Chapter 3: Radiopharmaceuticals for Cardiac Imaging

This chapter reviews the most important characteristics of the various single-photon emission computed tomography (SPECT) and positron emission tomography (PET) radiopharmaceuticals that are currently used in clinical practice. Myocardial perfusion scintigraphy has significantly evolved¹ since its introduction more than four decades ago. Three major factors have specifically contributed to this evolution: (1) technical improvements in scintigraphic data acquisition and analysis, (2) introduction of new technetium-99m (Tc-99m)-labeled MPI radiopharmaceuticals with different properties than thallium-201 (Tl-201), and (3) availability of PET radiotracers and PET-dedicated cameras. This chapter reviews the most important characteristics of the various SPECT and PET MPI radioactive agents that are currently used in clinical practice. The knowledge of the characteristics and kinetics of the various radiotracers is clinically relevant and essential for designing and implementing optimized imaging protocols.² The optimal timing of imaging after injection either at stress or at rest is determined by rate of uptake in the heart and adjacent organs, as well as the residence time of radiotracers within the myocytes. The efficiency of myocardial extraction over a wide range myocardial blood flows is relevant for reliable detection of obstructive coronary artery disease and absolute quantification of regional myocardial blood flow. Therefore, knowledge of the basic characteristics of MPI radiotracers is essential for optimal clinical use of these agents. In addition, a section on ^99mTc-pyrophosphate cardiac imaging is included, given its role in the diagnosis of cardiac amyloidosis, while other more specific agents such as ¹²³I-MIBG and ¹²³I-IPPA are covered in another chapter (Chapter 23). Radiation safety has become a more sensitive concern more recently. Therefore, data on radiation dosimetry are also discussed for each radiopharmaceutical.

General Physiologic Characteristics of Perfusion Imaging Agents

Although many different classes of radioactive MPI agents exist, they should all present a minimum of common basic characteristics.⁴

1. The myocardial uptake of the radiotracer must be proportional to the regional myocardial blood flow over a relatively wide range of blood flows.

2. The myocardial uptake should be high enough to allow for detection of regional inhomogeneity by external gamma scintigraphy.

3. The initial myocardial distribution of the radiotracer at the time of injection must remain stable during the acquisition time of the images.

4. The effect of blood flow on myocardial transport of the radiotracer must be predominant to the effect of metabolic cellular alterations.

5. Finally, the agent should be labeled to a radionuclide having adequate physical characteristics to provide high photon flux and optimal counting statistics.

Information on basic properties of radionuclide MPI agents is generally obtained from cultured myocardial cells, isolated perfused hearts, or in vivo animal models.⁵ Precise measurements of cellular or capillary-tissue tracer kinetics are usually obtained from cell cultures and isolated perfused heart models, whereas regional tracer distribution and uptake in other organs are studied with in vivo animal models. The two most important physiologic factors that affect the myocardial uptake of an MPI agent are the variations in regional myocardial blood flow and the myocardial extraction of the radiotracer. Three major parameters (Enet, Emax, and PS cap) can be determined using indicator-dilution techniques and radiolabeled albumin used as an intravascular reference. The difference between the intravascular albumin reference and the radiopharmaceutical that is evaluated (i.e., a diffusible perfusion agent) on a venous dilution curve is used to calculate its instantaneous cardiac extraction. The early peak of the curve, or Emax, represents the maximum fractional tissue extraction of the diffusible agent. This value is used to calculate the capillary permeability surface area product, PS cap. The net extraction (Enet) is the integral of the curve and is used as a measure of myocardial radiotracer retention, including both initial extraction and subsequent back-diffusion. A high value for Emax and PS cap indicates a rapid blood-tissue exchange and suggests that the diffusible radiotracer will be able to assess high levels of hyperemic flow accurately. Table 3-1 summarizes and compares these values for the commonly used radiopharmaceuticals.

The criteria necessary for a radiopharmaceutical to be used to determine regional perfusion were elaborated by Saperstein in the mid-1950s.³ As described by the Saperstein principle, a given type of radiopharmaceutical will be distributed in proportion to regional perfusion if its extraction by the organ of interest is high and if its clearance from the blood is rapid. One of the first classes of radiopharmaceuticals used to image myocardial perfusion was the potassium analogs—cesium, rubidium, and thallium, which enter the myocardium by the sodium–potassium–ATPase pump mechanism. Although different radioisotopes of these cations have been evaluated, Tl-201 was found to have the best physical and biological characteristics for imaging and has been the most popular radionuclide MPI agent used in noninvasive detection and evaluation of patients with known or suspected coronary artery disease for more than two decades.

Thallium-201 is a cyclotron-produced monovalent cation with a physical half-life of 73.1 hours. It decays by electron capture to mercury-201. Thallium-201 is a low-energy gamma emitter with principal photopeaks at 135.3 keV (2.7% abundance) and 167.4 keV (10% abundance) and x-rays emitted from the mercury daughter at 68 to 80.3 keV (95% abundance). Some contaminants such as Tl-200 (<0.3%), Tl-202 (<1.2%), and Pb-202 (<0.2%) can be present, but that usually represents less than 2% of the total Tl-201 activity at the time of calibration. Contrary to the Tc-99m-labeled MPI agents, there is no in-house preparation and no quality control procedure required before injection in humans.

