The primary goal of physicians with patients presenting with an acute coronary syndrome (ACS) is to choose a treatment strategy that optimizes outcome for each individual patient. Physicians also increasingly are asked to use medical resources in a way that minimizes the economic impact. These decisions, which involve consideration of the cost-effectiveness of an approach, necessarily involve an understanding of an individual patient’s risk for important cardiac events, such as death or myocardial infarction (MI). Interventions that are expensive and that carry their own risk are most beneficial for patients who have the greatest risk of future cardiac events. Conversely, low-risk patients are unlikely to benefit from any intervention, especially one that itself has risks of complications. Knowledge of the determinants of risk in patients presenting with ACS allows physicians to make cost-effective management decisions that most benefit individual patients. In this context, nuclear cardiology techniques offer the physician important tools to help make such decisions because these techniques can assess accurately two of the most important determinants of prognosis in patients with ACS, left ventricular dysfunction and current jeopardized viable myocardium. These factors represent two interactive influences, the extent of permanent damage and the extent of future myocardium at risk. This chapter reviews current data involving the use of nuclear cardiology techniques to determine cardiac risk in ACS and shows how they can be integrated into a comprehensive, rational management strategy.
Left Ventricular Function
Predictive Value
Determination of left ventricular function using radionuclide angiography has been shown to be a powerful predictor of outcome in patients who present with an acute MI. The left ventricular ejection fraction measured by radionuclide techniques probably has been the most valuable index of function because of its reproducibility and its consistent predictive value. In the MPRG (Multicenter Post Infarction Research Group) study, patients who had presented with an acute MI underwent predischarge radionuclide angiography. The 1-year cardiac mortality rate was exponentially inversely related to left ventricular ejection fraction ( Fig. 15-1 ). Cardiac mortality increased, especially after the ejection fraction decreased to less than 40%.
The influence of left ventricular function on outcome has persisted, even as the overall outcome after acute MI has improved with the introduction of thrombolysis and percutaneous coronary interventions. In the TIMI trial, overall cardiac death rate was lower than in the MPRG trial, but the increased relationship between ejection fraction and annual cardiac mortality rate was retained (see Fig. 15-1 ). Similarly, the CAMI study evaluated the outcome of patients with acute MI in the thrombolytic era of the 1990s. These investigators found an inverse relationship between ejection fraction and 1-year cardiac mortality that matched the MPRG findings closely ( Fig. 15-2 ). The differences were greatest at the low end (ejection fraction < 20%), possibly reflecting the ameliorative effects of angiotensin-converting enzyme inhibitor and beta blocker treatment that became the standard of care in the 1990s.
Several other studies have confirmed the important prognostic value of ejection fraction in patients receiving thrombolysis. Simoons and colleagues found that 5-year survival was only approximately 40% in patients with left ventricular ejection fraction less than 30% compared with greater than 90% survival when the ejection fraction was higher than 40%. Similarly, Dakik and coworkers found the risk of cardiac events increased as ejection fraction decreased in patients receiving thrombolysis for acute MI. Event-free survival was approximately 75% for patients with an ejection fraction of 40% compared with survival less than 25% for patients with an ejection fraction less than 40%.
Limitations of Measurement of Left Ventricular Function and Dynamic Changes
Although left ventricular ejection fraction has a powerful ability to predict cardiac mortality after acute MI, it has some limitations. First, although the measurement of ejection fraction itself with radionuclide angiography is highly reproducible and accurate, indices of left ventricular function measured early after MI may change over time, even in the short term. This change may reflect evolving post-MI cardiac physiology, including segmental transient stunning and hyperkinesis. Christian and associates have found that 16% of patients with acute MI showed a rise in ejection fraction of 8% determined by radionuclide angiography between discharge and 6 weeks. At discharge, the ejection fraction in this group was less than that predicted based on the size of the MI, as determined by myocardial perfusion imaging. This suggests that the rise in ejection fraction at 6 weeks reflects a resolution of myocardial stunning. Conversely, 19% of patients showed a decrease in ejection fraction of 8% by 6 weeks. Although remodeling can explain a decrease in left ventricular function over time after acute MI, these patients had a discharge ejection fraction that was greater than that predicted based on the infarct size, suggesting that the decrease in ejection fraction over 6 weeks more likely reflected resolution of early hyperkinesis in the noninfarct zone.
