight” bolus of the radioactive tracer are required to achieve a high-quality study. First-pass images are usually acquired from either anterior or left anterior oblique projections. The images derived from a first-pass study depict the movement of the radioactivity through the heart’s chambers with a sufficient temporal resolution to permit measurement of ventricular ejection fractions on a beat-to-beat basis. Measurements from 5 to 15 cardiac cycles are usually averaged to calculate the ventricular ejection fraction. The formula used to calculate the LVEF is:
End-diastolic and end-systolic counts represent background-corrected counts in the left ventricle at end-diastole and endsystole, respectively.
First-pass studies can be used to assess the ventricular response to exercise stress. A tracer such as technetium-99m DTPA is employed, because the renal clearance of this radiopharmaceutical from the vascular space is rapid enough to permit sequential injections at baseline and during peak exercise stress. Counts in the right and left chambers of the heart are usually separated by a sufficiently long time interval so that it is possible to calculate both right and LVEFs at rest and at peak stress.
Ventricular performance can also be assessed with gated radionuclide ventriculography (multiple gated acquisition [MUGA]). In this technique, a radioactive tracer that remains in the vascular space, such as technetium-99m labeled red blood cells, is administered to the patient. Images over hundreds of cardiac cycles are acquired, using the R wave of the patient’s ECG as the timing marker for image acquisition. The nuclear medicine technologist sets a “window," which defines the length of the cardiac cycles that are to be accepted for imaging. Only the cycles with appropriate R–R intervals are incorporated into the imaging study, with rejection of shorter or longer cycles. Usually the R–R interval is divided into 16 to 24 frames of equal time duration. With each accepted cardiac cycle, counts from that portion of the cardiac cycle (each frame) are added to those of the corresponding frame from preceding cycles. Count data from about 400 to 600 cardiac cycles are allocated to 16 to 24 images representing different sequential portions of the cardiac cycle. Once the acquisition is completed, the images are played in a continuous cine loop to give a time-averaged estimate of ventricular function. The ejection fraction is calculated using background-corrected end-diastolic and end-systolic counts, employing the same formula as that for the first-pass studies. Unlike echocardiography, ejection fraction measurements made with this technique represent time-averaged values, and are not beat-to-beat measurements.
Gated cardiac images may be obtained from differing projections (anterior, right anterior oblique, left anterior oblique, left lateral). In some centers, tomographic (SPECT) gated radionuclide ventriculography is performed. Visual estimates of chamber size and regional wall motion are obtained from the images. If there is a clinical need, additional gated images may be acquired during intervention (stress, inotropic stimulation, sublingual nitroglycerin), to ascertain if there are regional or global differences in ventricular function elicited by the maneuver. A stable R–R interval is needed for gated radionuclide ventriculography. Patients with frequent ectopic beats, irregularly paced rhythms, and/or atrial fibrillation/flutter with an uncontrolled ventricular response rate benefit substantially by medical stabilization of the cardiac rhythm before referral for gated imaging.
Radionuclide ventriculography can be used to assess the severity of regional and global right and left ventricular dysfunction and to determine the effect of medical or interventional treatments on cardiac function. In conjunction with exercise stress, radionuclide ventriculography can be utilized to determine the presence and extent of stress-induced ischemia. Both rest and stress ejection fraction measurements provide useful prognostic information to the clinician. Because ejection fraction measurements obtained with this imaging technique are count based and are not dependent on assumptions about the shape of the ventricle, they are highly reproducible if the imaging study is performed properly.
RADIOACTIVE TRACERS FOR SPECT AND PET
Several different radioactive tracers are available for clinical imaging. Each has unique properties that may make it more or less suitable for a specific imaging task. Factors that influence the choice of tracer for an imaging study include (a) whether SPECT or PET imaging is to be performed, (b) the physiologic property to be visualized (e.g., perfusion, metabolism, neuronal innervation), (c) the patient’s body habitus, (d) the physical characteristics and the biologic behavior of the agent, and (e) the radiation dosimetry of the tracer. A brief overview of the radioactive tracers used for cardiac imaging in the United States is provided, to provide a basic understanding of the agents essential to the daily practice of nuclear cardiology.