Physiologic Characteristics

The myocardial transmicrovascular transport of Tl-201 was evaluated by Leppo and Meerdink⁶ in a blood-perfused, isolated rabbit heart model. The averaged myocardial extraction (Emax) of Tl-201 was 0.73 ± 0.10. The net myocardial extraction measured over a 2- to 5-minute period was 0.57 ± 0.13. The mean PS cap of Tl-201 was 1.30 ± 0.45 mL/g/min. Although Tl-201 is considered biologically similar to potassium, its myocardial uptake is greater than that of potassium. The myocardial extraction fraction of Tl-201 in in vivo model is approximately 87% at normal flow rates.⁷ Studies on the relationship of myocardial uptake of Tl-201 to regional coronary blood flow as determined by radiolabeled microspheres showed that there is a nearly linear correlation over a wide range of coronary blood flows (Fig. 3-1). However, at high flow rates (at approximately 2.5 mL/min/g), there is a diffusion limitation and a fall in tracer extraction fraction and also an increased cellular washout, resulting in a plateau in myocardial uptake. This results in an overall underestimation of higher levels of coronary blood flows. Conversely, at very low flow rates (usually less than 10% of the baseline blood flow), there is an increased myocardial extraction fraction of Tl-201 relative to blood flow, resulting in an overestimation of coronary blood flow.

One of the most clinically important characteristics of Tl-201 is its myocardial redistribution. Myocardial uptake of Tl-201 is not static over time. This property forms the basis of the stress-redistribution imaging protocol used to diagnose the presence of coronary artery disease with Tl-201. The redistribution or the filling in of a myocardial perfusion defect generally occurs between 3 and 5 hours after the injection of Tl-201 at peak stress and is related to two factors: the rate of influx of Tl-201 into the myocardium from whole-body blood pool activity and the rate of clearance or washout of Tl-201 from the myocardium. A myocardial perfusion defect related to ischemic disease will show a normalization of the Tl-201 uptake at 3 to 4 hours after the injection at stress because of a delayed accumulation into the ischemic segment and a more rapid washout from normal myocardial segments than from hypoperfused segments.

Biodistribution and Dosimetry

The greatest concentration of Tl-201 is found in kidneys, heart, and liver. The activity in these organs remains high for few hours after the injection. Maximal myocardial uptake, which is approximately 3.7% to 4.0% of the injected dose, is achieved by 10 to 15 minutes after the administration of the radiopharmaceutical. Tl-201 myocardial uptake has two components: an early component with a half-life of approximately 4 hours (80%) and a delayed component with a half-life of 40 hours (20%). Disappearance of Tl-201 from the blood compartment is rapid with two components: 92% of the blood activity disappears with a half-life of 5 minutes, and the other 8% will be cleared from the blood with a half-life of approximately 40 hours. This rapid clearance of Tl-201 from the blood results in a decreased blood pool background when MPI is performed. Table 3-2 summarizes the radiation dose estimates for Tl-201 along with the two Tc-99m-labeled MPI agents. Kidneys are the target organs.⁸ Based on the dosimetric data, the recommended adult dose of intravenous Tl-201 for SPECT MPI is 74 to 111 MBq (2–3 mCi), although several authors are using doses up to 4.0 mCi.

Thallium-201 was the major radiopharmaceutical in nuclear cardiology for almost two decades. Although hundreds of studies have demonstrated the clinical value of Tl-201 myocardial perfusion scintigraphy, the physical characteristics of this radionuclide are suboptimal for scintillation camera imaging. Therefore, in the late 1970s and early 1980s, many investigators attempted to develop an MPI agent labeled with Tc-99m to circumvent the physical limitations of Tl-201. The potential advantages of a Tc-99m-labeled agent over Tl-201 are significant and include the following:

1. The 140-keV photon energy of Tc-99m, which is optimal for standard gamma camera imaging, results in an improved resolution due to less Compton scatter and less tissue attenuation in the patient (in comparison to the low photon energy of 68–80 keV for Tl-201).

2. The much shorter physical half-life of Tc-99m (6 hours vs. 73 hours for Tl-201) and the better radiation dosimetry permit the administration of a 10 times higher dose of a Tc-99m-labeled compound than Tl-201. This yields better image quality, and images can be performed in a shorter time period.

3. The resulting overall better counting statistics of Tc-99m allows for the perfusion images to be optimally obtained in a gated mode. Simultaneous assessment of perfusion and function (global and regional wall motion) can thus be obtained.

4. It is possible to perform first-pass function studies (if the initial lung transit is rapid enough) with a Tc-99m-labeled radiopharmaceutical agent.

5. Because Tc-99m is constantly available from a molybdenum generator in a nuclear medicine laboratory, special deliveries from a distribution center or a commercial radiopharmacy are not required. A Tc-99m-labeled MPI agent can thus be available almost 24 hours a day.

