PET Myocardial Imaging:
Advantages of cardiac PET over SPECT include higher spatial resolution, attenuation correction, and the ability to perform quantitative measurements [17]. In the detection of coronary artery disease, PET imaging has a sensitivity of 92%, a specificity of 89% (higher than SPECT- 75%), and a normalcy of 89 to 95%. The higher specificity is related to the ability of PET imaging to provide attenuation correction which decreases the false positive rate.
PET Perfusion Imaging
Physiology
All PET perfusion agents can be used to measure absolute blood flow in ml/gm/min. Quantitation of myocardial perfusion with PET requires mathematical models which describe the kinetic behavior of the tracers used and measurements of tracer activity in the blood and tissue. Blood activity can be measured accurately by direct sampling of blood from an artery, but this is invasive and the sample would still require correction for time delays and dispersion. Measurement from the reconstructed PET images would be more practical.
The most appropriate site to measure arterial blood pool activity is at the level of the coronary ostia since this is the point from which arterial blood irrigates the myocardial tissue. Unfortunately, measuring activity from an ROI placed within the ascending aorta is subject to error due to the limited resolution of the imaging system, the small dimensions of the aorta, and contamination from adjacent vascular and myocardial structures. Thus, activity is usually measured from an ROI placed within the left atria or left ventricle.
The drawback of a left ventricular ROI is that it is subject to motion artifact as well as spillover activity from myocardial tissue. A left atrial ROI tends to closely match values obtained from direct arterial sampling. It is important, however, to correct for the small time delay between the arrival of the blood in the left atria and its actual arrival in the myocardium or flow rates may be artifactually elevated.
Radiopharmaceuticals/Technique
13N is cyclotron produced by either irradiating methane gas with deuterons [12C (d,n) 13N] or by the reaction 16O (p,alpha) 13N with water as the target. 13N has a physical half-life of 10 minutes. There is rapid blood clearance with high initial extraction (over 90%) and high retention (80%) which provides high myocardial to background count ratios. Extraction decreases, however, at high flow states (to as low as 35%), but to a less significant degree when compared to 82Rb.
The agent diffuses into myocytes where it is metabolized to glutamine and becomes unable to leave the cell. Extraction of the tracer generally reflects regional perfusion . Marked lung uptake is seen in all smokers. Another consistent finding is low uptake of the tracer in the lateral/posterolateral wall on both stress and rest images which appears to be related to a regional alteration in 13N tissue metabolism/ retention [3]. This defect is not detected on dynamic imaging which can be used to differentiate a true perfusion defect from artifact [4].
The critical organ for N-13 is the urinary bladder. The typical dose used for imaging is 10 to 15 mCi I.V. given over a 20 to 30 second interval. Imaging is usually started 5 minutes after injection of the tracer to permit pulmonary background activity to clear and images should be acquired in a decay compensated mode. A FDG exam may be performed after a 50 minute interval to permit decay of the N-13 to background levels.
15O is cyclotron produced by the irradiation of nitrogen [14N (d,n) 15O or 15N(p,n)15O] and it has a physical half-life of only 2.2 min. (123 sec.). O-15 produced in the cyclotron is converted to [15O]CO2 by passing it over activated charcoal at 400 to 600 degrees Celsius. [O-15] water is then prepared by bubbling the [15O]CO2 into water.
Theoretically, 15O water is ideal for quantitative regional myocardial flow measurements (in ml/min/gm) for 2 reasons: it is a freely diffusible perfusion tracer with 100% extraction by the myocardium, and it is not affected by metabolic factors. The extraction of oxygen-15-water (H215O) remains linear even at very high flow rates (it is the only perfusion tracer with linear extraction) and therefore, myocardial distribution of the agent reflects regional perfusion.
Unfortunately, image quality is not as good as that obtained with other flow agents because tracer circulating in the blood pool remains within the ventricular chamber and must be subtracted in order to visualize the myocardium.
is a generator produced positron emitter. Strontium-82 (T1/2=25 days) decays by electron capture to. Rubidium-82 is a potassium analog with a physical half-life of only 75 seconds; this allows for repeated blood flow measurements in short time intervals.
Unfortunately, the short half-life has some negative effects as well- following injection imaging must be delayed for at least 2 minutes to permit clearance of blood pool activity and considerable decay of the tracer occurs during this time. Also, due to changing myocardial activity from decay, imaging times must be short to prevent backprojection reconstruction artifacts.