SPECT Perfusion Agents
The most commonly used SPECT tracers of myocardial perfusion are technetium-99m sestamibi and technetium-99m tetrofosmin. Thallium-201 chloride is also used for myocardial perfusion imaging, but it imparts a significantly higher radiation dose to the patient than the technetium-99m labeled tracers.
Technetium-99m-Labeled Agents
Technetium-99m labeled sestamibi and tetrofosmin can be prepared locally using kits and technetium-99m eluted from a molybdenum-99 generator. They also are available as prepared unit doses from commercial radiopharmacies. Technetium-99m-labeled tracers result in better image quality than thallium-201 because the higher-energy, monochromatic, 140-keV photons are less subject to scatter and attenuation than the lower energy photons of thallium-201. In addition, the physical half-life of 6 hours permits administration of significantly higher activity doses than thallium-201. The higher count rates achievable with the larger doses yield better image quality and allow image acquisition gated to the patient’s ECG. However, the extraction fractions of techne- tium-99m sestamibi and technetium-99m tetrofosmin (0.45 to 0.60) are lower than that of thallium-201. Therefore, net myocardial uptake of these tracers begins to plateau ("roll off") at lower tissue blood flow rates than thallium-201.
The technetium-99m labeled tracers are retained within myocardium predominantly by selective sequestration within the mitochondria of viable cells. Unlike thallium-201, technetium-99m sestamibi and technetium-99m tetrofosmin have minimal redistribution. As a result, these tracers are well suited for imaging of acute chest pain. Activity administered during chest pain can be imaged at a later time point with accurate depiction of the pattern of myocardial perfusion at the time the tracer was administered.
Thallium-201 Chloride
Thallium-201 is a cationic, metallic element with a physical half-life of 73 hours. Most of the photons emitted by thallium-201 are x-rays, with energies ranging from about 69 to 80 keV. The low energy photons are more susceptible to scatter and attenuation than the 140 keV technetium-99m photons. Because the longer half-life of thallium-201 results a higher radiation dose to the patient, thallium-201 doses are smaller than those of the technetium-99m agents. Thallium-201 images can therefore appear less distinct than technetium-99m images, due to lower image counts and lower photon energies.
Thallium-201 is a potassium analog that enters the cardiac myocyte via the Na+/K+-ATPase pump in the cell membrane. Although thallium-201 uptake by the myocardium is proportional to blood flow at lower blood flows, the amount of the tracer retained in the tissue at higher flow rates underestimates actual tissue perfusion. This “roll-off" of net tissue tracer accumulation at higher flows is common to all diffusible tracers; however, the “roll-off” begins at higher flow rates for thallium-201 than for technetium-99m-labeled agents, because of its greater first-pass extraction fraction.
Following its intravenous administration, blood levels of thallium-201 peak very rapidly, and there is a concentration gradient for the tracer from the vascular space to inside the cell. This gradient provides the stimulus for thallium-201 to enter the cell. As vascular tracer concentrations eventually decline, thallium-201 leaves the cell to go back into the circulation again. For myocardial regions with lower flows at stress (a perfusion defect), the flux of the tracer into and out of the area is slower than in normal tissue. As a consequence, initial images demonstrate a perfusion defect that appears to “fill in” or redistributeon delayed images. For areas with profound ischemia, a defect may be present on rest images, which then fills in on delayed images or on images obtained following the reinjection of a “booster” dose of thallium-201. The implication of redistribution on delayed or reinjection images is that there is viable tissue in that myocardial region, forming the basis for viability imaging with this tracer.
PET Tracers of Myocardial Perfusion
The perfusion tracers used for cardiac PET imaging include rubidium-82 chloride, nitrogen-13 ammonia, and oxygen-15 water.
Rubidium-82
Rubidium-82 has biologic properties similar to potassium. It is eluted directly off of a portable bedside strontium generator, obviating the need for an on-site cyclotron to perform PET perfusion imaging. The physical half-life of rubidium-82 is 75 seconds. Uptake by the cardiac myocyte is predominantly via membrane-bound Na+/K+-ATPase. Clearance from the blood pool is prompt and generally results in high-quality images. Like most other perfusion tracers, net myocardial uptake of rubidium-82 plateaus at higher myocardial blood flows.