For the above reasons, Tc-99m-labeled agents have enormous potential to assess myocardial perfusion and could be very useful clinically.^9–11

Tc-99m-Sestamibi

The first member of the Tc-99m-isonitrile family to be evaluated in humans was the hexakis (t-butyl-isonitrile)-technetium (I), also known as Tc-99m-TBI.^12,13 Although the myocardial uptake of Tc-99m-TBI was proportional to myocardial blood flow and was satisfactory for imaging purposes, its routine clinical use was limited by an increased lung uptake and prominent and persistent liver uptake. The initial lung uptake and subsequent washout of Tc-99m-TBI from the lungs also created significant imaging problems. A subsequently developed Tc-99m-isonitrile compound emerged from intensive search,^14–16 initially known by the coded name of RP-30A (nonlyophilized form) or RP-30 (lyophilized form), and then as either Tc-99m-hexakis 2-methoxyisobutyl isonitrile, Tc-99m-hexakis-2-methoxy-2-methylpropyl-isonitrile, Tc-99m-hexamibi, or Tc-99m-MIBI, and currently by its generic name Tc-99m-sestamibi or Cardiolite™. Tc-99m-sestamibi has favorable biological characteristics for clinical applications with transient liver uptake and rapid hepatobiliary excretion, along with minimal lung uptake. Tc-99m-sestamibi was approved by the U.S. Food and Drug Administration for clinical application in December 1990.

Intravenous injections of Tc-99m-sestamibi have been associated with very few adverse reactions. According to the product monograph, during clinical trials (Phase III study) approximately 5% to 10% of patients experienced transient parosmia and/or taste perversion (metallic or bitter taste) occurring a few seconds after the injection. Usually, this side effect disappears within 15 to 30 seconds. This adverse effect seems to be related to the presence of the copper salt in the kit formulation, and its incidence may be related to the concentration of Tc-99m-sestamibi used.

Physiologic Characteristics

Tc-99m-sestamibi is a cationic complex that is taken up by myocytes in proportion to regional myocardial blood flow. The cationic charge of the compound provides hydrophilic properties, and the six isonitrile groups allow hydrophobic interaction with cell membranes.

Tc-99m-sestamibi, which is retained within cells because of the negative charge generated on the mitochondria, has a high affinity for cytoplasm and shows very little extracellular exchange. Thus, metabolic derangements affecting myocytes' viability would also result in decreased Tc-99m-sestamibi uptake, independently of myocardial blood flow. Using aerobic metabolic blockade, Beanlands et al.¹⁷ showed that an irreversible cellular injury resulted in a marked increase in Tc-99m-sestamibi clearance rate. They concluded that the accumulation and clearance kinetics of Tc-99m-sestamibi were dependent on sarcolemmal integrity and on aerobic metabolism and were significantly affected by cell viability.

Okada et al.¹⁸ investigated the myocardial kinetics of Tc-99m-sestamibi in dogs undergoing a partial occlusion of the left circumflex coronary artery. They showed that Tc-99m-sestamibi was rapidly taken up by nonischemic and ischemic myocardium at rest in proportion to regional myocardial blood flow. There was a good correlation between the initial myocardial flow at normal resting flow rates and the Tc-99m-sestamibi myocardial distribution. Another study from the same group of investigators¹⁹ using the same animal model evaluated the myocardial kinetics of Tc-99m-sestamibi after pharmacologic vasodilation with dipyridamole and showed that the initial myocardial uptake was closely related in a linear fashion to the regional myocardial blood flow at rates up to approximately 2.0 mL/min/g. However, at higher flow rates, there is a plateau in the myocardial distribution versus flow curve, resulting in an underestimation of coronary blood flow. Similar findings have been reported other investigators, which parallel the kinetics of thallium-201.^20,21 There is overestimation of myocardial blood flow at low flows which is probably related to an increased extraction seen with diffusible indicators. Thus myocardial uptake of Tc-99m-sestamibi is proportional to regional myocardial blood flow over the physiological flow range with decreased extraction at hyperemic flows and increased extraction at low flows.

In contrast to Tl-201, Tc-99m-sestamibi shows a very slow myocardial clearance after its initial myocardial uptake. A fractional Tc-99m-sestamibi clearance of 10% to 15% over a period of 4 hours has been measured by Okada et al.¹⁸ in a canine model of partial coronary occlusion. The clearance was similar in the hypoperfused and normal zones. Animal studies have shown that after Tc-99m-sestamibi injection during brief periods (6–15 minutes) of coronary occlusion in dogs, the occluded zone shows a continued myocardial uptake during the reperfusion phase. Thus, following transient ischemia and reperfusion, there is some degree of myocardial redistribution of Tc-99m-sestamibi, although it is slower and less complete than Tl-201.²²

Sinusas et al.²³ studied the myocardial uptake of Tc-99m-sestamibi and Tl-201 showed that as long as myocardial cells were still viable, the myocardial uptake of the tracer was not affected by an ischemia. These data suggest that the agent can also assess myocardial viability. Tc-99m-sestamibi uptake is maintained in viable myocardium but reduced in necrotic tissue. Using a dog model with coronary occlusion and reperfusion, Verani et al.²⁴ demonstrated that the size of the perfusion defect during occlusion as detected by scintigraphic images correlated with the amount of myocardium supplied by the occluded vessel, the area at risk. A smaller perfusion defect was detected on Tc-99m-sestamibi imaging during reperfusion. This defect correlated with the amount of infarcted myocardium and the area showing improved perfusion pattern after reflow represented the salvaged myocardium.

