SPECT and PET Myocardial Perfusion Imaging: Tracers and Techniques

Abstract

The application of a radiotracer technique to measure physiologic parameters, such as pulmonary circulation, dates back to 1927. Despite considerable advances in the application of radiotracer technique for interrogating physiologic and pathophysiologic cardiopulmonary conditions in the early twentieth century, the spatial resolution of the scintigraphic instruments used to measure tracer concentration in the heart, lungs, and blood was limited. The advent of single-photon emission CT (SPECT) in the late 1970s and positron emission tomography (PET) in the 1980s dramatically changed the clinical utility of radiotracer technique for the assessment of myocardial perfusion, viability, and function.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The application of a radiotracer technique to measure physiologic parameters, such as pulmonary circulation, dates back to 1927. Despite considerable advances in the application of radiotracer technique for interrogating physiologic and pathophysiologic cardiopulmonary conditions in the early twentieth century, the spatial resolution of the scintigraphic instruments used to measure tracer concentration in the heart, lungs, and blood was limited. The advent of single-photon emission CT (SPECT) in the late 1970s and positron emission tomography (PET) in the 1980s dramatically changed the clinical utility of radiotracer technique for the assessment of myocardial perfusion, viability, and function.

Both SPECT and PET technologies use similar reconstruction processes to obtain tomographic images of the heart. However, they differ in the type of radiopharmaceuticals and kind of instrumentation used to acquire cardiac images. SPECT allows a noninvasive evaluation of myocardial blood flow by extractable tracers such as ²⁰¹Tl- and ^99mTc-labeled perfusion tracers. PET, on the other hand, allows a noninvasive assessment of regional blood flow, function, and metabolism using physiologic substrates prepared with positron-emitting isotopes such as carbon, oxygen, nitrogen, and fluorine. Radioisotopes commonly used with SPECT emit γ-rays of varying energies and have relatively long physical half-lives. The localization of γ-rays emitted by single-photon–emitting radiotracers in the heart is conventionally accomplished by an Anger scintillation camera (gamma camera), which converts the γ-rays to light photons via sodium iodide scintillation detectors. The gamma camera limits the direction of photons entering the detector by a collimator and then positions each event electronically. More recently, new designs of high-speed SPECT cameras have been introduced, which utilize a series of small, pixilated solid-state detector columns with cadmium zinc telluride or CSI (Tl) crystals, which provide considerably more information for each detected γ-ray. In addition, the design of the solid-state detector design with wide-angle tungsten collimators combined with a novel image reconstruction algorithm provide true three-dimensional, patient-specific images localized to the heart. Compared with the conventional SPECT cameras, the high-speed SPECT systems can provide up to an eightfold increase in count rates, thereby reducing imaging times significantly from 14 to 15 min with a conventional Anger camera to 5–6 min with the newer solid-state cameras, while achieving a twofold increase in spatial resolution from 9 to 12 mm for Anger cameras to 4.3–4.9 mm for cadmium zinc telluride cameras. The radioisotopes used for optimal scintigraphic registration with SPECT cameras are limited to those that emit γ-rays with an energy range that is suitable for the gamma camera and related single-photon devices such as ²⁰¹Tl, ^99mTc, and ¹²³I. Although clinically useful, estimates of relative myocardial blood flow by SPECT are significantly affected by attenuation artifacts that are not reliably corrected for when compared to PET attenuation correction algorithms.

Positron-emitting radioisotopes commonly used with PET emit two γ-rays, 511 keV each, and have relatively short physical half-lives. When the high-energy positron is emitted from a nucleus, it travels a short distance and collides with an electron. The result is complete annihilation of both the positron and the electron, and the conversion of the combined mass to energy in the form of electromagnetic radiation (two γ-rays, 511 keV energy each). Because the γ-rays are perfectly collinear (discharged at 180° to each other) and travel in opposite directions, the PET detectors can be programmed to register only events with a temporal coincidence of photons that strike directly at opposing detectors. This results in improved spatial (4–6 mm) and temporal resolution. Moreover, the PET system is more sensitive than a SPECT system (higher count rate) and provides a more robust soft tissue attenuation correction. The consequence of these advantages with PET is the possibility for a quantitation of the tracer concentration in absolute units.

Figure 2-1.

(a and b) Properties of single-photon emission computed tomography and positron emission tomography flow tracers. To reflect regional myocardial perfusion, radiotracers commonly used with single-photon emission CT (SPECT) and positron emission tomography (PET) must have high extraction by the heart and rapid clearance from the blood. Clinically available radiopharmaceuticals that meet these criteria are ²⁰¹Tl, ^99mTc-labeled sestamibi, and tetrofosmin with SPECT, and ⁸²Rb and ¹³N-ammonia with PET. If radiotracers are not highly extracted (<50 %) or if the residence time in the blood is prolonged (clearance half-time of >5 min), they cannot be used to assess regional perfusion.

Table 2-1. (A and B) Properties of SPECT and PET flow tracers. Radiotracers, mechanisms of myocyte uptake, and usual doses are listed.

Myocardial Perfusion, Uptake, and Clearance

Figure 2-2.

Myocardial blood flow and coronary anatomy: disparate yet complementary information. Regional myocardial blood flow is critically dependent on the driving pressure gradient and the resistance of the vascular bed. Advanced degrees of coronary artery disease may exist at rest (a) without myocardial ischemia due to compensatory dilatation of the resistance vessels. As illustrated in (b and c), at rest, regional myocardial blood flow is preserved in both patent and stenosed coronary artery branches. Such disparity between myocardial blood flow and coronary anatomy attests to the complementary information that a physiologic study such as myocardial perfusion single-photon emission computed tomography or positron emission tomography provides to that of coronary angiography with CT or diagnostic catheterization. In a canine model, over 80 % occlusion of the coronary artery was necessary before ischemia was observed under the basal state. Because the pressure drop across a stenosis varies directly with the length of the stenosis and inversely with the fourth power of the radius (Bernoulli’s theorem), resistance almost triples as the severity of coronary artery stenosis increases from 80 to 90 %. Consequently, during exercise or pharmacologic stress testing, when the resistance to the distal bed and the pressure distending the stenotic coronary artery declines, myocardial ischemia ensues (b and c).

Coronary blood flow in myocardial regions without coronary artery stenosis may increase about two- to threefold during vigorous aerobic exercise. However, in the setting of moderate-to-severe coronary artery stenosis, the degree of coronary flow increase may be attenuated when compared to myocardial regions without coronary artery stenosis (b). The insufficient coronary blood flow increase during stress results in impaired perfusion and myocardial ischemia (c). In patients with coronary artery disease, an inverse relationship has been shown between the increase in myocardial blood flow and the percentage of coronary artery stenosis once the lumen is narrowed by approximately 40–50 %. Thus, when a radiotracer such as thallium is injected at peak exercise, the relative differences in regional myocardial blood flow will be reflected in disproportionate concentrations of regional thallium activity on the stress images. Therefore, myocardial perfusion imaging identifies subcritical coronary artery stenosis when it is performed in conjunction with exercise or pharmacologic stress, but not at rest.

Figure 2-3.

Schematic illustration of radiotracer uptake in relation to regional myocardial blood flow. The radiotracer that most closely parallels myocardial blood flow would be expected to most accurately identify coronary artery narrowing. There are several classes of radiopharmaceuticals that meet these criteria, such as microspheres, ²⁰¹Tl, ^99mTc-labeled perfusion tracers, ¹⁵O-water, ¹³N-ammonia, and ⁸²Rb. Differences in the first-pass extraction of these tracers ultimately determine the regional myocardial tracer uptake relative to regional blood flow. The extraction fraction is determined experimentally in a Langendorff preparation and represents first- or single-pass extraction of the radiotracer from the blood into the myocardium. An ideal myocardial perfusion tracer would be expected to exhibit a linear relationship to myocardial blood flow over a wide range of flow rates in mL/g/min. ¹⁵O-water, a positron emission tomography (PET) myocardial flow tracer, exhibits such a relationship. A linear relationship between the tracer uptake and myocardial blood flow would, therefore, differentiate between regions with normal or high blood flow (supplied by normal coronary arteries) and abnormal or low blood flow (supplied by narrowed coronary arteries). However, this is not the case for all other radiotracers commonly used in clinical practice. In an open-chest canine model of regional myocardial ischemia with dipyridamole-induced hyperemia, thallium showed a more ideal linear relationship between the tracer uptake and myocardial blood flow assessed by microspheres when compared with ^99mTc-labeled myocardial perfusion agents. While the extraction fraction of ²⁰¹Tl is high at 85 %, the extraction fraction of ^99mTc-sestamibi is only 60 % and that of ^99mTc-tetrofosmin is approximately 54 %. Beyond the first-pass extraction, recirculation of the radiotracer in patients allows further extraction of the radiotracers from the blood into the myocardium during that particular physiologic state (rest, exercise, pharmacologic, or mental stress).