Nitrogen-13 Ammonia
Nitrogen-13 ammonia is a perfusion agent that is produced via a cyclotron, and because of its short half-life of 10 minutes, it must be produced on site. In the vascular space, there is a dynamic equilibrium between 13NH4+ and 13NH3 and hydrogen ion. 13NH3 diffuses across the membrane, and is trapped intracellularly via the glutamine synthetase reaction. Like rubidium-82, net myocardial uptake of nitrogen-13 ammonia also decreases at higher blood flows. It has a high single-pass extraction fraction and long tissue retention, which permits ECG-gated image acquisition.
Oxygen-15 Water
Oxygen-15 is cyclotron produced, and its short half-life of 2.1 minutes requires it to be produced on site. Unlike rubidium-82 and nitrogen-13 ammonia, it is freely diffusible and net myocardial uptake of the tracer increases nearly linearly with increasing blood flows. Moreover, tissue retention of the tracer is largely independent of tissue metabolism. It is not rapidly cleared from the systemic circulation and image quality is lower because of relatively high background activity. As a result of these factors, oxygen-15 water is used less frequently for clinical imaging than either rubidium-82 or nitrogen-13 ammonia.
PET Tracers of Myocardial Metabolism
Positron-emitting tracers of myocardial metabolism play an important role in distinguishing between viable and nonviable myocardium in patients with impaired left ventricular function. The most commonly used PET agent for this purpose is FDG. (Other, less commonly used, PET agents used to assess other facets of myocardial metabolism include 11C-palmitate and 11C-acetate). FDG is a glucose analog that enters the cardiac myocyte by facilitated transport. Once within the cell, it competes with glucose for the enzyme hexokinase. FDG is phosphorylated to FDG-6-phosphate, which is then trapped within the cardiac myocytes and not further metabolized. FDG imaging is utilized with perfusion imaging to identify myocardium that is normal (with normal uptake of both the FDG and the perfusion tracer), hibernating (defect on the perfusion images, with preserved uptake of FDG), or scarred (defect on the perfusion images with matching defect on the FDG metabolic images).
SUMMARY
An understanding of the fundamentals of the instrumentation and radioactive tracers employed in nuclear cardiology forms the basis for the appropriate utilization of these advanced imaging techniques in clinical cardiology (see Chapter 25). Although SPECT myocardial perfusion imaging is still primarily performed using conventional gamma cameras, newer instruments with solid-state detectors dedicated to cardiac imaging are assuming a larger role in clinical practice. These advanced cameras, along with the use of iterative reconstruction techniques, permit myocardial perfusion imaging to be performed more rapidly using lower doses of the radioactive tracers. PET/CT perfusion imaging with rubidium-82 chloride or nitrogen-13 ammonia is more accurate for the detection of coronary artery disease than SPECT imaging, imparts a lower patient radiation dose than conventional SPECT imaging, and can yield measurements of rest and hyperemic blood flows as well as flow reserves. If desired, CT angiography of the coronary arteries and/or coronary calcium scoring can be performed on the PET/CT camera in the same setting, providing both anatomic and physiologic information about the state of the heart.
ACKNOWLEDGMENT
The author would like to thank Dr. Gregory G. Bashian for his contributions to the previous edition of this chapter.
SUGGESTED READINGS
Beller GA, Bergmann SR. Myocardial perfusion imaging agents: SPECT and PET. J Nucl Cardiol. 2004;11:71– 80.
Cerqueira MD, Weissman NJ, Dilsizian V, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. J Nucl Cardiol. 2002;9:240–245.
Chandra R. Nuclear Medicine Physics: The Basics. 7th ed. Philadelphia: Wolters Kluwer Health/Lippincott Williams & Wilkins; 2012.
Dilsizian V, Narula J. Atlas of Nuclear Cardiology. China: Current Medicine Group; 2009.
Dilsizian V, Bacharach SL, Beanlands RS, et al. ASNC Imaging guidelines for nuclear cardiology procedures: PET myocardial perfusion and metabolism clinical imaging. Available at: www.asnc. org under Guidelines and Standards.
Heller GV, Hendel R. Nuclear Cardiology: Practical Applications. 2nd ed. China: McGraw-Hill Companies; 2011.
Heller GV, Calnon D, Dorbala S. Recent advances in cardiac PET and PET/CT myocardial perfusion imaging. J Nucl Cardiol. 2009;16:962–969.