Biodistribution and Dosimetry

The blood clearance, biodistribution, dosimetry, and safety of Tc-99m-sestamibi were initially reported by Wackers et al.¹⁵ Both rest and stress blood clearance curves approximate a dual exponential curve with an initial fast and later slow component. The maximal activity at rest was noted at 1 minute after injection (36 ± 18% of injected dose), and the maximal activity after injection during exercise was measured at 0.5 minute. At 1 hour after the intravenous injection of Tc-99m-sestamibi, the blood pool activity progressively decreased to 1.10 ± 0.01% and 0.7 ± 0.1% of the injected dose at rest and after stress, respectively. At 60 minutes after the injection at rest, the uptake in the heart was 1.0 ± 0.4% of injected dose. The 24-hour urinary excretion was 29.5% of injected dose, whereas the 48-hour fecal excretion was 36.9% of injected dose. The study of the upper-body organ distribution showed that the highest initial Tc-99m-sestamibi concentration (counts/pixel) is found in the gallbladder and liver followed by heart, spleen, and lungs. The myocardial activity remained relatively stable over time (27 ± 4% of initial activity has cleared from the heart at 3 hours), whereas activity in the spleen and lung decreased gradually. The maximal accumulation in the gallbladder occurred approximately at 60 minutes after the injection. Similar biodistribution and radiation dosimetry was noted after exercise.

Radiation dose estimates for Tc-99m-sestamibi have been evaluated from whole-body images.¹⁵ The estimated radiation absorbed dose at rest and at stress, assuming a 2.0-hour void, are summarized in Table 3-2. The uptake in the heart is 1.06 ± 0.4% of injected dose at 60 minutes after injection at rest and 1.4 ± 0.3% at 60 minutes for the stress study. The upper large intestine wall receives the highest dose of radioactivity, both at rest and at stress. To decrease dosimetry to the urinary bladder, increasing voiding frequency should be encouraged. If patients are administered a total dose of 30 mCi of Tc-99m-sestamibi, no individual organ dose will exceed 5 rads (50 mGy). Although there is accumulation of Tc-99m-sestamibi in the mammary glands, there is a minimal transfer into milk: approximately 0.01% to 0.03% of the injected Tc-99m-sestamibi activity can be excreted in human breast milk of a breastfeeding patient.²⁵

Tc-99m-Tetrofosmin

Tc-99m-tetrofosmin, a diphosphine complex of Tc-99m, was the third Tc-99m-labeled MPI FDA-approved agent, following Tc-99m-teboroxime (no longer clinically available) and Tc-99m-sestamibi. Tc-99m-tetrofosmin shows similar myocardial uptake, retention, and blood clearance kinetics to Tc-99m-sestamibi. However, its clearance from both the liver and the lung is faster than that of Tc-99m-sestamibi. These characteristics can have an impact on the injection and imaging protocols. Tetrofosmin is a ligand that forms a lipophilic, cationic complex with Tc-99m. Tc-99m-tetrofosmin is the generic name for 1,2,-bis [bis(2-ethoxyethyl) phosphino] ethane, also called p53 or Myoview^TM. Tetrofosmin has a molecular weight of 382, and an empirical formula of C₁₈H₄₀O₄P₂. The functionalized diphosphine complex of Tc-99m has a molecular weight of 895 and a formula of [TcO₂ (tetrofosmin)₂].¹ There are no known contraindications to intravenous administration.

Physiologic Characteristics

Sinusas et al.²⁶ tested the hypothesis that Tc-99m-tetrofosmin was a reliable coronary blood flow tracer over a physiologic range of flows seen in ischemia or infarction conditions. Dogs were injected with 30 mCi Tc-99m-tetrofosmin during peak pharmacological stress performed with either adenosine or dipyridamole. Myocardial Tc-99m-tetrofosmin activity at 15 minutes after the injection correlated linearly with radiolabeled microsphere flow during peak stress in each dog. Myocardial Tc-99m-tetrofosmin activity appeared to underestimate flow at flows exceeding 1.5 to 2.0 mL/min/g. The plot of Tc-99m-tetrofosmin activity versus blood flow achieved a plateau at approximately 2.0 mL/min/g (see Fig. 3-1). On the other hand, as with Tc-99m-sestamibi and Tl-201, Tc-99m-tetrofosmin activity overestimated coronary blood flow in low flow ranges, at less than 0.2 mL/min/g. Tc-99m-tetrofosmin activity cleared rapidly from the blood with 2.8% and 0.8% of peak activity remaining in the blood at 5 and 15 minutes, respectively. The myocardial clearance between 3 and 15 minutes was similar in both ischemic and nonischemic regions. The myocardial activity cleared 18 ± 11% in the ischemic region. Lung activity remained lower than myocardial activity and the liver activity remained elevated over the initial 15-minute period following injection.

