¹³N-Ammonia

¹³N-Ammonia is a cyclotron produced PET perfusion agent with a physical half-life of 10 minutes requiring an on-site cyclotron and radiochemistry synthesis capability. In the blood, ¹³N-ammonia exists in equilibrium between two states, neutral ammonia (NH₃) and charged ammonium (NH₄⁺). The neutral ammonia rapidly crosses cell membranes. Once it is inside the cell, glutamine synthase takes glutamate, ammonia, and ATP and produces glutamine, which remains intracellular, thereby fixing the ¹³N within the cell with very little diffusion back into the blood pool. ¹³N-Ammonia has a high extraction rate, with a significant percentage of the agent entering the myocyte.¹ Animal models demonstrate, however, a nonlinear extraction of ¹³N-ammonia compared with the gold standard of microsphere flow measurements. At flows in the normal resting physiologic range of about 50 to 150 mL/min/100 g, ¹³N-ammonia uptake is nearly linear. However, at higher flows of more than 200 mL/min/100 g, such as those produced with vasodilator stress, there is a plateau effect, and higher flow rates are not associated with a further linear increase of ¹³N-ammonia uptake and retention.² The nonlinear uptake at higher flow rates does not reflect a limitation of the diffusion of ¹³N-ammonia; rather, it is a limitation of metabolic trapping through the glutamine synthase pathway. This roll-off phenomenon may lead to underestimation of the true myocardial perfusion at higher flows, a statement true for all extracted radiotracers (Fig. 25-1). ¹³N-Ammonia image quality is generally excellent, with rapid clearance of the tracer from the blood pool and lung, although lung uptake can be an issue in patients with chronic lung disease or depressed left ventricular function.³

FIGURE 25-1 True myocardial blood flow as measured by microspheres (x-axis) is plotted versus radiotracer uptake. At lower (physiologic) flows, there is a linear relationship between the two. However, at higher flow rates, tracer uptake “rolls off.” This is not true of ¹⁵O-water, which is a freely diffusible tracer and can track flows across a wide range of rates.

(From Glover DK, Gropler RJ. Journey to find the ideal PET flow tracer for clinical use: are we there yet? J Nucl Cardiol 2007; 14:765-768.)

Imaging with ¹³N-ammonia is typically initiated either simultaneously with or shortly after a 10- to 20-mCi radiotracer infusion. For estimation of absolute blood flow, a dynamic image acquisition must be initiated with the infusion to determine the arterial input function and myocardial uptake and clearance kinetics for application of compartmental analysis. When ECG-gated PET imaging is performed only for the determination of relative perfusion and function, the acquisition begins 90 seconds to 3 minutes after infusion, which allows clearance of tracer activity from the blood pool. Image acquisition lasts 5 to 15 minutes, and attenuation correction is used in image generation. ¹³N-Ammonia perfusion information is used as part of a rest-stress protocol, defining myocardial ischemia, or it is used in conjunction with 2-fluorodeoxyglucose (FDG) to delineate the classic mismatch between myocardial perfusion and metabolic activity in the evaluation of myocardial viability.

¹⁵O-Water

¹⁵O-Water is also a cyclotron-produced agent with an ultrashort half-life of 2 minutes. Unlike ¹³N-ammonia and ⁸²Rb, ¹⁵O-water is a diffusible agent that is not extracted by or trapped within the myocytes but rather freely diffuses across membranes. The kinetics of ¹⁵O-water are dependent only on myocardial perfusion and are not affected by the metabolic rate-limiting steps that affect extracted tracers, eliminating the issues with tracer roll-off. However, PET ¹⁵O-water imaging generally is performed only as a dynamic study to determine uptake kinetics. Application of a single-compartment model to the dynamic image data allows a highly accurate estimation of absolute myocardial flow across a wide range of blood flow rates (0.29 to 5.04 mL/min/g).^{4 15}O-water can be administered as an intravenous bolus, as a slow infusion, or by inhalation of ¹⁵O-CO₂, which is converted to ¹⁵O-water in the lungs by carbonic anhydrase.