At rest, myocardial blood flow is approximately 1 mL/g/min. During physical exercise, myocardial blood flow usually increases two- to threefold, while with pharmacologic vasodilation (adenosine, regadenoson, or dipyridamole) myocardial blood flow exceeds 3 mL/g/min. All clinically available perfusion tracers, single-photon emission CT (SPECT) and PET, demonstrate “roll-off” at high coronary blood flow levels (a deviation from the line of identity). For SPECT tracers, this “roll-off” phenomenon is particularly marked for ^99mTc-sestamibi and ^99mTc-tetrofosmin, and less so for ²⁰¹Tl. This implies that at higher flow levels, relative myocardial tracer uptake may underestimate regional myocardial blood flow and thereby also the underlying coronary artery disease. Clinical studies have shown that this underestimation of regional blood flow deficits does not affect the detection of significant (>70 %) coronary artery stenosis. However, it is important to point out that coronary artery stenosis between 50 and 70 % may go undetected, especially with radiotracers with low extraction fraction and marked roll-off phenomenon.

Figure 2-4.

Blood clearance of radiotracers. Once a radiotracer is injected intravenously at peak stress, it is extracted rapidly from the blood and accumulated in the myocardium in proportion to regional blood flow. All clinically useful radiotracers have extraction fractions above 50 % and are cleared rapidly from the blood in 5–7 min after injection. Because ²⁰¹Tl has a higher first-pass extraction fraction and is cleared more rapidly from the blood than ^99mTc-sestamibi and ^99mTc-tetrofosmin, patients are encouraged to exercise for an additional 1 min after an injection of ²⁰¹Tl at peak exercise and for 2 min after an injection of ^99mTc-sestamibi or ^99mTc-tetrofosmin at peak exercise. If exercise is stopped too early with the ^99mTc perfusion tracers (that is, 1 min rather than 2 min after an injection), residual radiotracer activity in the blood may be taken up at a different physiologic state (under resting condition), thereby underestimating the presence and extent of myocardial ischemia.

Image Interpretation and Quantitation

Figure 2-5.

Myocardial segmentation, standard nomenclature, and vascular territories. Single-photon emission computed tomography myocardial perfusion images are interpreted on the basis of the presence, location, extent, and severity of perfusion defects using a standard 17-segment model [1] and visual scoring. (a) A standard segmentation model divides the left ventricle into three major short-axis slices: apical, midcavity, and basal. The apical short-axis slice is divided into four segments, whereas the midcavity and basal slices are divided into six segments. The apex is analyzed separately, usually from a vertical long-axis slice. Although the anatomy of coronary arteries may vary in individual patients, the anterior, septal, and apical segments are usually ascribed to the left anterior descending (LAD) coronary artery, the inferior and basal septal segments to the right coronary artery (RCA), and the lateral segments to the left circumflex (LCX) coronary artery. The apex can also be supplied by the RCA and LCX artery. (b) Data from the individual short-axis tomograms can be combined to create a bull’s-eye polar plot representing a two-dimensional compilation of all the three-dimensional short-axis perfusion data. Standard nomenclature for the 17 segments is outlined. (c) The two-dimensional compilation of perfusion data can then easily be assigned to specific vascular territories.

Figure 2-6.

Quantitative analysis. Because radionuclide images are intrinsically digital images, the true quantification of tracer uptake in myocardial regions is feasible [2]. (a) The methodology of semiautomatic quantitative circumferential profile analysis is applied to a short-axis ²⁰¹Tl tomogram obtained after exercise in a patient with coronary artery disease. (b) The left ventricular myocardium is divided into 64 sectors, representing four myocardial regions. (c) The patient’s thallium uptake during stress imaging in each section (orange line) and the normal range (mean  ±  2 SD for normal subjects; shaded area with blue line). The patient’s count profile displays the distribution of counts in the tomogram relative to maximal counts counterclockwise, starting at 0, which represents the high lateral region that is designated as the normal reference value of 100 % (maximal count density). Whenever a region of the circumferential profile falls below the lower limit of normal, that region of the patient’s myocardium is considered to have a perfusion defect. In this patient, thallium perfusion defects are apparent in the anterior and septal regions. LAD left anterior descending, LCX left circumflex, RCA right coronary artery.

Figure 2-7.

Common variations and artifacts of myocardial perfusion single-photon emission computed tomography. Although normal myocardial perfusion single-photon emission CT (SPECT) images appear to have homogeneous radiotracer uptake, regional inhomogeneities are commonly present that are related to normal structural variation, tissue attenuation, and abdominal visceral activity, as well as technical factors associated with image acquisition. During SPECT acquisition, the camera is physically closer to the lateral wall (which is in close proximity to the lateral chest wall) than to the other myocardial regions. Consequently, the lateral region is subject to less soft tissue attenuation and is associated with a more efficient count capture. This should not be interpreted as relative hypoperfusion in all other myocardial regions (which will appear to have slightly less uptake). Thus, in normal SPECT images, the lateral region is usually the area with maximal radiotracer uptake, and it is often difficult to detect subtle perfusion defects visually within the lateral region because the activity of the radiotracer may remain greater than or similar to other myocardial regions. This would especially be the case in patients with multivessel disease, in whom an equivalent reduction in perfusion in several myocardial regions would still result in greater tracer activity in the lateral territory, termed balanced reduction in flow. Normal structural variations include the “drop-out” of the upper septum (transition from the muscular to membranous septum) and apical thinning (an anatomically thinner apex may appear as a perfusion defect). (a) An example of apical thinning is shown. Soft tissue attenuation can present as breast attenuation (commonly in women with large or dense breasts). (b) This figure is an example of a patient with breast attenuation, which shows mildly decreased uptake in the anterior region. (c) Breast attenuation can be recognized on the rotating planar projection images as a photopenic shadow over the heart that has the contour of a breast. The demonstration of preserved wall thickening in the anterior region by gated SPECT imaging may be helpful in differentiating an attenuation artifact from myocardial infarction. (d) Similarly, inferior wall attenuation can be caused by the diaphragm or by other abdominal visceral structures, such as the liver or bowel, either overlapping or near the inferior wall or ascites and large pleural effusions. Soft tissue attenuation such as that caused by the diaphragm can be corrected with attenuation correction. However, attenuation due to radiotracer activity within abdominal visceral structures, such as the liver or bowel, either overlapping or near the inferior wall, cannot be resolved with an attenuation correction algorithm. Adjacent abdominal visceral activity may falsely increase the number of counts that are assigned within the heart (in which case the adjacent myocardium appears to be “hot”) or cause a “ramp filter” or “negative lobe” artifact (in which case the adjacent myocardium appears to be “cool”).

(e) This figure is an example of a patient motion artifact exhibiting as a “hurricane sign.” Images in the top row show the consequence of significant patient motion in creating artifactual regional perfusion defects. In the bottom row, when the images are reacquired in the same patient without motion artifact, the distribution of the radiotracer appears homogeneous in all myocardial regions without regional perfusion defects.

Figure 2-8.

Reversible and irreversible perfusion defects: myocardial ischemia and infarction. An imbalance between oxygen supply (usually due to reduced myocardial perfusion) and oxygen demand (determined primarily by the rate and force of myocardial contraction) is termed ischemic myocardium. A clinical presentation of such an imbalance may be symptomatic (angina pectoris) or asymptomatic (silent ischemia). If the oxygen supply–demand imbalance is transient (i.e., triggered by exertion), it represents reversible ischemia. The scintigraphic hallmark of myocardial ischemia is a reversible perfusion defect. (a) Examples of patients with reversible perfusion defects in the left anterior descending (LAD), left circumflex (LCX), and right coronary artery (RCA) territories are shown. On the other hand, if a regional oxygen supply–demand imbalance is prolonged (i.e., during myocardial infarction), high-energy phosphates will be depleted, regional contractile function will progressively deteriorate, and cell membrane rupture with cell death will follow (myocardial infarction). The scintigraphic hallmark of myocardial infarction is a fixed or irreversible perfusion defect. (b) Examples of patients with irreversible (fixed) perfusion defects in the LAD, LCX, and RCA territories are shown.

Figure 2-9.