Mechanisms of Tc-99m-tetrofosmin myocardial uptake have been studied using different experimental models. Dahlberg and Leppo²⁷ evaluated the effect of coronary blood flow on the uptake of Tc-99m-tetrofosmin in the isolated rabbit heart model. The Emax of 0.37 for Tc-99m-tetrofosmin suggests a PS cap similar to that of Tc-99m-sestamibi. In comparison, Emax for Tl-201 is 0.73 and for Tc-99m-sestamibi is 0.39. However, Tc-99m-tetrofosmin has the lowest Enet among all perfusion agents. Although this lower value of Enet for Tc-99m-tetrofosmin in rabbits suggests myocardial clearance of this compound, human studies²⁸ have shown a stable myocardial retention of Tc-99m-tetrofosmin, at least up to 4 hours. This difference between animal and human data is not really surprising, considering that similar interspecies variability has been previously observed for the kinetics of other Tc-99m-labeled phosphine compounds.²⁹ Platts et al.³⁰ found that the uptake of Tc-99m-tetrofosmin into rat myocytes was rapid, temperature-dependent and independent of extracellular Tc-99m-tetrofosmin concentration. The lack of effect of ion channel inhibitors on Tc-99m-tetrofosmin uptake is similar to that on uptake of other cations such as Tc-99m-sestamibi. Thus, Tc-99m-tetrofosmin differs from Tl-201 in that it does not appear to act as potassium analog.

Based on experimental studies, mitochondrial membrane potential appears to play a major role in the myocardial uptake and retention of Tc-99m-tetrofosmin, as seen with Tc-99m-sestamibi. Younes et al.³¹ demonstrated that Tc-tetrofosmin uptake into myocytes is likely the potential-driven transport of the lipophilic cation and myocardial uptake in vivo was related to the metabolic status of the myocytes.

Biodistribution and Dosimetry

Human biodistribution, dosimetry, and safety of Tc-99m-tetrofosmin administration at rest and during exercise were by Higley et al.³² By 10 minutes after the injection, there was less than 5% of the injected dose in the whole blood volume and less than 3.5% of the injected dose in the total plasma volume. The blood clearance was initially faster following exercise. At 2 hours after injection, the urinary clearance was 13.1 ± 2.1% in the resting study and 8.9 ± 1.7% in the exercise study (p < 0.001). At 48 hours postinjection, the rate of urinary clearance was almost identical for both physiological conditions: 39.0 ± 3.7% at rest and 40.0 ± 3.7% at exercise. Analysis of whole-body images showed that good quality images of the heart can be obtained as early as 5 minutes after the injection of Tc-99m-tetrofosmin, and this uptake persisted for several hours. Myocardial background clearance resulting from activity in the blood, liver, and lung was rapid. After exercise, there was less Tc-99m-tetrofosmin activity in certain organs, mainly liver, urinary bladder, and salivary glands, in comparison to the rest study. As with Tc-99m-sestamibi, this relatively reduced liver uptake at stress can be explained by an enhanced retention in peripheral muscles as a result of the increased blood flow induced by physical exercise and by splanchnic vasoconstriction during exercise.

After a stress injection, the myocardial uptake of Tc-99m-tetrofosmin, although relatively stable over time, slightly decreases from 1.3% of the injected dose at 5 minutes to 1.0% at 2 hours after the injection. From 5 to 60 minutes after injection, the heart-to-lung ratio increases from 4.0 ± 1.1 to 5.9 ± 1.3 and the heart-to-liver ratio increases from 0.8 ± 0.3 to 3.1 ± 3.0. After a rest injection, similar kinetics were noted, and the heart-to-lung ratio increases from 3.1 ± 1.8 to 7.3 ± 4.4, and the heart-to-liver ratio increases from 0.4 ± 0.1 to 1.2 ± 0.8 from 5 to 60 minutes after injection.

Sridhara et al.³³ compared Tc-99m-tetrofosmin and Tl-201 myocardial imaging in patients with documented coronary artery disease and showed that there was no significant Tc-99m-tetrofosmin myocardial redistribution with a slow myocardial washout of approximately 4% to 5% per hour after exercise and 0.4% to 0.6% per hour after a rest injection. The estimated absorbed radiation doses at rest and at stress are given in Table 3-2. The results show that both at rest and at stress, the gallbladder wall is the target organ, followed by the other excretory organs, such as upper large intestine, lower large intestine, bladder wall, and small intestine. Overall, the radiation dose to most organs is significantly reduced during exercise in comparison to rest study.

^99mTc-Pyrophosphate

Cardiac amyloidosis, a significantly underdiagnosed cause of heart failure, is usually detected late when the heart is already significantly affected. Differentiating immunoglobulin light-chain (AL) from transthyretin-related (ATTR) cardiac amyloidosis is important given the implications for prognosis, therapy, and genetic counseling.³⁴

During the last two decades, noninvasive detection of cardiac amyloidosis using nuclear medicine techniques has gained in popularity.^{35–38 99m}Tc-labeled phosphate derivatives, initially developed as bone-seeking radiotracers for bone scintigraphy, were noted to localize to amyloid deposits. Different agents were used to detect myocyte necrosis and calcifications in amyloid deposits (with localized increased tissue calcium deposits) such as ^99mTc-diphosphonate, ^99mTc-pyrophosphate, ^99mTc-MDP (methylene-diphosphonate) and ^99mTc-DPD (diphosphino-propanodicarboxylic acid). The latter radiotracer is extensively used in European countries but is not yet approved by the FDA. Therefore, ^99mTc-pyrophosphate (PYP), approved since more than 30 years for bone scintigraphy, blood pool imaging and detection of myocardial infarction, is currently used in clinical practice. The precise mechanism by which ^99mTc-PYP (and the other bone-seeking radiotracers) accumulates in the myocardium of patients with cardiac amyloidosis remains unclear but is probably related to high calcium levels in amyloidosis. Different studies have shown that ^99mTc-PYP cardiac imaging can distinguish ATTR from AL amyloidosis possibly because ^99mTc-PYP may bind TTR amyloid fibrils more intensely than AL fibrils as a result of higher calcium containing substances in ATTR hearts. Using quantitative ^99mTc-PYP cardiac imaging, Bokhari et al³⁹ were able to differentiate light-chain cardiac amyloidosis from the TTR-related familial and senile cardiac amyloidosis. Using a heart-to-contralateral uptake ratio of more than 1.5, they showed a sensitivity of 97% and a specificity of 100% for identifying ATTR cardiac amyloidosis.