The advantage of ¹⁵O-water as an “ideal” flow tracer is counterbalanced by the complex acquisition protocols and analyses required for use in clinical practice. Because the ¹⁵O-water is not extracted from the blood pool, the images are contaminated by the high level of residual activity within the blood pool, which must be subtracted from the final image to properly evaluate myocardial perfusion. Two techniques are used to accomplish this goal. The blood pool can be labeled with ¹⁵O-carbon monoxide (CO). Inhalation of ¹⁵O-CO leads to irreversible binding of the CO to hemoglobin, thereby labeling the red blood cells and the blood pool. Blood pool data obtained during this acquisition can be subtracted from the ¹⁵O-water perfusion, eliminating the contribution of the blood pool from the final ¹⁵O-water data set. Alternatively, very early acquisitions (20 to 40 seconds after tracer infusion) provide an image of the blood pool before tracer flow has reached the myocardium; these are then used to remove background blood pool signal from the myocardium.⁵ The difficulties in image processing as well as the increased radiation exposure to the lung with use of the inhaled ¹⁵O-CO technique limit the clinical application of ¹⁵O-water.

Rubidium 82

⁸²Rb is a radioactive monovalent cationic potassium analogue. Unlike the other perfusion agents used in cardiac PET, ⁸²Rb is produced in a commercially available generator, alleviating the need for an on-site cyclotron, which makes it a more practical perfusion tracer for widespread clinical application. ⁸²Rb is obtained by elution of 25 to 50 mL of normal saline with use of a computerized pump through a strontium 82 (⁸²Sr) generator. The generator is rapidly replenished, with 90% of the maximal activity available after 5 minutes, and full replenishment occurs by 10 minutes. ⁸²Rb has an ultrashort half-life of 75 seconds, which allows rapid acquisition of sequential stress and rest images (Fig. 25-2). The half-life of ⁸²Sr is 25.5 days, and therefore the generator needs to be replaced every 4 weeks.

FIGURE 25-2 Adenosine ⁸²Rb stress-rest study. ⁸²Rb allows high-quality images. This study demonstrates normal myocardial perfusion at stress and rest with homogeneous uptake throughout.

⁸²Rb is extracted from the blood pool by the Na⁺,K⁺-ATPase pump. Extraction of ⁸²Rb is similar to that of thallium 201 (²⁰¹Tl), another potassium analogue,⁶ but it is less than that of ¹³N-ammonia⁷; therefore, the issues associated with roll-off may be present in ⁸²Rb as they are in all other extractable tracers. Extraction of ⁸²Rb is affected by myocardial perfusion as well as by the metabolic milieu. Severe acidosis and hypoxemia can affect ⁸²Rb extraction.⁶ In addition, ⁸²Rb is a less ideal imaging agent because the ⁸²Rb positron has a high kinetic energy, which allows the positron to travel a relatively long distance before an annihilation event occurs. The increased positron range of the parent molecule results in a reduced spatial resolution of ⁸²Rb imaging compared with that possible with other PET agents. Despite these issues, ⁸²Rb has become one of the most commonly used PET perfusion agents because of the availability, ease of use, and acceptable image quality.

Summary of PET Perfusion Imaging

PET perfusion has become a routine tool in the management of cardiac patients. Multiple studies have evaluated the accuracy of perfusion PET for the detection of coronary artery disease. PET has consistently been found to have a high sensitivity and specificity for the detection of coronary artery disease. The accuracy of this technique makes the use of PET perfusion an attractive option in patients with chest pain or suspected coronary artery disease despite the increased cost and relative technical complexity (Table 25-2).^8–15

¹⁸F-Fluorodeoxyglucose (¹⁸F-FDG)

Fluorodeoxyglucose (2-fluoro-2-deoxy-d-glucose, FDG) is a glucose analogue that enters the cell through facilitated diffusion. Once inside cells, it is phosphorylated by hexokinase into FDG 6-phosphate, which is metabolically inert and is trapped within the myocyte. Therefore, retention of FDG is an energy-requiring process and is dependent on intact metabolic pathways (Fig. 25-3). FDG can be labeled with fluorine18 (¹⁸F), which decays by positron emission with a half-life of 110 minutes. The prolonged half-life of ¹⁸F allows local distribution of the agent from a central cyclotron site to regional imaging centers. Whereas it is the most commonly used agent in oncologic PET imaging, ¹⁸F-FDG is used primarily for the evaluation of myocardial viability in cardiac imaging.

FIGURE 25-3 Glucose and fatty acid pathways in the myocytes. FDG enters as glucose and after phosphorylation becomes trapped intracellularly. Free fatty acids such as acetate and palmitate enter the cell and participate in the beta-oxidation pathway.

(From Herrero P, Gropler RJ. Imaging of myocardial metabolism. J Nucl Cardiol 2005; 12:345-358.)