Clinically relevant extracardiac activity. Beyond myocardial perfusion, additional important abnormal findings can be present on the rotating planar projection images, such as lung uptake, parathyroid adenoma, or lung cancer, which should be observed and reported. (a) Increased lung uptake (arrows) is associated with extensive coronary artery disease and an adverse prognosis. In patients with extensive myocardial ischemia and/or left ventricular dysfunction, it is likely that increases in the left atrial and pulmonary capillary wedge pressures slow the pulmonary transit of the radiotracer, thereby allowing more time for extraction or transudation of the radiotracer into the interstitial spaces of the lung. Lung uptake has been more extensively validated with ²⁰¹Tl than ^99mTc perfusion tracers. Due to differences in the biodistribution and clearance of ²⁰¹Tl and ^99mTc perfusion tracers, thallium images are acquired within a few minutes after exercise (minimal splanchnic and background activity) while the ^99mTc-sestamibi and ^99mTc-tetrofosmin are usually acquired 15–30 min after exercise and 30–60 min after pharmacologic vasodilation (liver uptake is more prominent than the heart if imaged too early). Thus, lung uptake, even if it had been present early after stress, may be missed with ^99mTc-sestamibi and ^99mTc-tetrofosmin due to more delayed imaging after stress when compared with thallium.

The rotating planar projection images allow the visualization of noncardiac structures such as lung, breast, and the thyroid gland. An abnormal uptake of ²⁰¹Tl, ^99mTc-sestamibi, and ^99mTc-tetrofosmin in the neck can identify parathyroid adenoma, while in the chest they can identify primary (such as lung or breast cancer), ectopic parathyroid adenoma, or metastatic lesions. (b) An example of a patient with findings of a parathyroid adenoma (arrow) in a stress ^99mTc-sestamibi myocardial perfusion single-photon emission CT (SPECT) is shown. (c) Similarly, an example of a patient with findings of solitary lung nodule (arrow) on a stress ^99mTc-sestamibi myocardial perfusion SPECT is shown. (d) Intense ^99mTc-sestamibi uptake (arrow), superior to the heart, represents an anterior mediastinal mass, which is compatible with thymoma. While thymomas are the most common neoplasm of the anterior mediastinum, representing 20 % of all anterior mediastinal masses in the adult population, their incidence is rather rare, only 0.15/100,000 cases. The differentiation of benign from malignant thymoma cannot be made on the basis of size or intensity of ^99mTc-sestamibi uptake. Additional studies are required to further characterize the lesion. (e) Example of a patient with pericardial effusion. A photopenic halo of decreased activity surrounding the entire heart is consistent with pericardial effusion, which could be a consequence of myopericarditis; an infection such as tuberculous pericardial effusion; or a malignant effusion such as breast cancer, melanoma, or lung cancer (d – Courtesy of Jeffrey R. Folk and e – Courtesy of Rory Hachamovitch).

Figure 2-10.

Detection of angiographic coronary artery disease with radiotracers. Extensive literature exists on the diagnostic yield of stress single-photon emission CT (SPECT) myocardial perfusion imaging [3–14]. Among 1,827 patients referred for the evaluation of chest discomfort (pooled data from 12 studies performed between 1989 and 1999), the overall sensitivity of myocardial perfusion SPECT for the detection of angiographic coronary artery disease was 91 %, the specificity was 72 %, and the normalcy rate (in subjects with low likelihood for coronary artery disease who did not undergo coronary angiography) was 91 %.

Table 2-2. Clinical indications for myocardial perfusion imaging. The clinical indications for stress–rest myocardial perfusion single-photon emission CT (SPECT) imaging are well established [15]. Most patients are referred due to chest pain symptoms and suspected coronary artery disease (CAD). However, patients with known CAD are referred as well. In these patients, the purpose of testing may be to evaluate the effect of therapy or to determine the cause of changes in symptom patterns. In addition, many patients are referred for risk stratification after acute myocardial infarction. Stress–rest SPECT imaging plays an important role in the preoperative evaluation of patients who are scheduled to undergo major noncardiac surgery. The most important and useful clinical application of SPECT myocardial perfusion imaging is to stratify patients into low- and high-risk categories and, thus, contribute to the management of patients.

Table 2-3. High- and low-risk single-photon emission computed tomography images. Single-photon emission CT (SPECT) images should not be interpreted as either normal or abnormal. The prognosis of a patient is related to the degree of myocardial perfusion abnormality. Quantification or semiquantification provides that important prognostic information. High-risk SPECT images are characterized by large perfusion defects in the stress images that involve multiple coronary artery territories (if two or more coronary territories are involved, the study should be considered to be high risk). Large stress-induced reversible defects represent extensive myocardial ischemia, which may be associated with increased lung uptake, transient ischemic left ventricular cavity dilation, and transient increased right ventricular myocardial visualization.

SPECT Techniques: ²⁰¹TI

Figure 2-11.

Stress-redistribution ²⁰¹Tl protocol. Schematic diagrams of ²⁰¹Tl uptake and redistribution in normal and ischemic myocardium (a), and a stress-redistribution protocol (b). While the initial distribution of ²⁰¹Tl (early after intravenous injection) is proportional to regional blood flow, the later distribution of ²⁰¹Tl over a 3- to 4-h period, the redistribution phase, is a function of regional blood volume and is unrelated to flow. During the redistribution phase, there is a continuous exchange of ²⁰¹Tl between the myocardium and the extracardiac compartments, driven by the concentration gradient of the tracer and myocyte viability. Thus, the extent to defect resolution, from the initial to delayed redistribution images over time (a reversible defect), reflects one index of myocardial viability. When only nonviable, scarred myocardium is present, the initial ²⁰¹Tl defect (an irreversible defect) persists over time without redistribution. When both viable and scarred myocardium are present, ²⁰¹Tl redistribution is incomplete, giving the appearance of partial reversibility. Thus, the initial phase of ²⁰¹Tl studies reflects reductions in flow caused by coronary artery narrowing, while the delayed, redistribution phase of ²⁰¹Tl studies reflects myocardial potassium space, differentiating viable from scarred myocardium.

Figure 2-12.

Prognostic value of thallium scintigraphy. Beyond its value as a perfusion and viability tracer, the stress-redistribution ²⁰¹Tl studies provide useful information regarding patient outcome and prognosis. In patients with chronic ischemic heart disease, an increased lung-to-heart ratio after stress, transient left ventricular cavity dilatation, and extensive reversible and irreversible ²⁰¹Tl defects have been shown to be important predictors of an adverse outcome. Similarly, the combination of reversible ²⁰¹Tl defects and increased lung-to-heart ratio has been shown to differentiate between low- and high-risk patients after an acute myocardial infarction. Among patients with acute myocardial infarction who had a predischarge submaximal exercise treadmill test (ETT), ²⁰¹Tl scintigraphy, and coronary angiography, ²⁰¹Tl identified the low-risk subgroup much better than submaximal exercise treadmill testing or coronary angiography. CAD coronary artery disease (Adapted from Gibson et al. [16]).

Figure 2-13.

Late (24-h) ²⁰¹Tl redistribution. (a) Late redistribution protocol after stress-redistribution ²⁰¹Tl imaging is shown. In some patients with critically stenosed coronary arteries, the initial uptake of ²⁰¹Tl in the ischemic region is low and the accumulation of the tracer from the recirculating ²⁰¹Tl in the blood is slow. Consequently, ischemic but viable myocardium may appear to be irreversible over the 3- to 4-h redistribution period and may mimic the appearance of scarred myocardium. However, if more time is allowed for redistribution, a greater number of viable myocardial regions may be differentiated from scarred myocardium.

(b) Polar maps demonstrating the effect of late ²⁰¹Tl redistribution. A bull’s-eye image of ²⁰¹Tl immediately after exercise (top panel, left) shows a marked decrease in tracer uptake throughout the anterior, septal, apical, and inferior regions, with partial redistribution in the 4-h delayed image (center). However, in the late (17-h) redistribution image (right), there is complete reversibility in all myocardial regions, which is suggestive of extensive myocardial ischemia rather than a scar. After successful percutaneous transluminal coronary angioplasty, the bull’s-eye image of ²⁰¹Tl immediately after exercise (bottom panel, left) shows normal distribution of the tracer throughout all myocardial regions, documenting the accuracy of late redistribution ²⁰¹Tl and the absence of myocardial scarring. In patients who are treated with revascularization, 95 % of segments that demonstrated late redistribution showed improved ²⁰¹Tl uptake after revascularization. However, as with early (3–4-h) redistribution, the absence of late redistribution underestimates the presence of viable myocardium (right). Up to 37 % of segments that remained irreversible in both early and late redistribution studies showed an improvement in function after revascularization [17]. Moreover, despite implementing longer imaging time, a number of late redistribution studies had suboptimal count statistics at 24 h. The data suggest that although late ²⁰¹Tl imaging improves the identification of viable myocardium when compared with early redistribution imaging, it continues to underestimate segmental improvement after revascularization (b – From Cloninger et al. [18]; with permission).