Two hours after intravenous injection of ^99mTc-PYP, 40% to 50% of the injected dose is taken up by the skeleton. Within a period of 1 hour, 10% remains in the vascular system. The average urinary excretion is about 40% of the administered dose after 24 hours. The usual injected dose for cardiac scintigraphy is 15 mCi (555 MBq). The target organs are the bladder (1.46 rads/15 mCi with a 2-hour void and 3.45 rads/mCi with a 4-hour void) and the kidneys (2.1 rads/15 mCi).

There is general consensus that PET myocardial perfusion imaging is of consistently of better quality than SPECT imaging, particularly in obese patients. Attenuation correction (currently often with x-ray computed tomography) is a standard component of PET imaging. Only two PET radiotracers are currently used in North America for PET myocardial perfusion imaging: Rubidium-82 chloride (⁸²Rb) and nitrogen-13 ammonia (¹³NH₃). ¹⁸F-Flurpiridaz is a novel PET MPI radiotracer undergoing Phase III studies and not yet approved by the FDA for clinical use.

Table 3-3 summarizes the criteria for the ideal PET myocardial perfusion imaging radiotracer. High and rapid myocardial uptake followed by long retention in myocytes are among the most important characteristics. Mitochondria are abundant in high energy consuming such as in the heart. Various radiolabeled analogs of mitochondrial complex I inhibitors, including rotenone, have been investigated.^60,61

Rubidium-82

Rubidium (Rb)-82 is the only one PET radiotracer currently available for MPI produced by a generator. Rb-82 is a monovalent cation potassium analog and is produced by eluting a Strontium-82/Rb-82 generator^40,41 (Table 3-4). Strontium-82 decays to Rb-82 by electron capture. The elution of Rb-82 can be repeated every 8 to 10 minutes. The energy of the Rb-82 positron is relatively high (3.15 MeV) and thus the range of positron travel before annihilation is relatively long. The positron range is approximately 2.6 mm root mean square (this is the distance traveled by the emitted positron before it combines with an electron to produce two 511-keV annihilation photons), resulting in lower spatial imaging resolution than with other PET tracers (e.g., F-18 and N-13). The physical half-life of Rb-82 is 76 seconds, allowing for multiple repeat imaging sequences. The half-life of the parent Strontium-82 is 25.5 days, and therefore the strontium–rubidium generator can be used for approximately 1 month.

Being a potassium analog, Rb-82 enters the myocytes using the sodium/potassium pump, and the biologic behavior is very similar to that of thallium-201.^36,37 The first-pass extraction fraction of Rb-82 is 40% to 75% at resting flow levels, which falls to 25% to 30% at high flow, and therefore also demonstrates "roll-off" at higher blood flow levels (Fig. 3-1).^38,39 However, the roll-off starts at slightly higher blood flow levels than for Tc-99m-Tc-99m-Tc-99m-sestamibi and Tc-99m-tetrofosmin, but at lower flow levels than for Tl-201. However, this is offset by the fact that generally Rb-82 perfusion images have higher count density and are of better quality than those with single-photon imaging agents.

The myocardial extraction fraction of Rb-82 remains reduced in the myocardium recovering from transient ischemia.^42–46 It is retained in the myocardium and shows equilibrium with the potassium pool. Cell membrane disruption may cause rapid tissue loss of the radiotracer. Rubidium-82 myocardial retention can thus be used as a marker of tissue viability.⁴⁷ Furthermore, quantitative assessments of relative Rb-82 perfusion defects have correlated well with those obtained from microspheres.⁴⁸ Because of the short half-life of Rb-82 it is not practicable to perform imaging after symptom-limited physical treadmill exercise. It is therefore standard to administer coronary vasodilator stress with the patient lying on the imaging table and infuse Rb-82 during the vasodilatory state.

Patient radiation exposure with an injected dose of 50 to 60 mCi of Rb-82 is relatively low because of the short half-life of Rb-82. The estimated effective radiation dose is 0.04 mSv/mCi and total patient dose may range from 4.0 to 6.0 mSv for a complete rest–stress Rb-82 study. Even though myocardial extraction of Rb-82 is not linear at higher flow levels, several investigators have shown that absolute regional myocardial blood flow may be derived from mathematical analysis of tracer kinetics.⁴⁹ The kinetic behavior of Rb-82 can be described as a two-compartmental model with a rapid and a slow clearance component.

The arterial input function is measured from serial regions of interest (ROI) over the cardiac blood pool and the myocardial inflow from ROIs over myocardial segments. The generated time–activity curves, corrected for spillover from the blood and decay, are fitted to the two-compartmental kinetic model to derive estimates of myocardial blood flow in mL/min/g.