Hibernating myocardium is another factor that may affect left ventricular function in the setting of prior MI. To the extent that it is present, hibernating myocardium is important to identify because coronary revascularization can improve left ventricular function and survival. Although hibernating myocardium may be an important factor in determining prognosis and management in chronic ischemic cardiomyopathy, its role in acute MI is likely to be small. Aside from the transient influences of stunning or hyperkinesis, left ventricular ejection fraction probably primarily reflects the permanent damage done by acute MI, which accounts for its strong prognostic value. It is clear that ejection fraction, reflecting the extent of scar, is a powerful predictor of mortality, but it reflects what has already occurred. Although beta blockers and angiotensin-converting enzyme (ACE) inhibitors may increase survival, it is an unproved hypothesis that revascularization alters outcome in this setting in the absence of stunning, hibernation, or other critical artery stenoses.
An outcome that can be altered by surgical or percutaneous intervention is more likely related to the extent of jeopardized viable myocardium that is present after the insult of the acute MI. As is reviewed subsequently, there are much data to suggest that the presence and extent of jeopardized viable myocardium defined by stress nuclear myocardial perfusion imaging (MPI) is the most important predictor of outcome in patients with acute MI. There is an important interaction between the extent of permanent damage (scar) and that of additional viable myocardium at risk.
Prognostic Value of Jeopardized Viable Myocardium Determined by Stress Nuclear Myocardial Perfusion Imaging
Predictive Value
Because reversible defects on stress MPI accurately identify and quantify jeopardized viable myocardium, this technique can play an important role in assessment of risk after acute MI ( Box 15-1 ). Probably the most consistent observation reported in the literature regarding the prognostic value of stress nuclear MPI is that the presence and extent of transient defects, reflecting jeopardized viable myocardium, predict important cardiac events. , A direct relationship between myocardium at risk identified by nuclear MPI and patients at risk for cardiac events first was reported by Brown and colleagues in 1983. They compared the prognostic value of exercise thallium-201 imaging, exercise treadmill testing, coronary angiography, and clinical data and found that the best predictor of cardiac death or nonfatal MI was the number of segments with transient thallium-201 defects. These early findings were confirmed and expanded by many investigators. Ladenheim and associates found that among clinical and scintigraphic indices, the number of reversible perfusion defects on stress thallium-201 images was the best predictor of future cardiac events. Similar observations were made in a wide clinical spectrum of patients—patients with suspected coronary artery disease (CAD) or known angiographic CAD, patients undergoing noncardiac surgery, patients with remote prior MI, and, as discussed later, patients presenting with unstable angina or acute MI. The powerful predictive value of nuclear MPI is retained whether the stress agent is exercise, vasodilator stress, or an adrenergic agent and regardless of whether the perfusion tracer is thallium-201 ( 201 Tl), technetium-99m ( 99m Tc)-based agents such as sestamibi or tetrofosmin, or positron emission tomography (PET) imaging agents. For each of these modalities of imaging and patient cohorts, the most consistent finding has been that cardiac risk is related directly to the presence and, more importantly, the extent of jeopardized viable myocardium.
BOX 15-1 - •
Increased sensitivity for detecting ischemia and multivessel disease
- •
Increased prognostic value—jeopardized viable myocardium predicts death or myocardial infarction
- •
Ability to localize ischemia to individual coronary territories
- •
Distinguishing infarct zone from noninfarct zone myocardium at risk
- •
Evaluation of left ventricular function and perfusion simultaneously
- •
Can use vasodilator stress as adjuvant, allowing earlier risk stratification
Acute ST-Segment Elevation Myocardial Infarction
Exercise Nuclear Imaging
The prognostic value of exercise MPI in the post-MI setting first was reported by Gibson and coworkers. Predischarge submaximal exercise 201 Tl-MPI was compared with clinical, exercise, and coronary angiographic data for predicting subsequent cardiac events. Reversible 201 Tl defects, defects involving multiple coronary territories, and increased lung 201 Tl uptake (reflecting left ventricular dysfunction) were the most important prognostic MPI variables. Compared with clinical or coronary angiography, these indices were significantly more sensitive for detecting patients at risk for cardiac events ( Fig. 15-3 ). The greater sensitivity for detecting the patient at risk translated into a greater ability to identify the low-risk patient who is unlikely to benefit from further invasive or interventional procedures. Subsequently, many