Chapter 38 Assessment of Myocardial Viability with Positron Emission Tomography

INTRODUCTION

Positron emission tomography (PET) enables quantification of myocardial perfusion, glucose utilization, fatty acid uptake and oxidation, oxygen consumption, contractile function, and presynaptic and postsynaptic neuronal activity. One of the first clinical applications of PET was the assessment of myocardial viability, and it has been generally regarded as the gold standard technique for this indication. The most common approach with PET is comparing myocardial metabolism and perfusion using FDG and ammonia or other perfusion tracers.¹ It has been shown that PET can predict regional and global recovery of wall motion, symptoms, functional outcome, and prognosis after revascularization. Also, other tracers and newer techniques are available, such as gating and hybrid imaging using PET/computed tomography (CT) scanners. However, there are less data about their clinical value in the assessment of viability.

METHODS FOR ASSESSING MYOCARDIAL VIABILITY WITH PET

Myocardial Perfusion Tracers

Estimates of regional perfusion themselves have also been used for predicting functional recovery of dysfunctional myocardium. Several studies have investigated the use of nitrogen-13 (¹³N)-labeled ammonia ² and rubidium-82 (⁸²Rb) in the assessment of viability.^3,⁴ ⁸²Rb is a potassium analog that is similar to thallium-201 (²⁰¹Tl), and its uptake has been shown to be a function of both blood flow and myocardial cell integrity. Oxygen-15 (¹⁵O)-labeled water offers a unique feature, since the fraction of tissue that can exchange water can be measured, and this has been used as a marker of myocardial viability.^5,⁶ These techniques, however, are not widely used clinically.

Fluorine-18-Deoxyglucose (See Chapter 40)

Physiologic Basis

Normal myocardium uses a variety of energy-producing substrates to fulfill its energy requirements.⁷ In the fasting state, free fatty acids (FFA) are mobilized in relatively large quantities from triglycerides stored in adipose tissue. Thus, the increased availability of FFA in plasma makes them the preferred energy-producing fuel in the myocardium.⁷ In the fed state, however, the increase in plasma glucose and the subsequent rise in insulin levels significantly reduce FFA release from adipose tissue and, consequently, its availability in plasma. This leads to an increased utilization of exogenous glucose by the myocardium.⁸

FFA metabolism via β-oxidation in the mitochondria is highly dependent on oxygen availability, and thus it declines sharply during myocardial ischemia.^9,¹⁰ Under this condition, studies in animal experiments⁹ and in humans¹¹ have shown that the uptake and subsequent metabolism of glucose by the ischemic myocardium is markedly increased. This shift to preferential glucose uptake may play a critical role in the survival of functionally compromised myocytes (i.e., stunned and hibernating), inasmuch as glycolytically derived high-energy phosphates are thought to be critical for maintaining basic cellular functions. Consequently, noninvasive approaches that can assess the magnitude of exogenous glucose utilization play an important role in the evaluation of tissue viability in patients with myocardial dysfunction due to coronary artery disease (CAD). For these metabolic adaptations to occur, sufficient nutrient perfusion is required to supply energy-rich substrates (e.g., glucose) and oxygen, and for removal of the byproducts of glycolysis (e.g., lactate and hydrogen ion). A prolonged and severe reduction of myocardial blood flow rapidly precipitates depletion of high-energy phosphate, cell membrane disruption, and cellular death. Therefore, assessment of regional blood flow also provides important information regarding the presence of tissue viability within dysfunctional myocardial regions.

With PET, regional glucose uptake is assessed with fluorine-18 deoxyglucose (FDG), a marker of exogenous glucose uptake that provides an index of myocardial metabolism and thus cell viability. The technique was first described by Tillisch et al.¹ After intravenous administration, FDG traces the initial transport of glucose across the myocyte membrane and its subsequent hexokinase-mediated phosphorylation to FDG-6-phosphate (Fig. 38-1).¹² Since the latter is a poor substrate for further metabolism and is rather impermeable to the cell membrane, it becomes virtually trapped in the myocardium.

