Background

The first description of myocardial stunning was by Heyndricks in 1975, subsequently modified by Braunwald and Kloner in 1982. The first description of hibernation was by Diamond in 1978, subsequently used by Rahimtoola in 1985. Since 1980, and with the advent of imaging, there has been an increasing interest in this topic with an ever-increasing number of publications ( Figure 15-1 ). In 1989, Rahimtoola wrote, “hibernating myocardium is a relatively uncommon response to reduced myocardial blood flow (MBF) at rest whereby the heart down-regulates its function to the extent that the MBF and function are once again in equilibrium and, as a result, neither necrosis nor ischemic symptoms are present.” That description is probably still true in some, but not all, patients.

The number of publications on myocardial viability over a 30-year period.

There are several reasons why patients with CAD develop LV dysfunction. These include infarction (scar), hibernation, stunning, and remodeling, which refers to regional dysfunction in segments remote from culprit zones. Obviously, and not uncommonly, the dysfunction can be due to a coexisting problem, such as hypertension, valvular heart disease (either preexisting, such as aortic valve disease, or developing as a result of LV dysfunction, such as mitral regurgitation), or dilated cardiomyopathy, just to mention the most common causes.

Simply stated, the dysfunction can be described as either reversible, if the MBF is restored to normal (stunning and hibernation are prototypes of viable myocardium), or irreversible (scar); the fate of remodeled areas will depend on whether the initiating process was scar or hibernation/stunning (assuming there are no concomitant causes of dysfunction and coronary revascularization is feasible, complete, and maintained for a prolonged period of time).

Many PET studies validated Rahimtoola’s observation and showed a reduction in resting MBF and increased glucose utilization (measured via F-18-fluorodeoxyglucose, FDG), the flow-metabolism mismatch. The normal myocardium has several sources of energy production, but it uses glucose or fatty acids preferentially. Each mole of glucose produces 2 moles of ATP in the cytoplasm when metabolized anaerobically and 36 moles of ATP when metabolized aerobically in the mitochondria. Glycolysis generates 17% more ATP than fatty acids for an equal amount of oxygen used and is the reason why metabolism is shifted to glucose utilization in ischemic states.

Other studies, including those in animal models, however, challenged the temporal relationship between changes in MBF and LV dysfunction. What emerged is the following scenario: early after a severe stenosis develops, regional LV function decreases even though the resting MBF is normal. But because the hyperemic MBF is blunted (due to the severe stenosis), any increase in demand will produce an episode of stunning (regional dysfunction that persist after flow returns back to normal), which is a flow-function mismatch (reduced function but normal resting MBF). This stunning, however, is not isolated to a single episode; rather it keeps repeating itself with such frequency that the function never has a chance to return to normal. As function deteriorates, the MBF decreases because the demand is lower (setting the bar progressively lower). Unlike the original hypothesis, the reduction in MBF is a secondary phenomenon rather than the initiating element. Repetitive stunning, as this process is known, presents itself as chronic LV dysfunction. Adaptation results in structural changes in the myocardium that are collectively referred to as de-differentiation, along with degeneration in the intracellular and extracellular structures. The intensity of these changes and the degree of LV dilatation will determine how completely and rapidly the function will normalize after coronary revascularization. Because of the tenuous relationship between supply and demand, these patients are at a higher risk for infarction and death. Finally, some studies have suggested that sudden death due to ventricular arrhythmias is due to functional dennervation involving segments even larger than those with flow or metabolic changes (so called perfusion–innervation mismatch).

Various imaging methods have relied on a number of different principles to elucidate what is viable and what is not. These studies have come to be known as viability testing. The implication being, that the techniques can identify patients with CAD and LV dysfunction who could benefit from coronary revascularization because the dysfunction is secondary to hibernation or stunning. As a caveat, there is no standard definition of what “benefit” means. “Benefit” has been previously defined as an improvement in regional function, EF, symptoms, quality of life, and/or survival. There are also no guidelines on when to assess for “benefit,” because, as mentioned earlier, degenerative changes may take some time to recover, as long as 6 to 12 months!

Patient selection

Proper candidates for viability testing are patients with severe LV dysfunction (EF < 35%) and severe CAD. This is not a test for patients with a normal EF, diastolic heart failure, patients with mild CAD but with LV dysfunction, or patients with dilated cardiomyopathy. It may seem simple and straightforward, but we continue to be surprised by the nature of some of the requests!

This book will not discuss the role of other imaging modalities as they are discussed in the other companion books.

