[ Scientific Activities - Actividades Científicas ]Noninvasive Predischarge Risk Assessment in Patients With Uncomplicated Myocardial Infarction George A. Beller, M.D.
Chief Cardiovascular Division Department of
University of Virginia Health System
Charlottesville, Virginia, U.S.A.
|Noninvasive risk assessment|
|Vasodilator stress imaging|
There has been a great deal of interest in determining the most cost-effective approach to risk stratification after acute myocardial infarction (1-7). The major determinants of prognosis after acute myocardial infarction relate to the degree of left ventricular dysfunction as evidenced by a depressed ejection fraction and extent of wall motion abnormalities, extent of residual jeopardized myocardium both in the distribution of the infarct-related artery or in remote myocardium, and the arrhythmogenicity potential of the left ventricle. The extent of jeopardized myocardium can best be evaluated by determination of extent and severity of inducible ischemia by noninvasive techniques and/or extent of angiographic coronary artery disease. Certain high-risk clinical variables have been identified that are associated with an increased risk of subsequent cardiac death or reinfarction. These variables are summarized in Slide #1.
Invasive Versus Conservative Strategies
Controversy exists concerning the most cost-effective approach to risk stratification in the thrombolytic era and the age of coronary intervention. Many patients who are successfully reperfused and have uncomplicated courses are candidates for early hospital discharge. The issue is whether or not "routine" coronary angiography performed early in the course of hospitalization yields superior information with respect to prognostication than exercise testing or exercise pharmacologic stress imaging. Results of certain large randomized clinical trials have indicated that either immediate or delayed elective angioplasty of the infarct-related vessel in patients without recurrent angina did not improve outcome or enhance left ventricular function (8-11). Michels and Yusuf (12) combined data from five trials that examined angioplasty performed more than 12 hours after myocardial infarction to patients who received only thrombolytic therapy and reported no difference in mortality rate or nonfatal infarction at either 6 weeks or 1 year of follow-up. In the TOPS study, Ellis et al (13) showed no functional or clinical benefit from routine late angioplasty in postinfarction patients receiving thrombolytic therapy who had negative functional test results for ischemia prior to randomization to medical therapy or elective revascularization.
The TIMIIII-B study specifically compared an early invasive strategy with a more conservative strategy in patients with unstable angina or nonQ-wave myocardial infarction (14). At 1 year, no difference in death or nonfatal infarction rate was seen between groups, although a substantial number of patients in the conservative group crossed over for subsequent revascularization (64%) (15).
The VANQWISH trial examined outcomes following acute nonQ-wave myocardial infarction comparing a conservative versus an invasive approach using a randomized protocol (16). In the conservative arm, patients underwent stress 201Tl planar imaging with symptom-limited treadmill exercise or dipyridamole vasodilator stress imaging. Coronary angiography was only performed in this group for those patients developing spontaneous postinfarction angina associated with ischemic ST-segment changes or exercise-induced ST-segment depression, evidence of myocardial ischemia in two or more vascular territories, or increased lung thallium uptake on 201Tl scintigraphy. Of interest was that the death or recurrent myocardial infarction rate was higher in the invasive group at discharge, 1 month and 1 year, although the mortality curves tended to converge at the end of follow-up at 44 months (Slide #2). Thus, the results of the randomized trials cited above, which compared a routine invasive approach to a "selective" angiographic approach showed no difference in outcomes.
In contrast to these prior studies, the FRISC-II study showed a decrease in the composite endpoint of death or myocardial infarction at 6 months in patients with nonQ-wave myocardial infarction randomized to an invasive strategy (17). Symptoms of angina and re-admission were halved by the invasive strategy.
Outcome studies comparing more invasive versus more conservative strategies showed no relation between use of invasive cardiac procedures and subsequent mortality rates in patients with acute myocardial infarction (18-24). Rouleau and coworkers (18) reported that although coronary angiography was more commonly performed in the United States than in Canada (68% vs. 35%), as were revascularization procedures after infarction (31% vs. 12%), no difference in mortality (22% in Canada and 23% in the U.S.), or rate of reinfarction (14% in Canada and 13% in the U.S.) was observed at a mean follow-up of 42 months. Angina, however, was more prevalent in Canada (33% vs. 27%).