Figure 38-1 Metabolic fate of [¹⁸F]deoxyglucose as a marker of exogenous glucose utilization.

Protocols

Because dysfunctional myocardium that improves functionally after revascularization must retain sufficient blood flow and metabolic activity to sustain myocyte viability, the combined assessment of regional blood flow and glucose metabolism appears most attractive for delineating myocardial viability. With this approach, regional myocardial perfusion is first evaluated following the administration of ¹³N-ammonia,⁸²Rb, or ¹⁵O-water. Since information regarding the magnitude of both stress-induced ischemia and resting viability is important for management decisions, the ideal approach should include both rest and stress perfusion imaging. However, the selection of the approach (i.e., rest versus stress/rest) should be tailored to the clinical question being addressed in an individual patient (Fig. 38-2). Regional glucose uptake is then assessed with FDG.

Figure 38-2 Myocardial viability protocols using PET and the hybrid SPECT/PET approach. FDG, fluorine-18-deoxyglucose; PET, positron emission tomography; SPECT, single-photon emission computed tomography; Tc-99m; technetium-99m.

Some investigators have also used absolute quantification of glucose uptake.^13,¹⁴ These studies have documented reasonable results, but absolute quantification of glucose uptake using dynamic imaging and kinetic modeling does not appear to improve diagnostic accuracy, likely because of high variability in glucose utilization rates between individual patients.

Patient Preparation for FDG Imaging

As mentioned, utilization of energy-producing substrates by the heart muscle is largely a function of their concentration in plasma and hormone levels (especially plasma insulin, insulin/glucagon ratio, growth hormone, and catecholamines) and oxygen availability for oxidative metabolism. For a detailed step-by-step description of the available methods for FDG imaging, the reader should review the Guidelines for PET Imaging published by the American Society of Nuclear Cardiology and the Society of Nuclear Medicine ¹⁵ or European Council of Nuclear Cardiology.¹⁶ These approaches for FDG imaging include:

1. Fasting: This is the simplest method because it does not require any substrate manipulation. With this approach, ischemic but viable tissue shows as a “hot spot” as a result of the preferential FFA utilization by normal (nonischemic) myocardium. While imaging interpretation would seem straightforward, physiologic heterogeneity of glucose uptake with the lack of tracer uptake in normal (reference) myocardium may occasionally lead to an overestimation of the amount of residual viability within a dysfunctional territory. Indeed, the predictive accuracy of this approach is lower than with the glucose-loaded approach.

2. Glucose loading: This is the most commonly used approach to FDG imaging. The goal of glucose loading is to stimulate the release of endogenous insulin in order to decrease the plasma levels of FFA and to facilitate the transport and utilization of FDG. Patients are usually fasted for at least 6 hours and then receive an oral or intravenous glucose load. Patients, especially diabetics, may require the administration of intravenous insulin to maximize myocardial FDG uptake. With this approach, image quality is generally of diagnostic quality and the reported diagnostic accuracy very good.¹⁷

3. Hyperinsulinemic-euglycemic clamp: This approach is technically demanding and time-consuming. However, it provides the highest and most consistent image quality.^18,¹⁹ Because it is technically demanding, most laboratories reserve this approach for challenging conditions (e.g., diabetes and severe congestive heart failure [CHF]).

4. Acipimox: This nicotinic acid derivative inhibits peripheral lipolysis, thereby reducing plasma FFA levels and (indirectly) forcing a switch to preferential myocardial glucose utilization. The drug can be given in combination with a glucose load 60 minutes prior to FDG administration. The approach is practical and provides consistent image quality.^19,²⁰ However, Acipimox is not available for clinical use in the United States, and some centers have used nicotinic acid instead.²¹

Myocardial Perfusion and Glucose-Loaded FDG Patterns

Using the sequential perfusion FDG approach, four distinct perfusion-metabolism patterns can be observed in dysfunctional myocardium:

1. Normal perfusion associated with normal FDG uptake

2. Reduced perfusion associated with preserved or enhanced FDG uptake (so-called perfusion-metabolism mismatch), reflecting myocardial viability (Fig. 38-3)