The extraction of is about 50-60% at rest, and decreases at high flow rates (25-30%). The typical dose for the perfusion exam is 40 to 50 mCi infused over 20 to 30 seconds. Imaging is usually begun 60 seconds after the infusion is complete to permit clearance of blood pool activity.
82Rb has the worst resolution of all the positron emitting agents, with most positrons traveling about 12.4 mm prior to undergoing annihilation. The critical organ is the kidney.
Copper 62 PTSM: Copper-62-Pyruvaldehyde-bis-(4N-thiosemicarbazone)
Cu-PTSM is a generator-produced tracer which is presently undergoing clinical evaluation. 62Cu has a physical half-life of 9.7 minutes. The short half-life of the parent isotope 62Zn (9.3 hours) limits the practical life of the generator to 1 to 2 days. However, once loaded, the generator can be eluted every 30 minutes.
62Cu -PTSM is a neutral, lipophilic tracer that is taken up in tissues and then trapped intracellularly by reduction to a non-lipophilic compound. There is non-linear extraction of the tracer (ie: extraction decreases at high flow rates) and high liver activity which scatters into the inferior wall. The agent also binds to human albumin which precludes accurate recording of the arterial input function which is critical for quantification. [5]
PET Metabolic Imaging
Physiology of FDG myocardial imaging
The heart can oxidatively metabolize a variety of substrates to meet its high energy needs. In the fasting state (high circulating free fatty acids and low insulin levels) the myocardium preferentially utilizes fatty acids and lactate as aerobic substrates to provide the energy necessary for contractile function (unlike other cells in which glucose is the most important substrate). However, even in the fasting state, glucose can still account for 30-40% of the energy derived from oxidative metabolism [15]. In plasma, fatty acids are transported bound to albumin and enter the myocardial cell either by passive or active transport. Once in the cell, fatty acids undergo oxidative metabolism. A depression of fatty acid extraction/oxidation will occur in zones of myocardial ischemia or infarction.
When blood flow is significantly reduced, tissue delivery of oxygen and removal of waste products are also reduced. As a result, oxygen dependent substrate catabolism decreases. Since fatty acids are catabolized by beta oxidation to acetyl-CoA, which requires the Krebs cycle and ultimately oxidative phosphorylation to produce ATP, fatty acid consumption ceases with anoxia. This is followed by an increased glycolytic flux with glycogen depletion and increased exogenous glucose utilization. Enhanced anaerobic glycolytic activity persists for some time after reperfusion despite the availability of adequate oxygen.
FDG uptake in the myocardium is highly dependent on the patients dietary condition [14]. Myocardial glucose utilization is increased by glucose administration with stimulation of insulin secretion (ie: such as following a meal). The increased insulin levels stimulate glucose metabolism and tissue lipolysis is inhibited [15]. Therefore, FDG uptake will be promoted in association with high glucose/insulin levels, but decreased in the face of high free-fatty acid levels.
Radiopharmaceutical: F-18 2-Deoxyglucose: (18-FDG)
Fluorine-18 has an effective half life of 110 minutes. It is produced by bombarding 18O with protons and displacing a neutron [18O (p,n) 18F]. Chemically 18F is comparable in size to hydrogen. 18F has the best resolution of all positron emitters- approaching 2 mm. This is because most of the emitted positrons traveling only about 1.2 mm prior to undergoing an annihilation reaction.
The agent 18FDG competes with glucose for facilitated transport into myocardial cells and also competes with glucose for phosphorylation by hexokinase. Unlike glucose, however, the phosphorylated form is not further metabolized. Regional myocardial uptake of FDG therefore reflects regional rates of exogenous glucose utilization. Only about 1 to 4% of the injected dose is trapped in the myocardium, but the target to background ratios are favorable (Heart:Lung [20:1], Heart:Blood [14:1]). The typical dose used for the exam is 15 mCi infused over a 60 second interval.
Standard Technique with FDG
Evaluation of glucose metabolism using 18FDG remains the gold standard for the evaluation of myocardial viability [6]. Ischemic, but viable myocardium (hibernating) utilizes glucose in preference to other substrates. Ischemic myocardium and myocardium recovering from a recent ischemic event will therefore concentrate an increased amount of the glucose 18FDG. Energy production, however, is typically insufficient to maintain mechanical work and impaired contractile function results. FDG images are usually acquired 45-60 minutes following injection of the radiotracer.