Myocytes generally favor fatty acids derived from adipose tissue stores as their primary source of energy. However, after a glucose load, glucose becomes the primary metabolic energy substrate as systemic release of insulin limits free fatty acid release and increases transmembrane transport of glucose. In the setting of myocardial ischemia, free fatty acid metabolism decreases while glucose metabolism increases, resulting in increased uptake of ¹⁸F-FDG in the ischemic tissue. As the phosphorylation of ¹⁸F-FDG is an energy-requiring process, uptake of ¹⁸F-FDG can occur only in metabolically active (viable) myocytes. Therefore, in the clinical setting, ¹⁸F-FDG is used to determine viability in chronically ischemic myocardium.

Preparation of the patient is important in the proper acquisition of ¹⁸F-FDG images. In the fasting state, glucose uptake is low and inhomogeneous. To maximize glucose use by the myocardium, the patient is asked to fast for 6 to 12 hours. An oral glucose load (25 to 100 g) or an intravenous glucose load is administered and followed by insulin as needed before imaging. In diabetic patients, a more rigorous methodology is necessary to produce high-quality scans, and a euglycemic hyperinsulinemic clamp is often used. A dose of 5 to 15 mCi is injected, and imaging begins at least 45 minutes after tracer infusion. Viability studies with ¹⁸F-FDG are performed in conjunction with a perfusion study. Perfusion can be acquired with PET perfusion agents as described before or with SPECT perfusion agents such as ²⁰¹Tl or technetium Tc 99m. Preserved FDG uptake in the setting of reduced perfusion suggests the presence of viable myocardium. These segments are likely to have improved function after revascularization. Segments that have both diminished perfusion and diminished FDG uptake are unlikely to recover function with revascularization and consist mostly of infarcted tissue (Fig. 25-4).

FIGURE 25-4 Cardiac viability study. Top rows (A) represent FDG study; bottom rows (B) represent ⁸²Rb perfusion examination. Short-axis images demonstrate a large perfusion defect in the lateral segments (white arrow). There is uptake of FDG in these segments (yellow arrow), indicating viable myocardium. Horizontal long-axis images demonstrate that the perfusion defect involves the apex as well as the lateral wall; all segments are viable as indicated by FDG uptake.

¹¹C-Palmitate and ¹¹C-Acetate

As noted previously, free fatty acids are the primary source of energy in normal myocardium, accounting for 60% to 80% of ATP produced in nonischemic myocytes. Free fatty acids are activated as acyl coenzyme A. Acyl coenzyme A enters the mitochondria through the acyl carnitine transport system and becomes part of the beta-oxidation pathway. Whereas this pathway is a rich source of ATP, it is highly oxygen dependent. During periods of ischemia, there is a rapid shift away from fatty acid metabolism to increased glucose use (see Fig. 25-3). Free fatty acids account for the preponderance of myocardial energy formation, and these pathways are dramatically altered by ischemia. Therefore, free fatty acid imaging is an attractive target for noninvasive imaging of ischemia and the associated alterations in oxidative metabolism. PET imaging of fatty acid metabolism generally involves imaging with carbon 11 (¹¹C)–labeled medium-sized straight-chain fatty acids like palmitate, which undergo beta-oxidation. ¹¹C has also been used to label acetate as an alternative approach for imaging of oxidative metabolism. ¹¹C is cyclotron produced and has a relatively short half-life of 20.4 minutes. Initially produced in 1934 and first studied in humans in 1945, it has the advantage of being an organic molecule and therefore can potentially be used to target a wide variety of metabolic processes.¹⁶ Unfortunately, ¹¹C-labeled metabolic agents are more difficult to employ in clinical practice and have been primarily used for research purposes. However, both ¹¹C-palmitate and ¹¹C-acetate have been effectively employed for the evaluation of myocardial oxidative metabolism.

Several studies have evaluated fatty acid metabolism in normal volunteers and patients. Walsh and colleagues¹⁷ demonstrated decreased mitochondrial metabolism in infarcted myocardium. In addition, ¹¹C-palmitate infusion during dobutamine stress testing demonstrated the expected rise in fatty acid metabolism in the normal myocardial segments, whereas areas supplied by stenosed vessels did not show a rise in fatty acid use.¹⁸ Last, free fatty acid metabolism imaging has been used to demonstrate differences in myocardial metabolism in the elderly¹⁹ as well as in inherited nonischemic cardiomyopathies.²⁰ In both cases, fatty acid metabolism was decreased compared with normal controls. Although the clinical application of some of these findings is not fully elucidated, free fatty acid PET offers a powerful method for the in vivo evaluation of myocardial metabolism and therefore provides an important tool for gaining insight into disease processes associated with altered myocardial metabolism.