Figure 2-14.

Myocardial thallium uptake and clearance in relation to blood activity of thallium. The redistribution of ²⁰¹Tl is dependent, in part, on the blood levels of ²⁰¹Tl. The redistribution of ²⁰¹Tl in a given myocardial region depends not only on the severity of the initial defect poststress but also on the presence of viable myocytes, the concentration of the tracer in the blood, and the rate of decline of ²⁰¹Tl levels in the blood. During the redistribution phase, there is continuous exchange of ²⁰¹Tl between the myocardium and the extracardiac compartments, which is driven by the concentration gradient of the radiotracer across the myocytes, as well as blood and myocyte viability. (a) If the blood level of ²⁰¹Tl remains the same (or increases) during the period between stress and redistribution imaging, then a stress-induced defect in a region with viable myocytes that can accumulate ²⁰¹Tl in the redistribution phase will appear to be reversible. (b) If the blood level of ²⁰¹Tl is low (or decreases) during the imaging interval, the delivery of ²⁰¹Tl may be insufficient and the stress-induced ²⁰¹Tl defect may remain irreversible even though the underlying myocardium is viable. Thus, some ischemic but viable regions may show no redistribution on either early (3- to 4-h) or late (24-h) imaging, unless blood levels of ²⁰¹Tl are increased (Adapted from Dilsizian [19]).

Figure 2-15.

Thallium reinjection. (a) ²⁰¹Tl reinjection differentiates ischemic but viable myocardium from scarred myocardium by augmenting the blood levels of ²⁰¹Tl at rest. A viable segment may be asynergic on the basis of repetitive stunning and hibernation. Thus, an asynergic but viable region may have reduced (but not absent) blood flow at rest (hibernation) or transient reduction in blood flow after a period of ischemia (stunning). Although standard stress 3- to 4-h redistribution ²⁰¹Tl scintigraphy may underestimate the presence of ischemic but viable myocardium in many patients with coronary artery disease, reinjecting ²⁰¹Tl at rest after stress 3- to 4-h redistribution imaging substantially improves the assessment of myocardial ischemia and viability in up to 49 % of patients with apparently irreversible defects [2]. The theory that myocardial regions identified by ²⁰¹Tl uptake following ²⁰¹Tl reinjection represent viable myocardium is supported by improved regional function after revascularization and preserved metabolic activity by [¹⁸F]-fluorodeoxyglucose PET. In addition, a significant inverse correlation between the magnitude of ²⁰¹Tl activity after reinjection and the regional volume fraction of interstitial fibrosis has been demonstrated in comparative clinicopathologic studies [20].

It is possible that the initial myocardial uptake of ²⁰¹Tl (postinjection) reflects regional blood flow while the redistribution of ²⁰¹Tl in a given defect depends not only on the severity of the initial defect but also on the presence of viable myocytes, the concentration of the tracer in the blood, and the rate of decline of ²⁰¹Tl levels in the blood. Thus, the heterogeneity of regional blood flow observed in the initial stress-induced ²⁰¹Tl defects may be independent of the subsequent extent of ²⁰¹Tl redistribution. If the blood level of ²⁰¹Tl remains the same (or increases) during the period between stress and 3- to 4-h redistribution imaging, then an apparent defect in a region with viable myocytes that can retain ²⁰¹Tl should improve. On the other hand, if the serum ²⁰¹Tl concentration decreases during the imaging interval, the delivery of ²⁰¹Tl may be insufficient, and the ²⁰¹Tl defect may remain irreversible although the underlying myocardium is viable. This suggests that some ischemic but viable regions may never redistribute, even with late (24-h) imaging, unless serum levels of ²⁰¹Tl are increased. (b) This hypothesis is supported by a study in which ²⁰¹Tl reinjection was performed immediately after 24-h redistribution images were obtained [6]. Improved ²⁰¹Tl uptake after reinjection occurred in 40 % of defects that appeared irreversible on late (24-h) redistribution images. Thus, the reinjection of 1 mCi of ²⁰¹Tl at rest immediately after either stress 3- to 4-h redistribution or stress 24-h redistribution studies, followed by image acquisition 10–15 min later, significantly improves the assessment of myocardial ischemia and viability (Adapted from Dilsizian [19]).

Figure 2-16.

Beneficial effect of ²⁰¹Tl reinjection in the clinical setting. (a) Short-axis tomograms demonstrate extensive ²⁰¹Tl defects in the anterior and septal regions in stress images (top row) that persist in redistribution images (center row) but improve markedly in reinjection images (bottom row). Among patients who had coronary artery revascularization, 87 % of the myocardial regions identified as viable by reinjection studies had normal ²⁰¹Tl uptake and improved regional wall motion after revascularization. In contrast, all regions with irreversible defects in reinjection imaging before revascularization had persistent wall motion abnormality after revascularization [2]. (b) Similar results were obtained when ²⁰¹Tl reinjection was performed immediately after late (24-h) redistribution imaging. Improved ²⁰¹Tl uptake after reinjection occurred in 40 % of regions (involving 60 % of patients) that appeared to be fixed in late redistribution imaging (a – Adapted from Dilsizian et al. [2]; (b) Adapted from Kayden et al. [21]).

Figure 2-17.

Postrevascularization functional outcome of asynergic regions in relation to prerevascularization ²⁰¹Tl patterns of normal, reversible, partially reversible, mild-to-moderate irreversible, and severe irreversible defects using the stress-redistribution-reinjection ²⁰¹Tl protocol. The probabilities of functional recovery after revascularization were over 90 % in normal or completely reversible defects, 63 % in partially reversible defects, 30 % in mild-to-moderate irreversible defects, and 0 % in severe irreversible defects. Asynergic regions with reversible defects (complete or partial) in the prerevascularization ²⁰¹Tl study were shown to be more likely to improve function after revascularization when compared with asynergic regions with mild-to-moderate irreversible defects (79 vs. 30 %, respectively; P  <  0.001). Even at a similar mass of viable myocardial tissue (as reflected by the final ²⁰¹Tl content), the presence of inducible ischemia (reversible defect) was associated with an increased likelihood of functional recovery (Adapted from Kitsiou et al. [22]).

Figure 2-18.

Incremental prognostic value of ²⁰¹Tl reinjection. In patients with prior myocardial infarction and left ventricular dysfunction in whom the assessment of myocardial viability is of clinical relevance, ²⁰¹Tl reinjection (Tl-RI) imaging provides incremental prognostic information to clinical, exercise tolerance testing (ETT), and ²⁰¹Tl stress-redistribution (Tl-RD) imaging. Similarly, in patients with chronic coronary artery disease and prior myocardial infarction, the scintigraphic variable that was the strongest predictor of hard events (cardiac death or myocardial infarction) was the presence of more than three irreversible defects that remained irreversible after ²⁰¹Tl reinjection (Adapted from Petretta et al. [23]).

Figure 2-19.

Rest-redistribution ²⁰¹Tl protocol. (a) The stress-redistribution-reinjection ²⁰¹Tl protocol provides important diagnostic information regarding both inducible ischemia and myocardial viability. In most cases, the identification of myocardial ischemia is much more important clinically in terms of patient management and risk stratification than is knowledge of myocardial viability. However, if the clinical question is one of the presence and extent of viable myocardium within a dysfunctional region and not inducible ischemia, then it is reasonable to perform rest-redistribution ²⁰¹Tl imaging only.

(b) Rest-redistribution short-axis ²⁰¹Tl tomograms are shown for a patient with chronic coronary artery disease. There are extensive ²⁰¹Tl perfusion defects in the anterior, distal anteroseptal, and inferior regions in the initial rest images (top row). In the delayed (3- to 4-h) redistribution images (bottom row), the anterior region remains fixed (scarred myocardium), while the inferior and distal anteroseptal regions show significant reversibility, which is suggestive of hibernating but viable myocardium [24].

Figure 2-20.