Examination of patients in the fasting state will demonstrate an area of ischemic, but viable myocardium as a region with normal or increased glucose utilization. Unfortunately, this technique may be overly sensitive- detecting small, clinically insignificant areas of myocardial viability because of low background myocardial activity (no FDG uptake by the surrounding normal and non-viable myocardial segments during fasting). Additionally, the normal myocardium can be difficult to define in the fasting state and the tracer may not clear adequately from the blood which can result in image degradation. Nonetheless, the average positive predictive accuracy for detecting viable myocardium by fasting FDG studies is about 85% and is similar to post-glucose loading exams (81%). Imaging following the administration of glucose will shift myocardial substrate utilization from fatty acids to glucose, however, this approach may miss small areas of viable myocardium.
Several protocols are available to promote cardiac FDG uptake- these include oral glucose loading, hyperinsulinemic euglycemic clamping, and administration of nicotinic acid derivatives (acipimox 250 mg orally given 2 hours prior to the exam inhibits peripheral lipolysis and reduces plasma free-fatty acid levels) [14]. Most studies are performed following glucose loading- typically using an oral solution containing 50 to 75 gm of glucose given approximately 60 minutes before FDG administration [15]. Alternatively, a glucose infusion can be performed using a 10% dextrose in water solution at a rate of 15 umole/kg/min. Up to 90% of exams will be of adequate-to-excellent quality in non-diabetic patients following glucose loading [15]. However, in patients with diabetes, the increase in plasma insulin levels following glucose loading may be attenuated [15]. Consequently, tissue lipolysis is not inhibited, plasma fatty acid levels remain high, and myocardial glucose uptake can be poor [15]. This can result in suboptimal quality exams. In diabetic patients, an IV bolus of regular insulin may be performed (according to a sliding scale) and the plasma glucose is checked every 15 minutes in order to maintain a stable serum glucose level of approximately 140 mg/dL. Another alternative in diabetic patients is hyperinsulinemic-euglycemic clamping which requires the simultaneous infusion of glucose and insulin to achieve a stable plasma insulin level. Unfortunately, the procedure is tedious and not practical for clinical imaging [15].
Imaging Findings:
Severely decreased FDG uptake below 50% of normal myocardial activity is generally considered indicative of scar, while uptake greater than 50% of normal myocardial activity is indicative of viability. However, determination of ischemic, but viable myocardium can only be made in relationship to the perfusion study [7]. When regional myocardial FDG uptake is disproportionately enhanced as compared to regional myocardial blood flow, the pattern is termed a perfusion-metabolism "mismatch." In the setting of chronic coronary artery disease, a perfusion-metabolic mismatch is highly predictive of myocardial viability and indicates a high likelihood of improved cardiac function following revascularization (improved contractility is seen in 82-85% of cases). For the prediction of improvement in regional LV function after revascularization FDG PET imaging has a pooled sensitivity of 88%, a specificity of 73%, a negative predictive value of 86%, and a positive predictive value of 76% [14]. Unfortunately, in the evaluation of FDG PET for myocardial viability there is wide variability in the literature regarding patient inclusion criteria and examination protocol [14].
A scar is characterized by concordant reduction in perfusion and FDG uptake [14]. Improved wall motion is seen in only 8-17% of matched perfusion-metabolic defects which are felt to most likely represent scar [14]. Based on the severity of the perfusion and FDG deficit, the "matched" pattern may be categorized as a transmural match (absent or markedly reduced perfusion and FDG uptake) or a non-transmural match (mild to moderately reduced perfusion and FDG uptake). When tracer activity equals 50% to 60% or more, the matches are mild and probably represent a non-transmural scar [14]. If quantification is performed, viable myocardium is very unlikely to exist if the basal myocardial blood flow is less than 25 ml/min/100 gm [8].
Although the uptake of 18FDG can discriminate between viable and non-viable tissue, the regional myocardial blood flow-FDG uptake pattern is similar for varying types of myocardial dysfunction and clinical symptoms are necessary to determine the exact etiology of the scan findings. Different conditions may also coexist in the same myocardial regions.
Patterns of Disease (Table)
CONDITION rMBF FDG UPTAKE CLINICAL STATE Acute Ischemia Decreased Norm or Incr. Acute CP Hibernation Decreased Norm or Incr. Chronic stable Stunning Normal Norm or Incr. After acute event Necrosis/Scar Decreased Decreased Chronic sx's
Prognostic Implications of FDG PET Myocardial Imaging (General):
PET studies have shown that cardiac morbidity and mortality is increased in patients with flow-metabolic mismatches. Up to 50% of patients that demonstrate a perfusion-metabolic mismatch will have a cardiac event in the subsequent 12 months in the absence of intervention. The incidence of cardiac events drops to 15% in these patients if revascularization is performed [9]. In two other studies, mortality ranged between 4 to 12% in the group with matched defects, and between 33 and 41% in the mismatch group. In the mismatch group, if revascularization was performed, mortality dropped to between 4 and 12% [7].