In practice, quantification of myocardial blood flow with Rb-82 is a challenging task because of the low signal/noise ratio of the time–activity curves. Recently, Prior et al ⁵⁰ have validated a new method of quantification of myocardial blood flow using a correction methodology for the flow-dependent variable extraction of Rb-82.

N-13 Ammonia

Nitrogen-13 (N-13) ammonia was the first PET radiotracer developed for MPI and the first to be approved by the FDA for this specific clinical indication. N-13 ammonia is a cyclotron-produced cation with a physical half-life of 9.8 minutes (Table 3-4). The positron range is very short at 0.7 mm (1.4 mm in water). N-13 ammonia is rapidly extracted from the blood and trapped in the myocardium by the glutamine synthesis reaction. Myocardial uptake of N-13 ammonia depends on flow, extraction, and retention. The first-pass extraction fraction is nearly 100%, since N-13 ammonia freely diffuses across membranes.^51,52 In the myocardial tissues, N-13 ammonia is either incorporated into synthesis of N-13 glutamine or back-diffused into the vascular space. The net extraction fraction is approximately 80% (70–90%) at resting coronary blood flow range. Although linearity of myocardial uptake is better than the conventional myocardial perfusion imaging agents, roll-off does occur at higher myocardial blood flow levels (Fig. 3-1). N-13 ammonia crosses the cell membrane by passive diffusion as well as by means of active transport of the ammonium ion by the sodium/potassium pump.

Its relatively long physical half-life, high myocardial extraction fraction, and low liver and low background activity make N-13 ammonia one of the best available PET myocardial perfusion tracers. Different tracer kinetic models were used but the most common is the three-compartmental model that relates the N-13 ammonia activity: 1—in the vascular space, 2—the free interstitial space, and 3—the metabolically trapped space. Comparative studies in the animal model with microsphere measurements suggested that regional myocardial blood flow can be quantitatively assessed over a wide range of blood flow with the three-compartmental N-13 ammonia model.^53–55

For imaging usually 10 to 20 mCi's are injected as a bolus, followed by dynamic acquisition for 10 to 15 minutes. The dynamic list mode data are then fitted to three-compartmental model. Absolute myocardial blood flow is expressed in mL/min/g. The total effective radiation dose is approximately 2.0 mSv for an injection of 20 mCi of N-13 ammonia. The target organ is the bladder, which receives 6 mSv.

O-15 Water

Oxygen (O)-15 water was one of the first radiopharmaceutical used for PET application.⁵⁶ O-15 gas is produced either by ¹⁴N (d,n) ¹⁵O reaction or ¹⁵N (p,n) ¹⁵O method. ¹⁵O₂ gas is combined to hydrogen gas to produced O-15 water.

O-15 water is cyclotron-produced, has a half-life of 2 minutes and a positron range of 1 mm (Table 3-4). After injection it diffuses freely from the blood into myocardial cells. The extraction fraction is very high and not affected by myocardial blood flow rates and is independent of metabolic state of the myocardium. Because of these characteristics, the biologic behavior of O-15 can be modeled with a simple one-compartmental Kety's model. As a consequence myocardial uptake is perfectly linear over a wide range of myocardial blood flows (Fig. 3-1). Therefore, O-15 water is considered the gold standard for clinical noninvasive absolute myocardial blood flow measurements. Cardiac PET imaging with O-15 water is very demanding because of persisting high blood pool activity, requiring subtraction of blood pool activity from the original image to visualize the myocardium. Using advanced processing of dynamic images with dedicated research softwares, relative myocardial blood flow can be estimated. Despite its success in research studies, the clinical use of O-15 water in patients is limited. Although O-15 water is used clinically in some European hospitals, it is not FDA approved yet and still considered investigational in North America.

F-18 Fluorodeoxyglucose (FDG)

The myocardium can choose various energy substrates, including free fatty acids, glucose, lactate, and ketone bodies. In the fasting state, plasma free fatty acid levels are high so that 70% to 80% of the myocardial oxygen consumption is obtained by oxidation of free fatty acid.⁵⁷ Conversely, the postprandial state elevates plasma glucose level so that myocardium shifts its fuel selection to glucose.

Fluorodeoxyglucose (FDG) is a glucose analog that enters cardiac myocytes by facilitated diffusion via glucose transporters. Once FDG is in the cell it is phosphorylated by hexokinase to FDG-phosphate. However, this is not further metabolized and FDG is trapped within the myocytes. Myocardial accumulation of FDG therefore parallels exogenous glucose utilization. FDG can be labeled with Fluor-18 (F-18), which has a half-life of 110 minutes and a positron range of 0.6 mm.

The myocardial extraction of FDG from the blood is rapid but only a small percentage of injected dose, 1% to 5%, is taken up by the heart, depending on the nutritional state or insulin stimulation. To enhance FDG uptake in the heart, the patient should be either glucose loaded or undergo the hyperinsulinemic euglycemic clamp; these techniques are beyond the scope of this chapter. These preparations drastically reduce plasma free fatty acid levels and stimulate glucose utilization.