Henzlova MG, Cerqueira MD, Hansen CL, et al. ASNC Imaging guidelines for nuclear cardiology procedures: Stress protocols and tracers. Available at: www.asnc.org under Guidelines and Standards.
Henzlova MJ, Duvall WL. The future of SPECT MPI: Time and dose reduction. J Nucl Cardiol. 2011;18:560–567.
Holly TA, Abbott BG, Al-Malla M, et al. ASNC Imaging guidelines for nuclear cardiology procedures: Single photonemission computed tomography. J Nucl Cardiol. 2010;17:941–973.
Iskandrian AE, Garcia EV. Nuclear Cardiac Imaging: Principles and Applications. 4th ed. Oxford: Oxford University Press; 2008.
Sharir T, Slomka PJ, Berman DS. Solid-state SPECT technology: fast and furious. J Nucl Cardiol. 2010;17:890–896.
Travin MI, Bergmann SR. Assessment of myocardial viability; Semin Nucl Med. 2005;35:2–16.
Ziadi MC, Beanlands RSB. The clinical utility of assessing myocardial blood flow using positron emission tomography. J Nucl Cardiol. 2010;17:571–581.
QUESTIONS AND ANSWERS
Questions
1. Which of the following statements is not true?
a. Collimator selection for a gamma camera reflects a trade-off between spatial resolution and efficiency.
b. Photons that pass through the collimator and strike the imaging crystal produce visible light that is then converted to electrical current and amplified by photomultiplier tubes.
c. By setting a more narrow energy window around the isotope’s photopeak, one can increase the sensitivity of a camera to include more scattered photons.
d. Planar images are typically obtained from several fixed views, whereas SPECT images are obtained from multiple projections that are then reconstructed.
2. Assuming comparable biologic behavior, which of the following combinations of gamma-emitting isotopes and administered activity doses would result in the lowest absorbed dose of radiation to a patient?
a. Short half-life, high dose
b. Long half-life, high dose
c. Short half-life, low dose
d. Long half-life, low dose
3. Assuming a “right dominant” coronary artery distribution, match each of the ventricular segments (from the standard 17-segment model) with its most likely corresponding coronary arterial distribution.
(1) Basal anterior
(2) Basal anteroseptal
(3) Basal inferoseptal
(4) Basal inferior
(5) Basal inferolateral
(6) Basal anterolateral
(7) Mid anterior
(8) Mid anteroseptal
(9) Mid inferoseptal
(10) Mid inferior
(11) Mid inferolateral
(12) Mid anterolateral
(13) Apical anterior
(14) Apical septal
(15) Apical inferior
(16) Apical lateral
(17) Apex
a. Left anterior descending (LAD)
b. Right coronary artery (RCA)
c. Circumflex
4. Given the following simplified count data obtained during radionuclide ventriculography, what is the ejection fraction?
End-diastolic counts = 1,000
End-systolic counts = 650
a. 54%
b. 65%
c. 35%
d. 30%
5. Match the following combinations of stress and rest imaging findings with the appropriate interpretation:
6. For which of the following myocardial perfusion imaging protocols would the radioactive tracer(s) utilized provide the lowest estimated effective radiation dose to the patient?
a. Stress and 4-hour redistribution SPECT imaging, using 3.0 mCi of thallium-201 for the stress dose
b. Dual isotope rest and stress SPECT imaging, using 2.5 mCi of thallium-201 for the rest dose and 25 mCi of technetium-99m tetrofosmin for the stress dose
c. Stress only SPECT perfusion imaging with 25 mCi of technetium-99m sestamibi
d. Rest and regadenoson stress rubidium-82 PET imaging with rest and stress doses of 30 mCi each
7. In which of the following patients would gated radionuclide ventriculography be expected to underestimate the left ventricular ejection?
a. A 73-year-old woman with an inferobasal aneurysm
b. A 43-year-old man with a nonrestrictive ventricular septal defect
c. A 64-year-old woman with new onset atrial fibrillation and a poorly controlled ventricular response rate
d. A 59-year-old man who is pacemaker dependent
8. Which of the following statements regarding cardiac positron emission tomography (PET) is true?
a. Radioactive tracer doses used for a PET scanner operating in the three-dimensional (3-D) mode usually are less than for a comparable device operating in the two-dimensional (2-D) mode.
b. Myocardial perfusion imaging must be performed using pharmacologic stress.
c. An on-site cyclotron is necessary for myocardial perfusion imaging.
d. 18F-2-fluoro-2-deoxyglucose (FDG) imaging is used to assess local myocardial uptake of fatty acids in hypoperfused myocardium, distinguishing viable but hypoperfused tissue (preserved uptake) from myocardial scar (matching perfusion and metabolic defects).