studies confirmed the prognostic value of exercise MPI in patients presenting with an acute MI. , Wilson and colleagues showed that in patients with acute MI and single-vessel CAD, late cardiac events were related to the presence and extent of transient defects on submaximal exercise MPI but not to clinical or exercise electrocardiographic data. Travin and associates have used regression analysis of clinical, exercise electrocardiography, and exercise sestamibi single-photon emission computed tomography (SPECT) MPI variables in patients with acute MI and found that only the number of reversible defects is a significant predictor of cardiac events ( Fig. 15-4 ). As in patients with chronic stable CAD, there is compelling evidence that the risk of future cardiac events in patients with acute MI is strongly related to the presence and, more importantly, the extent of jeopardized viable myocardium. Interesting data using a novel, ischemia-sensitive perfusion agent has confirmed the incremental prognostic value of residual jeopardized viable myocardium post-MI. Iodine-123 beta-methyl iodophenyl pentadecanoic acid ( 123 I-BMIPP) uptake is related to myocardial perfusion but is also sensitive to metabolic changes induced by ischemia. When imaged in conjunction with resting 201 Tl imaging, a mismatch in uptake (normal 201 Tl uptake, reduced 123 I-BMIPP uptake) indicates an area of myocardium with normal resting flow but with a recent history of ischemic insult. Nanasato and coworkers have found that in patients with acute ST-segment elevation MI (STEMI) treated with primary coronary intervention 123 I-BMIPP– 201 Tl mismatch and total 123 I-BMIPP defect score added significant prognostic value to left ventricular function and extent of angiographic CAD for predicting cardiac events. For all-cause mortality, mismatch plus extent of 123 I-BMIPP defects doubled the global chi-square test results of the predictive model compared with cardiac catheterization data.
Indirect Markers of Ischemia
Several indirect markers of ischemia on stress MPI have been shown to have adverse prognostic implications, although they have been incompletely studied in patients with acute MI. Increased 201 Tl lung uptake on exercise MPI has been shown to reflect stress-induced rises in left ventricular filling pressure , and has been associated with severe CAD and resting and exercise-induced left ventricular dysfunction. It has been shown to predict an increased risk of cardiac events in patients with acute MI and with chronic coronary disease. Transient left ventricular dilation on stress compared with rest imaging also has been related to extensive CAD and left ventricular dysfunction, and has been associated with increased risk of cardiac events. More data are needed to understand the full clinical implications of these findings, especially in patients with ACS.
Vasodilator Perfusion Imaging
Vasodilator stress may have a particular advantage over exercise as an adjunct to MPI in patients with acute MI. It produces a greater hyperemic stimulus compared with submaximal exercise and consequently has been shown to have greater sensitivity for detecting CAD when used with stress MPI, an important issue for risk stratification in the post-MI setting. Leppo and coworkers were the first to show that vasodilator stress, using intravenous dipyridamole, with 201 Tl-MPI predicted cardiac events when performed 10 to 16 days after MI. Compared with clinical and radionuclide ventriculographic left ventricular function data, these investigators found that reversible 201 Tl defects were the only significant predictors of late cardiac death or MI and identified 92% of patients at risk for such future cardiac events. Subsequently, other investigators confirmed the predictive value of reversible defects on dipyridamole-MPI, reflecting jeopardized viable myocardium in patients with acute MI. ,
Adenosine Stress Perfusion Imaging
Adenosine also has been found to be valuable as a vasodilator adjuvant for stress MPI after MI. Mahmarian and colleagues , have described the early and late prognostic value of adenosine 201 Tl-SPECT MPI in patients with acute MI. Imaging detected jeopardized viable myocardium in 59% of infarct zones and in 92% of noninfarct zones supplied by a stenotic artery. Angiographic patency did not predict the presence or extent of jeopardized myocardium. In-hospital cardiac events occurred in 43% of patients with significant reversible defects compared with 9% without significant reversibility. Over a mean 16-month follow-up, the best predictors of cardiac events were extent of reversible 201 Tl defects and ejection fraction. The authors showed that adenosine MPI data had significant incremental prognostic value when added to clinical and angiographic data, improving the ability to predict cardiac events twofold to fivefold. The negative impact of MPI ischemia on outcome was additive to that of the left ventricular ejection fraction ( Fig. 15-5 ), consistent with the paradigm that the extent of residual jeopardized viable myocardium and left ventricular function are the primary determinants of outcome after MI.