3. Proportional reduction in perfusion and FDG uptake (so-called perfusion-metabolism match), reflecting nonviable myocardium (see Fig. 38-3)

4. Normal or near-normal perfusion with reduced FDG uptake (so-called reversed perfusion-metabolism mismatch) (Figs. 38-4 and 38-5)^22,²³

Figure 38-3 Cardiac PET images in selected short-axis, vertical long-axis, and horizontal long-axis views demonstrating perfusion-metabolic match (left panel) and mismatch (right panel) patterns of myocardial viability with PET. Images of myocardial perfusion were obtained with nitrogen-13-ammonia and images of metabolism with FDG. FDG, fluorine-18-deoxyglucose.

Figure 38-4 Cardiac PET images of a human heart in short-axis views obtained in corresponding midventricular levels before and after coronary artery bypass surgery (CABG). Images of blood flow (left column) were obtained with [¹³N]ammonia and images of metabolism (left column) with FDG. Top, A patient with three-vessel CAD and a severe wall motion abnormality in the anterior and septal walls. The resting perfusion images (left) demonstrate a small perfusion defect in the anterior wall and near-normal perfusion in the interventricular septum. A technetium-99m-sestamibi perfusion scan (not shown) demonstrated exercise-induced ischemia in these regions. Glucose metabolism in the anterior and septal walls is markedly reduced compared to normal myocardium (lateral wall) (reverse perfusion-FDG mismatch pattern). Bottom, Same patient 4 weeks after CABG. Blood flow and glucose metabolism are largely homogeneous. A two-dimensional echocardiogram after CABG demonstrated improvement in systolic wall motion in the anterior and septal regions.

Figure 38-5 Positron emission tomography images of a dog heart in short-axis views obtained at corresponding midventricular levels obtained after reperfusion after four 5-minute LAD coronary occlusions, each followed by 5 minutes of reperfusion. Images of blood flow (left column) were obtained with ¹³N-ammonia, and images of glucose metabolism (middle column) with FDG. Images of oxidative metabolism (MVO₂) (right columns) were obtained with ¹¹C-acetate; the early phase denotes delivery of the tracer to the myocardium; the late phase represents regional washout of the tracer through the tricarboxylic acid cycle (myocardial oxidation). Top, Corresponding midventricular short-axis sections of regional blood flow, glucose, and MVO₂ 4 hours post-reperfusion. The flow images (left) demonstrate near-normal perfusion in the stunned regions (i.e., anterior and anteroseptal). However, stunned regions demonstrated reduced FDG utilization (arrow) and slow clearance of ¹¹C-acetate (impaired oxidation) relative to normal myocardium (lateral wall). Middle, 1 day post-reperfusion. Myocardial perfusion in stunned myocardium is near normal, glucose uptake (arrow) remains depressed, and the MVO₂ is still lower (arrow) than in normal myocardium. Bottom, 1 week after reperfusion. Blood flow, glucose uptake, and MVO₂ are largely homogeneous. Wall motion and metabolism demonstrated a parallel recovery with time. ¹¹C, carbon-11; FDG, fluorine-18-deoxyglucose; LAD, left anterior descending coronary artery; MBF, myocardial blood flow; MVO₂, myocardial oxygen consumption; ¹³N, nitrogen-13.

(From Di Carli MF, Prcevski P, Singh TP, et al: Myocardial blood flow, function, and metabolism in repetitive stunning, J Nucl Med 41:1227–1234, 2000.)

The patterns of normal perfusion and metabolism or of a PET mismatch identify potentially reversible myocardial dysfunction, whereas the PET match pattern identifies irreversible myocardial dysfunction. The reversed perfusion-FDG mismatch has been described in the context of repetitive myocardial stunning ²² and in patients with left bundle branch block.²³ Images are usually interpreted semiquantitatively, but relative quantitation of regional myocardial perfusion and FDG tracer uptake and their difference can be helpful to objectively assess the magnitude of viability (Fig. 38-6).