9. Which of the following tracers of myocardial perfusion has the highest net tissue uptake at a myocardial blood flow of 4.0 mL/min/g tissue?
a. Oxygen-15 water
b. Thallium-201
c. Technetium-99m sestamibi
d. Rubidium-82
10. Which of the following isotopes liberates the lowest energy photons when it decays?
a. Fluorine-18
b. Rubidium-82
c. Technetium-99m
d. Thallium-201
Answers
1. Answer C: The purpose for setting a narrow energy window around the photopeak of an isotope is to exclude lower-energy photons, which are more likely to be photons that have been scattered and thus lost some of their energy. In doing so, some of the true photons are also excluded, thus decreasing the sensitivity.
2. Answer C: Assuming comparable biologic behavior for the gamma-emitting isotopes, the lower the administered dose, the lower the exposure, thus excluding choices a and b. For the same dose, the isotope with the shorter half-life will result in a lower absorbed dose of radiation because of the shorter duration of exposure.
3. Answers:
a. (1), (2), (7), (8), (13), (14), (17)
b. (3), (4), (9), (10), (15)
c. (5), (6), (11), (12), (16)
4. Answer C: 35%, because LVEF = (end-diastolic counts - end-systolic counts)/(end-diastolic counts) = (1,000 - 650)/(1,000) = (350)/(1,000) = 0.35, or 35%.
5. Answers: a. (iii); b. (i); c. (iv); d. (ii)
6. Answer D: Because of the 73 hour physical half-life of thallium-201, the effective doses from protocols a and b are relatively high (22 and 23.5 milliSieverts, respectively). The estimated effective radiation dose from the stress only SPECT protocol with 25 mCi of technetium- 99m sestamibi is 7.3 milliSieverts, while that for the rubidium-82 PET study is 2.8 milliSieverts.
7. Answer C: The woman with poorly controlled atrial fibrillation will have widely varying R–R intervals over the period of image acquisition, adversely impacting allocation of counts to the appropriate frames (time bins). This usually results in an underestimation of the left ventricular ejection fraction (LVEF). Each of the other patients should have stable R–R intervals and the cardiac rhythm should therefore pose no problems in ECG gating of the study.
8. Answer A: PET scanners operating in a 3-D mode have a greater sensitivity than when operated in a 2-D mode. As a result, a smaller tracer dose is usually used for PET imaging in the 3-D mode as compared to the 2-D mode. While PET myocardial perfusion imaging is usually performed using pharmacologic stress, centers using nitrogen-13 ammonia for imaging can use treadmill exercise because this tracer’s half-life is long enough to permit imaging following recovery from exercise. An on-site cyclotron is not required for PET myocardial perfusion imaging if generator-produced rubidium-82 is selected as the perfusion tracer. FDG imaging depicts regional myocardial glucose uptake, not fatty acid uptake.
9. Answer A: As myocardial blood flow increases, the net myocardial uptake of oxygen-15 water increases linearly. At a hyperemic blood flow of 4.0 mL/min/g, net myocardial uptake should be about four times that at 1.0 mL/min/g. For each of the other tracers listed, the incremental increase in tissue tracer uptake gets progressively smaller as blood flows rise much above 2.0 mL/min/g, that is, tissue uptake plateaus at hyperemic blood flows. For these tracers, net myocardial uptake at a blood flow of 4.0 mL/min/g will be appreciably less than four times that at 1.0 mL/min/g.
10. Answer D: When thallium-201 decays, most of the photons emitted are low energy x-rays (68 to 80 keV). Technetium-99m emits a 140 keV gamma ray when it decays. Fluorine-18 and rubidium-82 decay by positron emission, resulting in the production of 511 keV annihilation photons that are used for imaging.