Dobutamine Stress Perfusion Imaging
Dobutamine adrenergic stimulation offers an alternative to vasodilator stress as an adjunct to MPI in patients unable to exercise. This alternative is particularly useful for patients with contraindications to vasodilators, such as bronchospasm, recent caffeine or methylxanthine exposure, high-degree atrioventricular block, or hypotension. Previous reports have described sensitivity and specificity for detecting CAD comparable to exercise or vasodilator stress nuclear imaging. Several studies have shown that dobutamine stress MPI has significant prognostic value in patients with stable CAD or before noncardiac vascular surgery. Data in stable coronary disease suggest that, as with exercise MPI, the risk of death or MI is related directly to the extent of jeopardized viable myocardium on dobutamine stress MPI. Few data are available, however, for dobutamine MPI in the post-MI setting. In contrast to vasodilators, dobutamine produces a much more marked increase in heart rate and blood pressure, leading to induction of true ischemia in the setting of coronary lesions, rather than just heterogeneity of hyperemia. Consequently, although dobutamine stress MPI can be performed safely, it needs to be applied more cautiously in the post-MI setting, especially the early post-MI setting.
Coma-Canella and colleagues have performed dobutamine stress 201 Tl-SPECT and radionuclide angiography a mean of 16 days after MI. They found that extent of ischemia on SPECT imaging correlated with abnormal dobutamine-induced regional wall motion. No angiographic or prognostic data were reported. Elhendy and associates have reported a sensitivity of 74% and 71% for detecting remote and infarct-related coronary disease, respectively, using dobutamine stress 201 Tl-SPECT in 71 patients more than 3 months after acute MI. Specificity was 80% to 83%. These authors also reported similar sensitivities, specificities, and predictive values for dobutamine stress MPI and echocardiography in detecting remote and infarct-related coronary disease in patients with prior MI. In this cohort, most of the patients were studied several years after MI. In contrast, Lancellotti and coworkers have compared dobutamine stress MPI and echocardiography performed a mean of 5 days after MI. In this small select cohort of 75 patients, there were no serious dobutamine-related complications. MPI was a sensitive and specific predictor of infarct-related stenoses (70% and 83%) and multivessel disease (67% and 93%) and was comparable to stress echocardiography.
Dobutamine stress MPI seems to assess accurately post-MI coronary anatomy, and there is internal consistency between dobutamine-induced perfusion defects and inducible wall motion abnormalities. Determination of the prognostic implications and the safety of dobutamine stress MPI awaits further data, however.
Early Post–Myocardial Infarction Risk Stratification
In addition to its greater sensitivity for detecting CAD, vasodilator stress MPI has other advantages that can allow it to play an important role in the early in-hospital management of post-MI patients. In contrast to exercise, vasodilator stress induces only modest changes in determinants of myocardial oxygen demand. In addition, the hemodynamic effects are brief (adenosine) or rapidly reversible (dipyridamole). , As a consequence, risk stratification with vasodilator stress MPI can be performed safely, potentially much earlier than exercise after acute MI. Management decisions can be made sooner than the standard 5- to 7-day post-MI, predischarge evaluation, potentially shortening hospitalization and reducing costs. In addition, identifying high-risk patients and directing appropriate treatment sooner can prevent early cardiac events. Brown and colleagues first reported a series of 50 patients who underwent dipyridamole 201 Tl-MPI 1 to 4 days (mean, 2.6 days) after acute MI. No serious adverse effects occurred with dipyridamole administration. Clinical, electrocardiographic (ECG), cardiac catheterization, and 201 Tl-MPI data were analyzed, and the only significant predictor of in-hospital ischemic cardiac events was the presence of reversible defects in the infarct zone. Nine of 20 patients (45%) with infarct zone reversible defects had in-hospital ischemic cardiac events compared with 0 of 30 patients without ( P = .0001). Over a mean 12-month follow-up period, there were three additional cardiac events in patients with reversible defects, whereas patients without reversible defects remained free of cardiac events. It seems that early risk stratification with dipyridamole MPI could identify high-risk patients, who could be referred early to invasive procedures and revascularization, and low-risk patients, who could be discharged early safely without further interventions.
These pilot data led to a much larger multicenter study involving 451 patients that was designed to compare the prognostic value of dipyridamole 99m Tc-sestamibi SPECT MPI performed 2 to 4 days after acute MI with standard submaximal exercise MPI obtained at 6 to 12 days. Confirming the pilot safety data, no significant adverse effects were attributable to the dipyridamole infusion. Clinical and stress test data were compared with MPI data, including the following: a summed stress score, reflecting the size and severity of the defect on the stress images; and a summed difference score, reflecting the degree of reversibility between stress and rest images. The dipyridamole MPI summed stress and reversibility scores were significant multivariate predictors of in-hospital cardiac events. Patients were followed for a mean of 2 years. The dipyridamole MPI summed stress, rest, and difference scores were each significant multivariate predictors of future cardiac death or MI. Consistent with the theme emphasized in this review, indices of scar (summed rest score) and ischemia (summed difference score) were significant determinants of outcome after MI. Dipyridamole MPI was better able to risk-stratify patients than submaximal exercise MPI. This was manifested as a greater ability to separate low-risk from high-risk patients. Not only were significant prognostic data available earlier with dipyridamole MPI compared with submaximal exercise MPI, but the information was superior at separating high-risk from low-risk patients.