Nuclear imaging for viability assessment

The use of PET is discussed in Chapter 18 . What follows is a discussion on the use of single photon tracers, Tl-201– and Tc-99m–labeled tracers (sestamibi and tetrofosmin). Both Tl and Tc-tracers are considered flow tracers and their sequestration inside the cells (cytoplasm and mitochondria, respectively) require intact cell membrane (hence, the technique is sometimes described as testing cell membrane integrity). The uptake is proportional to regional MBF at rest, in fact, it may overestimate resting MBF. There is one difference, however, between the tracers and that is the process of redistribution; thallium redistributes and sestamibi/tetrofosmin do not (or do so minimally). That difference, therefore, dictates how imaging is done. With thallium, rest and 4-hour redistribution, stress and 4-hour delayed reinjection imaging, or stress, 4-hour delayed, and 24-hour delayed imaging can be performed. Tc tracers are used for rest, stress and rest, or rest and stress imaging. Many studies have found that thallium results are comparable to Tc tracers for viability testing, but we have used thallium in most of our studies (with or without sestamibi) using a protocol we have used for almost two decades. This is because we believe that a reduction in resting MBF does occur in some patients, or myocardial segments in a given patient, that can be missed by Tc tracers.

Our viability protocol

We use rest and 4-hour delayed thallium imaging in patients with known severe CAD by coronary angiography or hemodynamically unstable patients. In these patients, we do not use stress testing as the question being asked is simple; is there sufficient viable myocardium that revascularization would benefit the patient? We continue to do these studies at the bedside in unstable patients using a mobile gamma camera and planar imaging.

In the remaining patients, we perform rest and 4-hour delayed thallium imaging followed by vasodilator sestamibi imaging. We compare the two thallium images to each other and compare the better of the two to the sestamibi images (all images are with gated SPECT). At one time, we gave a low dose of dobutamine after the first set of sestamibi images were acquired and acquired a second set of sestamibi images (fourth set for the study) during the infusion to study contractile reserve. This is no longer done because the additional information was not helpful. We have, thus, limited the study to three sets of images.

On evaluating the images, we examine the L/H ratio and look for any evidence of redistribution on the delayed thallium images and stress-induced ischemia. One way of looking at the segments showing redistribution on the 4-hour delayed rest thallium images is that they are a poor man’s PET perfusion–metabolism mismatch; the initial image is a flow image, but the delayed image is a metabolic image related to tracer exchange across metabolically active cell membranes. We use automated methods to quantify tracer uptake. If the two thallium images reveal severe and extensive fixed defects, we omit the stress portion. If the initial thallium images are normal, we omit the 4-hour delayed images.

Characterization of dysfunctional myocardium

When reviewing perfusion and regional function, we define segments (or more practically vascular territories or walls) as follows:

• 1.
  

  Scar (nonviable): Severe perfusion defect on initial thallium image (< 50% tracer activity), no evidence of redistribution on the delayed thallium images, and no stress-induced ischemia (if a stress study was obtained). These segments have severe wall motion/thickening abnormalities at rest.

  
  
• 2.
  

  Hibernating: Reduced tracer activity on initial thallium images with evidence of redistribution on delayed thallium imaging. The stress images, if available, will show a more severe or extensive abnormality. These segments have severe wall motion/thickening abnormalities at rest.

  
  
• 3.
  

  Stunning: Reduced tracer activity on initial thallium images with no redistribution on delayed thallium images but with stress-induced ischemia (a stress study is needed to define stunning). These segments have severe wall motion/thickening abnormalities at rest.

  
  
• 4.
  

  Remodeled: These segments have reduced initial activity on thallium images with no redistribution on the delayed thallium images and no stress induced ischemia.

The entire protocol can be completed in 5 hours and provides comprehensive information on rest and stress perfusion and function. We do not routinely give nitroglycerin before thallium injection, but most of our patients are on long-acting nitrates. Further, we do not image at 24 hours as the image quality by then is poor and unreliable.

In some segments (walls or patients), there is slight redistribution on the thallium images or a slight stress-induced ischemia on the sestamibi images, but the activity level at its best is still very low. For example, the activity increases from 10% to 20% from the initial to the delayed thallium images. Twenty percent is still too low to have an impact on a positive outcome. Thus, not all reversible defects are created equal. Similarly, not all fixed defects are the same either. Fixed defects with greater than 50% activity have a sufficient amount of viable myocardium (subepicardial), which could be quite useful in preventing further remodeling, serious arrhythmias, and preserving diastolic function. They can even preserve the systolic function with exercise, which may produce symptomatic improvement, although it may not produce an improvement in the regional function at rest.

Tracer activity and recovery of function exist on a continuum. A good rule of thumb is that 80% tracer activity corresponds to an 80% chance for improvement and 40% represents a 40% chance for improvement with coronary revascularization. In general, the presence of ischemia is greater than 90% accurate in predicting functional recovery, if good target vessels are present. This is the reason for including a stress study in the protocol; reversibility is more commonly detected with stress than at rest.