Anatomic assessment of the coronary arteries alone does not differentiate viable from scarred myocardium. A patient with a prior history of hypertension and hyperlipidemia presents with a 2-month history of substernal chest pain. The electrocardiogram shows inverted T waves in the inferolateral leads (a), while the coronary angiogram shows total occlusion of the left circumflex (LCX) (arrow) (b) and the proximal right coronary artery (RCA) (c) (arrow). There are extensive perfusion defects in the lateral and inferior regions (d) in the initial rest images (top row; arrows) that become reversible in the delayed (3- to 4-h) redistribution images (bottom row), providing evidence for hypoperfused but viable myocardium in the LCX and RCA vascular territories. The anatomic assessment of the coronary arteries alone is insufficient to determine whether the myocardium subtended by the totally occluded vessels is viable or scarred. Accurate distinction between viable (hibernating or stunned) and scarred myocardium has important clinical implications. Ideally, such information may be used to guide therapeutic decisions for revascularization and risk stratification. The scintigraphic finding of reduced regional blood flow (rest ²⁰¹Tl images) but preserved cell membrane integrity (redistribution ²⁰¹Tl images) in the dysfunctional myocardial regions provides the most direct evidence of myocardial hibernation. In view of the findings on the rest-redistribution ²⁰¹Tl study, the patient was referred for coronary artery bypass surgery with an uneventful postoperative course.

Figure 2-21.

Prognostic value of rest-redistribution ²⁰¹Tl single-photon emission computed tomography. In patients with chronic ischemic left ventricular dysfunction, the demonstration of redistribution on rest ²⁰¹Tl imaging protocols portends a higher mortality rate with medical therapy than do patients with a comparable degree of left ventricular dysfunction without evidence of redistribution. (a) An actuarial survival curve in 81 medically treated patients is shown; 38 patients (mean left ventricular ejection fraction [LVEF]  =  26  ±  7 %) showed redistribution in rest ²⁰¹Tl images, and 43 patients (mean LVEF  =  27  ±  8 %) showed no redistribution. Moreover, in a nonrandomized, retrospective study with rest-redistribution ²⁰¹Tl, survival and survival without myocardial infarction tended to be significantly higher in patients with chronic ischemic left ventricular dysfunction treated with coronary artery revascularization compared with those treated with medical therapy alone. (b) An actuarial survival curve in 85 patients with evidence of myocardial viability by rest-redistribution ²⁰¹Tl is shown; 38 patients underwent coronary artery revascularization, and 47 patients were treated medically (Adapted from Gioia et al. [25, 26]).

Figure 2-22.

Extent of myocardial viability assessed by rest-redistribution ²⁰¹Tl and patient outcome. Considering the survival advantage of coronary artery revascularization when compared with medical therapy in patients with chronic ischemic left ventricular dysfunction, one might question whether the preoperative assessment of myocardial viability is necessary in making revascularization decisions. Should coronary artery revascularization be considered in all patients with chronic ischemic left ventricular dysfunction with or without evidence of myocardial viability? Event-free survival in a retrospective study in patients with preoperative rest-redistribution ²⁰¹Tl testing who had coronary artery bypass surgery is shown. Perioperative and long-term postoperative survival is significantly better in patients with evidence of significant myocardial viability on rest-redistribution ²⁰¹Tl compared with those patients with less evidence of myocardial viability. EF ejection fraction (Adapted from Pagley et al. [27]).

Table 2-4. Single-photon emission computed tomography techniques: ²⁰¹Tl. The myocardial extraction of ²⁰¹Tl is dependent on energy utilization, membrane ATPase, and active transport. ²⁰¹Tl does not actively concentrate in regions of infarcted or scarred myocardium. Thus, decreased myocardial ²⁰¹Tl uptake early after injection could be caused either by reduced regional blood flow or by infarction. Experimental studies with ²⁰¹Tl have shown that the cellular extraction of ²⁰¹Tl across the cell membrane is unaffected by hypoxia unless an irreversible injury is present. Similarly, pathophysiologic conditions of chronic hypoperfusion (hibernation) and postischemic dysfunction (stunning), in which regional contractile function is impaired in the presence of myocardial viability, do not adversely alter extraction of ²⁰¹Tl. SPECT single-photon emission CT.

SPECT Techniques: ^99mTc-labeled Perfusion Tracers

Figure 2-23.

Alternatively, dual-isotope gated single-photon emission computed tomography imaging could be performed, which combines rest-redistribution ²⁰¹Tl (for viability) with stress ^99mTc-sestamibi or -tetrofosmin (for perfusion), thereby taking advantage of the favorable properties of each of the two tracers. RD redistribution.

Figure 2-24.

Cellular kinetics of ²⁰¹Tl and ^99mTc-sestamibi during metabolic inhibition in cultured chick embryo cardiac myocytes, independent of perfusion. (a) Oxidative phosphorylation and glycolysis were inhibited simultaneously by rotenone (10 um) and iodoacetate (1 mmol/L), respectively, producing a decline in the myocellular ATP content. Under these conditions, the initial extraction efficiency of ²⁰¹Tl and ^99mTc-sestamibi responded in divergent ways to ATP depletion. The extraction efficiency of ²⁰¹Tl declined within 20 min of metabolic inhibition by 50–70 %, while the extraction efficiency of ^99mTc-sestamibi (^99mTc-MIBI) increased significantly by 10–20 min and remained elevated for the first 40–60 min of metabolic inhibition. The observed disparity in initial uptake rates between ²⁰¹Tl and ^99mTc-MIBI during mild-to-moderate metabolic injury may explain, on a metabolic basis alone, the clinical observation that ^99mTc-MIBI defects are smaller than those assessed by ²⁰¹Tl. (b) Images taken with ²⁰¹Tl 5–10 min after stress (top row) and with ^99mTc-MIBI 2 h after stress (bottom row) are shown for a patient who performed the same level of exercise with both tracers. A quantitative left ventricular mass algorithm provided similar measures of total mass for ²⁰¹Tl (197 g) and for ^99mTc-MIBI (189 g). However, the stress-induced defect mass derived from ²⁰¹Tl imaging (41 g) is significantly larger than that detected by ^99mTc-MIBI (30 g). No transmural defects are present on the ^99mTc-MIBI images (a – Adapted from Piwnica-Worms et al. [28]; b – From Narahara et al. [29]; with permission).

Figure 2-25.

Pharmacologic stress. In canine models of moderate (a) and severe (b) coronary artery occlusion, ²⁰¹Tl and ^99mTc-sestamibi myocardial perfusion defect sizes are compared during pharmacologic stimulation and with postmortem staining to define the extent of the hypoperfused region. These bull’s-eye displays are from four representative experiments of moderate coronary artery stenosis during pharmacologic stimulation for ²⁰¹Tl and ^99mTc-sestamibi, and the corresponding pathologic polar displays from the same four experiments are shown. The extent of ²⁰¹Tl myocardial perfusion defect size (but not ^99mTc-sestamibi) approaches the hypoperfused area in the corresponding pathologic display. The ^99mTc-sestamibi defect size occupies only 37 % of the area of the defect in the ²⁰¹Tl images of the same dog, and the counts within the defects are 39 % higher for ^99mTc-sestamibi compared to ²⁰¹Tl (a). On the other hand, when coronary artery occlusion is near total (severe), ²⁰¹Tl and ^99mTc-sestamibi show similar defect contrast and areas (b). These observations in canines are similar to the experimental observations made in cultured myocytes (From Leon et al. [30]; with permission).

Figure 2-26.

^99mTc-sestamibi and ²⁰¹Tl activities in myocardial biopsies. A change in defect size of ^99mTc-sestamibi (^99mTc-MIBI) with time (redistribution) has been shown both in animal models and in patients with chronic coronary artery disease [31, 32]. Depending on the level of blood activity of ^99mTc-MIBI after stress, continued uptake by the myocardium after the first pass may reduce the defect severity and area in the hypoperfused region. In the early comparative studies of ²⁰¹Tl and ^99mTc-MIBI, ²⁰¹Tl images were acquired 5–10 min after injection, while ^99mTc-MIBI images were acquired 1–2 h after injection. The 1- to 2-h delay between ^99mTc-MIBI injection and imaging was based on the best compromise between a high myocardial count rate and low background activity, and on the assumption that ^99mTc-MIBI does not “redistribute” over time.

Following transient ischemia and reperfusion after 5 min in a canine model, there was evidence for change in the defect size of ^99mTc-MIBI with time (a), albeit more slowly and less completely when compared to ²⁰¹Tl redistribution (b). For both ^99mTc-MIBI and ²⁰¹Tl, the consistent fall in the normal zone activity and rise in the ischemic zone activity over the 3-h time interval, which is consistent with redistribution, is noted. It is important to point out, however, that there is no change in the ^99mTc-MIBI defect size between the 5- and 30-min time intervals. In view of these and other similar reports, it is now recommended that ^99mTc-MIBI images be acquired earlier, approximately 30 min after injection of the tracer (Adapted from Li et al. [31]).