Figure 38-6 Quantification of relative myocardial perfusion and FDG uptake using circumferential profile analysis. FDG, fluorine-18-deoxyglucose; Rb-82, rubidium-82.

(Courtesy of Dr. Cesar Santana, Emory University.)

Special Considerations for the Hybrid Myocardial Perfusion SPECT and FDG PET Approach

In current clinical practice, FDG PET images are often performed and interpreted in combination with SPECT myocardial perfusion images (see Fig. 38-2). The interpretation of the specific viability patterns shown in Figures 38-3 and 38-4 should be performed carefully, especially when comparing non–attenuation-corrected SPECT myocardial perfusion images with attenuation-corrected FDG PET images. Myocardial regions showing an excessive reduction in tracer concentration due to attenuation artifacts on the perfusion images, such as the inferior wall or the anterior wall in females, may result in falsely positive perfusion-FDG mismatches. Two approaches have proved useful for overcoming this limitation. First, because assessment of viability is relevant only in myocardium with regional contractile dysfunction, gated SPECT or gated PET images offer means for determining whether apparent perfusion defects are associated with abnormal regional wall motion. Second, quantitative analysis of regional myocardial perfusion using polar map displays that are compared to tracer-specific and gender-specific (for SPECT images) databases may be a useful aid to the visual interpretation.

Gated FDG Study

Parameters of global and regional left ventricular (LV) function derived from gated FDG PET images correlate closely with those obtained by MRI.²⁴ As mentioned earlier, gated images are particularly useful when FDG PET patterns are interpreted in relation to non-attenuation-corrected SPECT perfusion images.²⁵ In addition, measures of global LV function and remodeling (i.e., LVEF and volumes) are also useful for predicting improvement in LV function after revascularization.²⁶ Myocardial infarction (MI), especially one that is large and transmural, can produce alterations in both the infarcted and noninfarcted regions that result in changes in LV architecture known as remodeling.²⁷ Increased LV volumes and cavity size are important predictors of poor outcome in patients after infarction²⁸ and may offset the potential benefits from revascularization on ventricular function and survival even if there is evidence of viable (ischemic) myocardium.²⁹^–³²

Hybrid FDG Positron Emission Tomography and Computed Tomography

Since modern PET scanners now include a multislice CT scanner, an attractive approach could be performing FDG PET imaging with CT (Fig. 38-7). CT could be used for attenuation correction, calcium score measurement, and visualization of coronary arteries and accompanying stenoses. Recently, multislice CT has been found to have sufficient accuracy to image coronary arteries and assess the coronary stenoses.³⁰ An example of such a viability study using FDG and PET/CT hybrid imaging is displayed in Figure 38-7. The application of contrast-enhanced CT angiography in heart failure patients who need viability investigation may be problematic, owing to advanced coronary atherosclerosis and concomitant problems such as kidney disease. The clinical value of hybrid imaging is at this time still open, and feasibility needs to be explored.

Figure 38-7 Example of viability study using FDG and PET/CT hybrid imaging. FDG PET images are overlaid with CT angiography. Scarred region is visualized as blue color. CT, computed tomography; FDG, fluorine-18-deoxyglucose; PET, positron emission tomography.

Carbon-11-Acetate

Physiologic Basis

The consumption of oxygen by the myocardium is required for the continued generation of high-energy phosphates to maintain contractile function and other cellular processes. As mentioned, the heart uses a variety of substrates to support overall oxidative metabolism and thus fulfill its energy requirements.⁷ PET imaging with ¹¹C-acetate offers a direct approach to evaluating myocardial oxygen consumption (MVO₂) and oxidative metabolism.^31,³² Indeed, the clearance of ¹¹C-acetate from myocardium on PET imaging reflects overall regional myocardial oxidative metabolism. Acetate is a short-chained fatty acid that is readily extracted by myocardial tissue. The extracted acetate is then converted into acetyl-CoA within the mitochondria and undergoes near-complete oxidation by the tricarboxylic acid (TCA) cycle and oxidative phosphorylation.