F-18 FDG is cleared via the kidneys and excreted in the urine. The effective dose from the administration of F18-FDG for a PET-CT scan is estimated to be approximately 0.019 mSv/MBq for adults using 370 to 740 MBq (10–20 mCi). The bladder is the organ receiving the highest radiation dose. The effective dose from the CT portion of the PET/CT study can range from approximately 5 to 80 mSv depending on the CT system and protocol being used but is often comparable to the effective dose from the PET portion of the study.

F-18 FDG is predominantly used in cardiac imaging to assess myocardial viability in association to MPI radiotracers.^58,59 For this specific clinical application, glucose metabolism and uptake should maximized by the above-mentioned patient preparation. However, when F-18 FDG imaging is used to detect resting myocardial ischemia, for example, in unstable angina, or to detect active cardiac sarcoid lesions, one may consider to administer FDG in fasting state in order to bring out areas with enhanced glucose metabolism adjacent to areas with normal fatty acid myocardial metabolism.

F-18 Flurpiridaz

One of the most promising radiolabeled analogs of these mitochondrial complex I inhibitor is ¹⁸F-flurpiridaz with a pyridaben pharmacophore.⁶² Initially named RP1012 or BMS747158, the chemical structure of ¹⁸F-flurpiridaz is: 2-tert-butyl-4-chloro-5-[4-(2-fluoro-ethoxymethyl)-benzyloxy]-2H-pyridazin-3-one. ¹⁸F-flurpiridaz can be synthesized by a one-step direct fluorination in a radiopharmacy, using an automated synthesis unit.

In a rat cardiac myocytes kinetics assay, ¹⁸F-flurpiridaz showed rapid uptake with a half-life of 35 seconds for maximal uptake and showed washout with a half-life of more than 2 hours.⁶³ Yu et al.⁶⁴ evaluated the tissue biodistribution and first-pass extraction fraction of ¹⁸F-flurpiridaz in rat, rabbit, and nonhuman primate models at rest and reported first-pass extraction fraction of more than 90%. The results showed that myocardial uptake of ¹⁸F-flurpiridaz was significantly higher than that of ^99mTc-sestamibi not only at normal MBF range in rats but also at high MBF ranges (<5 mL/min/g) in isolated rabbit hearts, suggesting that ¹⁸F-flurpiridaz may be less affected by the roll-off phenomena commonly seen with SPECT myocardial perfusion imaging agents. The cardiac uptake of ¹⁸F-flurpiridaz at 15 minutes in rats was 3.5 ± 0.3% of injected dose/g in comparison to 1.9 ± 0.1 for ^99mTc-sestamibi. The heart/lung and heart/liver ratio at 60 minutes after injection was 12.7± 1.4 and 3.7 ± 0.2, respectively (5.9 ± 0.5 and 2.4 ± 0.3 for ^99mTc-sestamibi, respectively). Myocardial uptake of ¹⁸F-flurpiridaz was shown to be stable for at least 1 hour after injection, and there was no significant myocardial redistribution.

In rat and pig models of myocardial infarction and ischemia, the ¹⁸F-flurpiridaz defect area was clearly detected.^65–67 There was a good correlation between the infracted regions detected by ¹⁸F-flurpiridaz and the necrotic area identified by ex vivo tissue histology. In comparison to ¹³N-ammonia, ¹⁸F-flurpiridaz showed higher activity ratios between the myocardium and blood, liver, and lungs.⁶⁷ Regional myocardial blood flow assessed with 18F-flurpiridaz showed a good correlation (r = 0.88, slope = 0.84) and agreement with that measured with radioactive microspheres over a wide flow range from 0.1 to 3.0 mL/min/g. A linear correlation existed between MBF quantified by radiolableled microspheres and myocardial uptake of F-18 flurpiridaz after adenosine hyperemia.^62,68

The high spatial resolution of PET imaging, along with high target-to-background ratios achieved with ¹⁸F-flurpiridaz, allows the acquisition of very high-quality electrocardiographic-gated PET images in both animal and human studies. Initial studies in humans have shown a very good diagnostic accuracy in detection of significant CAD.^69,70 Early clinical trials have shown that ¹⁸F-flurpiridaz intravenous injection is well tolerated and no clinically safety concerns have been raised. In humans, the mean effective dose is 0.02 ± 0.002 mSv/MBq at rest and 0.019 and 0.015 mSv/MBq with adenosine and exercise stress, respectively. This corresponds to that of ¹⁸F-FDG (0.019 mSv/MBq). The organ with the highest radiation exposure is the heart with both adenosine and exercise stress.

The longer half-life (110 minutes) of the F-18 label of flurpiridaz makes it possible to administer the radiotracer at peak physical exercise and perform PET imaging immediately thereafter, which represents another potential advantage of ¹⁸F-flurpiridaz PET imaging in comparison to all the other PET radiotracers which are suited for pharmacological stress tests and not for treadmill stress test.

Multiple radionuclide MPI agents are now commercially available, for both SPECT and PET imaging. Although many of their characteristics differ, they all share the same utilization, that is, the diagnostic and evaluation of patients with coronary artery disease. We are still learning how to obtain the best diagnostic results from Tl-201, Tc-99m-MPI agents, and PET radiotracers for myocardial perfusion scintigraphy. It is likely that with the newest agents and constantly evolving technology in data acquisition and analysis, our knowledge will improve and all these agents will be able to fulfill a more specific role in clinical practice.