Figure 2-27.

Clinically relevant change in the defect size of ^99mTc-sestamibi with time (redistribution) in two patients who experienced exercise ^99mTc-sestamibi studies. Myocardial single-photon emission computed tomography images were obtained from two different patients and are presented in the short-axis plane (top row) and in the vertical long-axis plane (bottom row) after exercise and at rest. In the short-axis plane, there is no change in ^99mTc-sestamibi defect size from 20 min to 2 h after exercise. However, by 6 h there is a significant change in the defect size in the inferoseptal region (open arrows) but not in the anteroseptal region (closed arrows). In the injected image taken at rest, complete normalization of all perfusion defects is seen, which suggests that delayed ^99mTc-sestamibi images alone do not provide accurate information regarding defect reversibility. In the vertical long-axis plane, there is significant change in ^99mTc-sestamibi defect in the inferior region (closed arrows) from 20 min to 2 h after exercise (redistribution), without further fill-in at 6 h or in the rest-injected ^99mTc-sestamibi images. Although interpretation of ^99mTc-sestamibi data should be viewed cautiously when imaging is delayed by 2 h or more after stress (due to underestimation of the defect size and extent of myocardial ischemia), the same concept does not apply for rest-injected ^99mTc-sestamibi studies. On the contrary, delaying the ^99mTc-sestamibi images by 2 h or more after rest injection may improve myocardial viability assessment (From Franceschi et al. [33]; with permission).

Figure 2-28.

Nitrate administration before rest ^99mTc-sestamibi or ^99mTc-tetrofosmin injection. Considering the kinetics of ^99mTc-sestamibi and ^99mTc-tetrofosmin, uptake of these radiotracers in myocardial regions with reduced perfusion and partially impaired viability appears to be influenced by regional perfusion rather than myocyte viability. In view of the limitations in the clinical setting of rest-injected ^99mTc-sestamibi and ^99mTc-tetrofosmin for assessing myocardial viability, some investigators have proposed injecting the radiotracers during nitrate infusion. In addition to lowering the preload and afterload, nitrates may cause vasodilatation of the flow, limiting epicardial coronary arteries as well as collateral vessels. The injection of ^99mTc-sestamibi during nitrate infusion (10 mg of isosorbide dinitrate in 100 mL of isotonic saline solution infused over 20 min) is shown to improve the accuracy of ^99mTc-sestamibi for predicting the recovery of regional and global left ventricular function after ­revascularization. In this patient example with anterior myocardial infarction and single vessel left anterior descending (LAD) coronary artery disease, the prerevascularization baseline images (left) show anteroapical akinesis and global left ventricular ejection fraction (LVEF) of 38 % in first-pass radionuclide angiography associated with a large anterior and apical ^99mTc-sestamibi perfusion defect (63 % of the LAD vascular territory in the bull’s-eye image at rest). The ^99mTc-sestamibi images acquired after nitrate infusion (center) show improvement in the anteroapical wall motion associated with an increase in global LVEF to 42 % and a decrease in the extent of ^99mTc-sestamibi perfusion defect size to 42 % of the LAD vascular territory. After revascularization of the LAD (right), there is improvement in the anteroapical wall motion at rest, an increase in global LVEF to 45 %, and a decrease in the extent of ^99mTc-sestamibi perfusion defect to 38 % of the LAD vascular ­territory [34] (Courtesy of Roberto Sciagra).

Figure 2-29.

Quantitation of severity of reduction in myocardial perfusion at rest. Another approach that may overcome, in part, the limitations of ^99mTc-sestamibi and -tetrofosmin in assessing myocardial viability is to quantify the severity of regional tracer activity, i.e., the severity of myocardial perfusion at rest. Among 18 patients with coronary artery disease who had revascularization, a good correlation between the quantitative regional activities of ²⁰¹Tl (on redistribution imaging after rest injection) and ^99mTc-sestamibi (at rest) is shown. Moreover, this scatterplot shows that at a 60 % threshold level for both radiotracers, dysfunctional myocardial regions that improve function after revascularization (blue circles) can be differentiated from dysfunctional myocardial regions that do not improve function after revascularization (orange circles). The positive and negative predictive accuracies attained when the severity of radiotracer defects were quantitated are 80 and 96 %, respectively. Considering the kinetics of ^99mTc-sestamibi, in myocardial regions with decreased blood flow and partially impaired viability, the uptake of ^99mTc-sestamibi appears to be influenced by regional perfusion rather than myocyte viability [35] (Adapted from Udelson et al. [36]).

Figure 2-30.

Underestimation of myocardial viability by rest ^99mTc-sestamibi single-photon emission computed tomography. Using dual-isotope injection at rest (same physiologic state) and simultaneous acquisition using single-photon emission CT (SPECT; accurate anatomical alignment), these images show mismatch between rest cardiac perfusion assessed by ^99mTc-sestamibi and metabolism assessed by [¹⁸F]-fluorodeoxyglucose (FDG). After oral glucose loading, the patient was injected with 10 mCi of FDG and 25 mCi of ^99mTc-sestamibi at rest. Dual-isotope single acquisition SPECT was performed approximately 60 min later by positioning two 20 % pulse-height analyzer windows symmetrically around the 140-keV photopeak of ^99mTc and the 511-keV photopeak of FDG. The digital electronics of the camera permitted frame-by-frame decay correction for short-lived FDG. Thus, two separate sets of slices mapping the ^99mTc-sestamibi and FDG distribution were simultaneously obtained, resulting in one-to-one correspondence in spatial registration. Rest ^99mTc-sestamibi images in the horizontal long-axis plane (top row) show reduced perfusion in the apical and lateral regions (arrow). Corresponding FDG images (bottom row) show preserved metabolism in the apical and lateral regions suggestive of viable myocardium (arrow) (From Delbeke et al. [37]; with permission).

Table 2-5. Single-photon emission computed tomography techniques: ^99mTc-labeled sestamibi and ^99mTc-labeled-­tetrofosmin. ^99mTc-sestamibi (isonitrile) and ^99mTc-tetrofosmin are both lipophilic cationic complexes with similar myocardial uptake and blood clearance kinetics. However, the clearance of tetrofosmin from the lungs and the liver is faster than ^99mTc-sestamibi, which may improve the resolution of cardiac images and reduce the overall radiation burden. Both ^99mTc-sestamibi and -tetrofosmin are taken up across the sarcolemmal and mitochondrial membranes of myocytes by passive distribution and are retained within the mitochondria at equilibrium due to a large negative transmembrane potential. Experimental studies with ^99mTc-sestamibi have shown that myocardial uptake and clearance are related to the mitochondrial transmembrane potential and do not differ from ischemic to nonischemic regions. In addition, experimental studies of myocardial infarction, with and without reperfusion, have fueled optimism in the clinical use of ^99mTc-sestamibi for myocardial viability assessment. In the clinical setting, however, with the exception of a few studies, both ^99mTc-sestamibi and -tetrofosmin appear to underestimate myocardial viability. Compared with ²⁰¹Tl and positron emission tomography tracers, factors that may contribute to the impaired ^99mTc-sestamibi or -tetrofosmin accumulation in viable regions at rest include differences in the extraction fraction, blood clearance, redistribution, and response to altered metabolic states. Perhaps a likely improvement in viability assessment with ^99mTc-sestamibi and -tetrofosmin could be achieved through nitrate administration before rest ^99mTc-sestamibi injection and the quantitation of regional radiotracer uptake. SPECT single-photon emission CT.

Table 2-6. Single-photon emission computed tomography techniques: ^99mTc-teboroxime. ^99mTc-teboroxime is a neutral, lipophilic BATO (boronic acid adducts of technetium dioxime) compound with a reported first-pass extraction of 88 % at rest and 91 % under hyperemic conditions. Unlike ^99mTc-sestamibi and -tetrofosmin, clearance of teboroxime from the myocardium is rapid and the washout rate is proportional to blood flow. In experimental studies, approximately two-thirds of the teboroxime activity has been shown to clear from the heart, with a half-life of 3.6 min. Thus, both uptake and clearance of teboroxime from the myocardium are proportional to regional blood flow and are not confounded by tissue metabolism or other binding characteristics within the myocardium. SPECT single-photon emission CT.

PET Tracers and Techniques

Figure 2-31.