There was an important interaction between the size of the initial stress defect and the degree of reversibility for determining individual patient risk. For a given size of stress defect, cardiac risk increased as the degree of reversibility, reflecting jeopardized viable myocardium, increased ( Fig. 15-6 ). Patients with a small defect (low summed stress score) had an overall annual cardiac death or MI rate of 2%, but this decreased to 0% in patients with little or no ischemia and increased to 4% in patients whose defect was primarily reversible. The interaction was greatest in patients with intermediate-sized stress defects. The overall annual event rate was 5%, but knowing this degree of jeopardized viable myocardium allowed further risk stratification—0% with no or small ischemia increasing to 6% with intermediate ischemia and 17% with extensive ischemia. The interaction was least in patients with large stress defects because the event rate remained high in patients with extensive infarction but little or no ischemia.
The predictive value of early post-MI vasodilator stress MPI has been confirmed by the larger INSPIRE trial, comprised of 728 patients undergoing gated adenosine MPI a median of 3 days post-MI. MPI results stratified patients into low-, intermediate-, and high-risk patients, with 1-year death-MI rates of only 1.8% in patients with small perfusion defects and increasing progressively as the perfusion defect exceeded 20% of the left ventricular (LV) myocardial volume (9.2%; 1-year death-MI) and also had moderate or greater ( > 10%) ischemia (11.6%; 1-year death-MI). This study also incorporated the evaluation of LV function into the analysis and found that the cardiac event rate increased as LV ejection fraction (LVEF) decreased. Multivariate analysis has shown that total stress perfusion defect size, which incorporates scar plus ischemia, is the best and only significant predictor of death or MI.
These studies show that vasodilator stress MPI can be performed safely early in the post-MI period and provides powerful prognostic data not only sooner than, but also superior to, submaximal exercise MPI data. It allows identification of high-risk patients, who are candidates for early intervention, and low-risk patients, who can be considered for early discharge.
Selective Versus Nonselective Invasive Approach After an Acute Myocardial Infarction (ST-Segment Elevation)
Implicit in an ischemia-guided selective approach to cardiac catheterization and revascularization for the patient with acute MI is the assumption that patient outcome will be at least as good as a nonselective approach, in which every patient is referred for catheterization, and revascularization decisions are based on coronary anatomy. There are now substantial data to support such an assumption ( Box 15-2 ). The TIMI IIB trial compared the outcome of 3262 patients presenting with acute STEMI who were randomized to angiography and anatomy-guided revascularization versus a conservative, or ischemia-guided approach, in which patients were referred to angiography and revascularization only if there was symptomatic or exercise-induced electrocardiographic evidence of ischemia. The composite end point of death or MI at 6 weeks occurred in 10.9% of the invasive group compared with 9.7% of the ischemia-guided group ( P = not significant). Although the more sensitive nuclear MPI was not used in the conservative ischemia-guided group, the outcome was at least as good as in the nonselective invasive group. Similar findings were reported from the SWIFT trial, which compared the outcome of patients with acute ST-segment elevation treated with thrombolysis and randomized to early invasive, anatomy-guided intervention versus conservative, ischemia-guided intervention. The death or MI rate at 1 year was 19.1% versus 16.6% in the invasive and conservative groups, respectively ( P = not significant). The TOPS study included patients who presented with acute STEMI, received thrombolysis, had angiographically significant infarct vessel disease (>50% stenosis), but had a negative stress functional study (generally, 201 Tl-MPI). Patients were randomized to medical treatment without intervention or to delayed coronary angioplasty at days 4 to 14. The infarct-free survival at 12 months was 100% in the medical treatment group versus 89% in the angioplasty group ( P = .07). The infarcts that occurred in the angioplasty group all were procedure-related. Functional noninvasive risk stratification was able to identify a population of patients at low risk for cardiac events treated medically who are not benefited (and possibly harmed) by intervention.
BOX 15-2 