It is well recognized that in patients with CAD on medical therapy, the presence of LV dysfunction is associated with a high mortality. In the CASS study, mortality of medically treated patients was related to the severity of LV dysfunction with up to 25% annual mortality in patients with resting LVEF's below 25%. In patients with LVEF's below 35% survival was better in those patients undergoing revascularization. Unfortunately, revascularization cannot be recommended in all patients with poor LV function since the surgery itself is associated with a 5 to 35% mortality in this subgroup. Ideally, the surgery should be performed only in those patients with a high likelihood of improved LV function.
In patients with symptoms of cardiac failure a PET pattern of perfusion-metabolic mismatch identifies a subgroup of patients who are at very high risk for cardiac death on medical therapy (30-40%) and who are most likely to show improvement as a result of revascularization. Survival in patients with a mismatched pattern can be significantly improved by myocardial revascularization [1] (cardiac mortality drops to 4 to 12%) compared to patients with a matched pattern. In patients with heart failure symptoms and a PET mismatch, 70% had improvement in symptoms following revascularization, while only 30% without a mismatch experienced improvement. A blood flow-metabolism mismatch is 68-95% accurate for predicting post-revascularization improvement in regional wall motion, whereas a matched defect is 75-100% accurate in predicting that segemental wall motion will not improve [16]. In general, the larger the extent of the perfusion-metabolic mismatch, the greater the anticipated improvement in LVEF following revascularization. In contrast, the average LVEF remained unchanged or decreased in revascularized patients who did not have a PET mismatch. Therefore, the mismatch pattern is also predictive of improvement in heart failure symptoms after revascularization [10]. When 25% of more of the LV was viable, a significant improvement in LVEF can be expected [14].
Drawbacks of FDG imaging include limited data in diabetic patients and recent reports which suggest that increased FDG accumulation can be seen in acute evolving myocardial infarction [1,11].
Perfusion-Metabolism Mismatch: The case below is from a 62 year old female patient 8 years status post CABG and 6 months following an anterior wall myocardial infarction. The patient presented with CHF and an LV ejection fraction of 22%. Perfusion exam was performed using NH3. An extensive perfusion-metabolism mismatch can be seen involving the apex, anterior, and septal walls. At angiography the patient was found to have an occluded bypass graft and a second bypass was performed. Six months later the patient was asymptomatic and the LV ejection fraction had increased to 47%. Case courtesy of CTI PET Systems, Inc. |
Prognostic Implications in the Immediate Post Infarct Period:
In the immediate post infarction period (out to 1 week post event) the functional outcome of infarcted segments with a flow-metabolism mismatch pattern is more variable and does not necessarily imply recovery of contractile function following revascularization (ie: in the post infarct period PET may overestimate the potential for recovery) [12].
Other Approaches to Myocardial Imaging
In addition to standard perfusion -metabolic imaging, other approaches have also been utilized to assess for viability, including determination of oxidative metabolism with 11C palmitate, uptake and retention of 82Rb, 18F fluoroisonidazole imaging, and the water perfusable tissue index
PET Myocardial Fatty Acid Imaging
Long chain fatty acids account for 90% of the myocardial energy requirements in the fasting state. When glucose or insulin levels are high, such as after a meal, glucose oxidation increases and fatty acid use is suppressed [2]. During ischemia, glucose again plays a major role in oxidative metabolism, whereas oxidation of long chain fatty acids is greatly reduced [2]. Long-chain free fatty acids easily pass through the sarcolemmic membrane to be activated as acylcoenzyme A (CoA). Acyl CoA is carried into mitochondria through an acyl carnitine carrier system to enter the beta-oxidation pathway. After beta-oxydation , acetyl CoA is formed and enters into the tricarbonic acid cycle for further oxidation to become water and carbon dioxide. Palmitic acid comprises approximately 25 to 30% of the circulating fatty acid in the blood and serves as the one of the primary sources for energy production by the heart. The distribution of C-11 labeled palmitic acid reflects efficiency of myocardial energy production.