Positron emission tomography (PET) techniques: ⁸²Rb. Examples of three-dimensional surface-rendered models of normal (a) and abnormal (b) ⁸²Rb PET myocardial perfusion images are shown with the potential for overlaying coronary anatomic information from hybrid PET/CT angiography.

(c), Although the first few minutes after the infusion of ⁸²Rb are not usually included in clinical acquisition protocols, it is precisely this period that is of interest if myocardial perfusion is to be quantified. Dynamic imaging of the heart during this time allows analysis of the ⁸²Rb concentration in both arterial blood and myocardial tissue as a function of time. ⁸²Rb time–activity curves are shown at rest (a) and after adenosine stress (b). Blue circles represent the activity concentration in the left atrium, and the red circles represent the activity concentration in myocardial tissue. (d) A disparity between myocardial perfusion single-photon emission CT (SPECT) and ⁸²Rb PET studies is shown. Clinically indicated adenosine dual-isotope gated SPECT images (left panel) without attenuation correction show a regional

^99mTc-sestamibi perfusion defect in the anterior and inferior regions (arrows). In the rest ²⁰¹Tl images, the anterior defect became reversible while the inferior defect persisted. Corresponding ⁸²Rb PET myocardial perfusion tomograms performed in the same patient are shown in the right panel. PET images were acquired from a PET/CT scanner after an infusion of adenosine and 30 mCi of ⁸²Rb (top) and at rest following another 30-mCi infusion of ⁸²Rb (bottom). The ⁸²Rb PET images show normal distribution of the radiotracer in all myocardial regions, without evidence for a reversible or fixed defect to suggest myocardial ischemia or infarction. Although the high-energy positrons of ⁸²Rb degrade spatial resolution and the short half-life increases statistical noise, high-quality images free from attenuation artifacts can be produced with ⁸²Rb PET with only a 30-mCi injected dose.

(e) A three-dimensional display of gated ⁸²Rb PET images acquired during pharmacologic stress with adenosine is shown. While gated myocardial perfusion SPECT images are acquired poststress, reflecting regional and global left ventricular function in the resting state, gated ⁸²Rb PET images are acquired during pharmacologic stress as well as at rest. As such, gated ⁸²Rb PET images provide an indirect evaluation of abnormal myocardial perfusion as reflected in regional wall motion abnormalities during stress as opposed to poststress with SPECT. In this patient example, a surface rendering of end-diastolic and end-systolic images from gated adenosine ⁸²Rb PET is shown along with the time–activity curve. The left ventricular ejection fraction is calculated to be 36 % during adenosine ⁸²Rb PET and 28 % at rest. A postexercise gated SPECT left ventricular ejection fraction acquired in the same patient was calculated to be 32 % [41].

Figure 2-32.

Experimental validation of ⁸²Rb for measuring myocardial blood flow. The kinetic model of ⁸²Rb is relatively “simple” since cellular trapping of ⁸²Rb is nonlinearly proportional to blood flow. (a) The initial verification of absolute myocardial perfusion with ⁸²Rb compared with microspheres is shown, measured by epicardial radiation detectors. (b) Coronary flow reserve (CFR) measured by ⁸²Rb positron emission tomography is validated using a “simple” flow model for flow-dependent ⁸²Rb extraction compared with the more complex complete compartmental modeling, both having a comparable correlation with CFR measured by flowmeter ((a) From Goldstein et al. [42]; with permission; (b) From Yoshida et al. [43]).

Figure 2-33.

Severity of anatomic coronary artery stenosis and coronary function. Relative uptake positron emission tomography (PET) myocardial perfusion images are shown for a patient with known coronary artery disease and multiple risk factors; results at baseline (rest and stress) and 6 years later (follow-up stress) are shown. The follow-up PET study shows the progression of the stress-induced perfusion defect in the mid-left anterior descending (LAD) distribution (which was only 65 % of the activity in the adjacent proximal areas of the heart) along with a new perfusion defect in the ramus intermedius or first obtuse marginal branch distribution. The coronary flow reserve (CFR) in the distal LAD is reduced to 1.8 by PET compared with an average of 3.0 in the rest of the heart proximally and compared with 4.0 in healthy young volunteers. A coronary arteriogram shows concentric 57 % mid-LAD stenosis by automated quantitative coronary arteriographic analysis (QCA) and that a fractional flow reserve (FFR) assessed by pressure wire measurements in the aorta and distal to the stenosis is 0.65. In view of concordant low CFR and FFR, a LAD stent was placed with improvement in FFR to 0.92, indicating residual pressure gradient due to diffuse disease proximal to the stent (Adapted from Gould. [44]).

Figure 2-34.

Clinical impact of quantitative myocardial positron emission tomography perfusion imaging. (a) Single views of stress positron emission tomography (PET) relative perfusion images are shown of two different patients at baseline (top row) and follow-up (bottom row), one demonstrating regression and the other progression of coronary artery disease. (b) This schematic of stress PET images at baseline is compared with follow-up PET in various quadrant pairings. Serial changes in PET perfusion images can be used to assess the response to treatment, predict outcomes, and provide insight into the progression or regression of coronary artery lesions and myocardial ischemia (Adapted from Gould [44]).

Figure 2-35.

Prognostic value of MFR by stress ⁸²Rb-PET. In a large cohort of patient referred for assessment of myocardial ischemia with ⁸²Rb-positron emission tomography (PET), the added and independent prognostic value of myocardial blood flow reserve (MFR) was determined beyond the conventional relative myocardium radiotracer uptake on attenuation-corrected PET images. Patients with impaired ⁸²Rb MFR had a higher incidence of hard and major adverse cardiac events (MACE) (cardiac death, nonfatal myocardial infarction, late revascularization, or hospitalization for cardiac reasons) at approximately 1-year follow-up. In the multivariable model analysis, ⁸²Rb MFR was an independent predictor of hard events (a) and MACE (b) over the summed stress score (SSS). MACE within subgroups of SSS for different levels of MFR is shown in (c). At any level of SSS, the prevalence of MACE is higher in patients with the lowest MFR (<1.5) and statistically significantly different compared with MFR greater than or equal to 2 among patients with overt ischemia (Adapted from Ziadi et al. [45]).

Figure 2-36.

Positron emission tomography techniques: ¹⁵O-water. ¹⁵O-water is a freely diffusible tracer that correlates closely with perfusion as assessed by microspheres with a first-pass extraction fraction approaching unity. Because ¹⁵O-water is both in the vascular space and myocardium, visualization of myocardial activity requires correction for activity in the vascular compartment. (a) Corrections for activity in the left atrium (top left), left ventricle (top right), and thoracic aorta (bottom left) are shown in a healthy patient after inhalation of 30–40 mCi of ¹⁵O-carbon monoxide, which labels erythrocytes in vivo. The distribution of ¹⁵O-water is shown in the left ventricular myocardium (bottom right) after correcting for vascular space. (b) The ability of ¹⁵O-water to assess myocardial viability through modification of the blood flow information is shown. This method, termed water perfusable tissue index (PTI), is based on a measurement of perfusable tissue fraction (PTF) as a method to correct for the partial volume effects in ¹⁵O-water studies [46]. PTF is defined as the fractional volume of a given region of interest occupied by myocardium that is capable of exchanging water rapidly. Using transmission and ¹⁵O-blood pool images, the anatomic tissue fraction (ATF), a quantitative estimate of extravascular tissue density, is derived. The ratio of PTF to ATF thus represents the proportion of the extravascular tissue that is perfusable by ¹⁵O-water. Because water can freely exchange across all normal tissue cells, the PTF should approach unity in normal myocardium and be reduced in scarred myocardium.

A myocardial region of interest containing a mixture of ¹⁵O-water perfusable and nonperfusable tissue is diagrammed. The volume of the region of interest is shown (A). ATF for the region of interest is produced by subtracting the blood pool (¹⁵O-carbon monoxide) from the transmission images after normalizing the latter to tissue density (1.04 g/mL). The total ATF (B) represents the total extravascular tissue and contains both perfusable and nonperfusable tissue components. The ¹⁵O-water PTF for the region of interest calculated from the ¹⁵O-water data set identifies the mass of tissue within the region of interest that is capable of rapid trans-sarcolemmal exchange of water. Note that the nonperfusable or necrotic region is excluded from this parameter. The ¹⁵O-water PTI is calculated by dividing ¹⁵O-water PTF (C) by the total ATF (B) and represents the fraction of the total anatomic tissue that is perfusable by water. ANT anterior, LAT lateral, SEP septal ((a) From Bergmann et al. [47]; with permission; (b) Adapted from Yamamoto et al. [48]).

Figure 2-37.

Positron emission tomography techniques: ¹³N-ammonia. ¹³N-ammonia is the extractable perfusion tracer most commonly used with positron emission tomography (PET). At physiologic pH, ammonia is in its cationic form with a physical half-life of 10 min. Myocardial distribution of ammonia is related inversely and nonlinearly to blood flow. Although the exact mechanism of ¹³N-ammonia transport across the myocardial membrane has not been conclusively established, it has been suggested that ¹³N-ammonia may cross cell membranes by passive diffusion or as ammonium ion (¹³NH+₄) by the active sodium–potassium transport mechanism influenced by the concentration gradient across the cell membrane. Once in the myocyte, myocardial retention of ¹³N-ammonia involves predominantly the conversion of ¹³N-ammonia and glutamic acid to ¹³N-labeled glutamine mediated by ATP and glutamine synthetase. Hence, absolute quantification requires two- and three-compartment kinetic models that incorporate both extraction and retention rate constants. Quantification of ammonia is further complicated by the rapid degradation of ammonia, which occurs within 5 min after administration, producing metabolic intermediates such as urea and glutamine that are also extracted by the heart. Experimental studies suggest that the myocardial uptake of ammonia reflects absolute blood flows up to 2–2.5 mL/g/min and plateaus at flows in the hyperemic range. In the clinical setting, 10–20 mCi of ¹³N-ammonia is administered intravenously.

Figure 2-38.

¹³N-ammonia positron emission tomography and coronary angiography in a patient with coronary artery disease. (a) These stress and rest ¹³N-ammonia positron emission tomography images of the heart in short-axis, vertical long-axis, and horizontal long-axis slices are shown from a 62-year-old patient who has type 2 diabetics. The stress images demonstrate a moderately decreased perfusion defect involving the lateral extending to the inferolateral region of the left ventricle, which is completely reversible in rest images. (b) This corresponding coronary angiography shows an occluded marginal branch of the left circumflex artery (left arrow), with diffuse 50 % stenosis of the proximal LAD (right arrow) and a 50 % stenosis in the mid-RCA (not shown) (Adapted from Schindler et al. [40]).

Figure 2-39.

Mechanism of ¹³N-ammonia uptake. The interplay between blood flow and metabolism in the extraction and retention of ¹³N-ammonia is complex. The early extraction phase of freely diffusible ¹³N-ammonia reflects blood flow while the later, slow turnover phase reflects the metabolic trapping of ¹³N-ammonia. In experimental animals, several investigators have shown that the myocardial extraction and retention of ¹³N-ammonia are related not only to regional blood flow but also to myocardial oxygenation and metabolism. Under hypoxic or ischemic conditions, the reduction of intracellular ATP to concentrations in the range of the K_m for the enzyme–ATP complex could reduce intracellular ¹³N-ammonia metabolism by glutamine synthetase. Because the extent of ¹³N-ammonia metabolism may depend on the ATP state of the myocyte, intracellular levels of ¹³N-ammonia may reflect cellular viability.

(a) In patients with chronic coronary artery disease and left ventricular dysfunction, receiver-operating characteristic (ROC) curves were used to compare the abilities of late ammonia uptake (final 10–15 min of image acquisition) and absolute blood flow (early extraction phase, approximately 3 min after injection) to predict the functional improvement of asynergic regions after revascularization. The results show that late ammonia uptake (metabolic trapping) is a significantly better predictor of functional improvement after revascularization when compared to absolute blood flow. (b) There is a linear relationship between percent late ammonia uptake and [¹⁸F]-fluorodeoxyglucose (FDG) uptake (left) and between ²⁰¹Tl uptake on redistribution imaging (right) in reversible (orange circles) and irreversible (blue circles) asynergic regions after revascularization. (c) Sequence and timing of ¹³N-ammonia and FDG positron emission tomography imaging for the assessment of myocardial viability. Thus, beyond ammonia’s value as a perfusion tracer, late ammonia images provide important insight regarding cell membrane integrity and myocardial viability (Adapted from Kitsiou et al. [49]).

Table 2-7. Positron emission tomography (PET) myocardial perfusion tracers and image acquisition. PET approaches for the assessment of regional myocardial blood flow in ml/g/min entail the intravenous injection of a positron-emitting perfusion tracer, such as ⁸²Rubidium, ¹³N-ammonia, or ¹⁵O-water, and the dynamic acquisition of images of the radiotracer passing through the central circulatory system to its extraction and retention in the left ventricular myocardium. Tracer kinetic models (one to three compartments) and operational equations are then applied to correct for the physical decay of the radioisotope, partial volume-related underestimation of the true myocardial tissue concentrations by assuming a uniform myocardial wall thickness of 1 cm, and spillover of radioactivity between the left ventricular blood pool and myocardium, to yield regional myocardial blood flows in absolute terms, ml/g/min. The relative distribution of the radiotracer in the ­myocardium can also be assessed visually or semiquantitatively (as percent uptake relative to a reference region) from the final static image of the myocardium, obtained from the last (e.g., 900 s) frame of the PET image series, which can be displayed as a polar map. 2D two-­dimensional, 3D three-dimensional [38] (Adapted from Schindler et al. [39]).

Table 2-8. Identification of flow-limiting coronary artery lesions by positron emission tomography (PET). Similar to single-photon emission CT (SPECT), the identification of stress-induced scintigraphic perfusion defects by positron emission tomography PET imaging provides important diagnostic and prognostic information. However, unlike SPECT imaging, soft tissue attenuation correction with PET imaging is reliable and accurate. This accurate attenuation correction in concert with the higher spatial resolution may explain a 10 % higher diagnostic accuracy of PET when compared to conventional SPECT imaging for the detection of flow-limiting coronary artery lesions. The advantages of PET imaging, however, pertain not only to the high spatial and depth-independent resolution but also to the ability to quantify the radiotracer uptake in the myocardial tissue and to assess rapid alterations of radiotracer activity concentrations in the arterial blood and myocardium owing to a high temporal resolution in seconds. The latter advantages of PET imaging combined with tracer kinetic compartment models afford the noninvasive assessment of myocardial blood flow in absolute terms (Adapted from Schindler et al. [40]).

Table 2-9. Positron emission tomography techniques: ⁸²Rb. ⁸²Rb is a generator-produced, short-lived, positron-emitting cation with biologic properties that are similar to potassium and ²⁰¹Tl. As with potassium and ²⁰¹Tl, the intracellular uptake of ⁸²Rb across the sarcolemmal membrane reflects active cation transport via the Na–K ATPase transport system. In patients with chronic coronary artery disease, myocardial uptake of ⁸²Rb is preserved in viable regions and is severely reduced in scarred regions. In the setting of acute myocardial injury and reperfusion, the initial uptake of ⁸²Rb reflects blood flow.

Figure 2-40.

Predictive value of major adverse cardiac events including cardiac death by stress ¹³N-ammonia positron emission tomography. The improved diagnostic accuracy of positron emission tomography (PET) over single-photon emission computed tomography for detecting coronary artery disease (CAD) can be attributed to the higher photon energy of PET radiotracers, as well as to the improved resolution and intrinsic attenuation correction of PET cameras. The latter has enabled PET to assess absolute myocardial blood flow, from which coronary flow reserve (CFR) can be quantified noninvasively. However, outcome data with PET in patients with CAD are scarce. In patients with suspected myocardial ischemia who underwent ¹³N-ammonia PET, the long-term prognostic value of CFR was assessed over a mean of 5.4 years. Major adverse cardiac events (MACE) (cardiac death, nonfatal myocardial infarction, late revascularization, or hospitalization for cardiac reasons) occurred in 34 % of the patients. Abnormal PET perfusion was associated with a higher incidence of MACE (P  <  0.001) and cardiac death (P  <  0.05). In patients with normal perfusion, abnormal CFR was independently associated with a higher annual event rate over 3 years compared with normal CFR for MACE (1.4 vs. 6.3 %; P  <  0.05) (a) and cardiac death (0.5 vs. 3.1 %; P  <  0.05) (b). This is reflected by the higher annual event rate (%/year) in abnormal CFR (c). In patients exhibiting abnormal perfusion, CFR remained predictive throughout the 10-year follow-up (P  <  0.001). Beyond detection of CAD, ¹³N-ammonia PET perfusion and CFR data are strong predictors of adverse outcome. CFR is an independent predictor of adverse outcome. It provides a 3-year “warranty” period of event-free survival for patients with normal CFR and normal PET perfusion. Conversely, in patients with abnormal perfusion, an impaired CFR has added value for predicting adverse outcomes (Adapted from Herzog et al. [50]).