After initial uptake of the tracer, there is a rapid clearance component which corresponds to levels of beta-oxidation. This is followed by a slower clearance component which reflects incorporation of the tracer into triacylglycerols and phospholipids. In normal myocardium, there is homogeneous distribution of the tracer. During periods of ischemia, the uptake of the agent and the rate of the rapid clearance phase and its magnitude markedly diminish, consistent with diminished oxidative metabolism. Delayed clearance is also observed from ischemic and infarcted regions as a result of decreased fatty acid oxidation. C-11 palmitate exams reveal segmental reductions of blood flow in acutely ischemic myocardium are associated with delayed clearance of 11C-palmitate and an increase in 18-FDG uptake.
The rapid blood clearance of this agent and loss of radioactivity from myocardial tissue due to straight chain fatty acid catabolism by beta oxidation complicate data analysis. A drawback of this agent is that it is affected by plasma substrate concentrations- in the fasting state, when glucose levels are low, fatty acid metabolism is accelerated and C-11 palmitate washout is rapid. Washout decreases markedly with glucose loading. The critical organ is the liver.
Another agent, 1-[11C] acetate can also be used to assess overall oxidative metabolism. The tracer is well extracted by the myocardium and is then converted to acetyl-CoA. Acetyl-CoA is the entry point for all metabolic pathways into the tricarboxylic acid cycle within the mitochondria. Oxidation of acetyl-CoA leads to the production of C-11 labeled CO2 reflecting the activity of the tricarboxylic acid cycle which, under normal conditions, reflects overall oxidative metabolism.
Even in the face of changing patterns of myocardial substrate use, the clearance of 1-[11C] acetate accurately delineated overall oxidative metabolism (ie: unlike C-11 palmitate, this agent is not affected by plasma substrate concentrations). The maintenance of oxidative metabolism in jeopardized myocardium was identified as a prerequisite to recovery of function following revascularization. Myocardial uptake and clearance are homogeneous in normal subjects. In infarcted myocardium, there is decreased uptake and clearance of the tracer due to the reduced oxygen consumption.
The predictive accuracy of this tracer in viability assessment is superior to FDG. In one study, 15% of nonviable myocardial regions demonstrated increased uptake of FDG relative to flow, but diminished levels of oxidative metabolism, while 20% of viable segments (demonstrating normal oxidative metabolism) had reduced FDG accumulation [11].
For Rb-82 imaging in the evaluation of myocardial viability there is unfortunately no association between the severity of a fixed defect on 82Rb imaging and the likelihood of myocardial viability in that area. In defects containing less than 50% relative activity, 30% were shown to be viable by FDG imaging [13].
18-F-Fluoromisonidazole has been shown to have increased retention in hypoxic, non-infarcted myocardium and it can therefore be useful for the selective delineation of hypoxic, but viable tissue.
Finally, (H215O) can be used to determine the water-perfusable tissue index. Contractile function is thought to improve after revascularization only if the perfusible tissue fraction is at least 70%.
REFERENCES:
(1) Radiographics 1999; Jadvar H, et al. SPECT and PET in the evaluation of coronary
artery disease. 19: 915-926
(2) J Nucl Med 2000; Tamaki N, et al. The role of fatty acids in cardiac imaging. 41: 1525-1534
(3) J Nucl Med 1994; Apr: p694
(4) J Nucl Med 1995; Apr: p.581-85
(5) J Nucl Med 1998; Human biodistribution and dosimetry of the PET perfusion agent Copper-62-PTSM. 39: 1958-1964
(6) J Nucl Med 1995; Nov: p.2110
(7) J Nucl Med 1994; Apr: Suppl., 8-14S
(8) J Nucl Med 1995; Sept: p.1550
(9) J Am Coll Cardiol 1992; 20: 559-65
(10) J Nucl Med 1994; Apr: p673
(11) J Nucl Med 1994; Apr. Supp.: 15-22S
(12) J Nucl Med 1995, Sept: p.1550
(13) Radiology 1995; Jan.: p.151-55
(14) Semin Nucl Med 2000; Bax JJ, et al. 18-Fluorodeoxyglucose imaging with positron emission tomography and single photon emission computed tomography: Cardiac applications. 30: 281-298
(15) J Nucl Cardiol 2001; Dilsizian V, et al. Fluorine-18-deoxyglucose SPECT and coincidence imaging for myocardial viability: Clinical and technical issues. 8: 75-88
(16) J Nucl Med 1993; Schelbert H, et al. Positron statement: Clinical use of cardiac positron emission tomography. Position paper of the cardiovascular council of the society of nuclear medicine. 34: 1385-88
(17) J Nucl Med 1993; Garcia EV, et al. What should we expect from cardiac PET? 34: 978-980 (No abstract available)