Krumholz et al (25) found no significant difference in short- or long-term mortality rates or hospital costs in patients admitted to hospitals with on-site cardiac catheterization facilities compared to those admitted to hospitals without such facilities. The cardiac catheterization rate was higher in the hospitals with facilities (38.6% vs. 26.9%; p<0.001), but the revascularization rate was similar (20.5% vs. 19.5%) during the initial episode of care and at 3 years (29.7% vs. 29.7%).
In the GUSTO-II trial, the incidence of coronary angiography after acute myocardial infarction was 81% in U.S. centers compared to 26% in Canadian centers (26). The incidence of revascularization after myocardial infarction in U.S. centers was 59%, compared to 15% in Canadian centers. Despite these large differences in postinfarction angiography and revascularization, mortality rates were comparable at 5.6% for U.S. centers and 5.7% for Canadian centers.
Taken together, most of the randomized studies and virtually all of the epidemiologic studies examining variation in practice patterns have shown no significant advantage of an early invasive strategy compared to a more early conservative strategy in patients undergoing selective angiography for spontaneous ischemia, clinical high-risk variables (e.g. prior infarction, congestive heart failure), or inducible myocardial ischemia. FRISC-II was the only randomized study showing a statistically significant advantage for an invasive strategy (17). However, that study employed only exercise electrocardiographic (ECG) treadmill testing for identification of patients with inducible ischemia following an uncomplicated nonQ-wave infarction.
Noninvasive Risk Assessment
Slide 3 lists the various noninvasive tests presently available for risk assessment after acute myocardial infarction. Studies in the literature published in the 1980s indicated that submaximal or symptom-limited exercise testing or pharmacologic stress testing with vasodilator stress, performed in conjunction with 201Tl perfusion imaging, could separate high- and low-risk subsets of patients who experienced an uncomplicated myocardial infarction (27). Gibson and colleagues (28) reported that approximately 50% of patients 65 years of age or younger with an uncomplicated myocardial infarction who showed either multiple 201Tl perfusion defects in more than one coronary vascular supply region, a reversible thallium defect within or outside the infarct zone or abnormal lung thallium uptake on predischarge submaximal exercise perfusion imaging subsequently experienced either cardiac death, recurrent infarction or hospitalization for Class III-IV angina. The cardiac event rate was only 6% for patients who had either a normal stress perfusion scan after infarction or solely persistent defects in the supply region of the infarct-related artery.
Other studies published in the prethrombolytic and pre-interventional era also reported a high sensitivity in perfusion imaging for detecting multivessel CAD after infarction. In a pooled analysis of 508 patients, there was a 72% sensitivity for multiple thallium defects in more than one coronary vascular supply region for identifying multivessel disease, with an 86% specificity. In these studies, the sensitivity of exercise ECG testing alone for detecting multivessel disease was only 59% (29).
Gibson and Beller (30) performed a pooled analysis of 17 studies in the literature comprising 1,426 patients who underwent both exercise ECG testing and exercise 201Tl perfusion imaging after uncomplicated myocardial infarction. The prevalence of ischemic ST-segment depression was 24% in that patient population, compared to 55% who manifested 201Tl redistribution (Slide #4). Thus, the sensitivity of myocardial perfusion imaging is significantly greater than the sensitivity of exercise ECG testing for identifying residual ischemia after uncomplicated infarction.
Shaw et al (27) performed a pooled analysis of test results from studies reported in the literature from 1980 to 1995. The death or myocardial infarction rate was 15.7% in patients with exercise-induced ST-segment depression on predischarge risk stratification using exercise electrocardiography, compared to a 9.9% subsequent combined event rate in patients without ST-segment depression (Slide #5). In contrast, the cardiac death or myocardial infarction rate was 16.7% in patients exhibiting multiple perfusion defects in more than one coronary supply region, compared to a 2% combined event rate in patients with exercise perfusion scans showing defects confined to one vascular region (Slide #6). Thus, this pooled analysis of data published in the literature also shows the superiority of exercise perfusion imaging in identifying patients at high risk for subsequent cardiac events.
The value of stress perfusion imaging can still be demonstrated in the thrombolytic era. DeCock and coworkers found that reversible defects on 201Tl scintigraphy were superior than exercise ECG variables for identifying high-risk patients after a Q-wave myocardial infarction (31). Basu and colleagues (32), employing exercise 201Tl imaging after thrombolytic therapy, reported that reversible ischemia on stress scintigrams predicted death, reinfarction, or subsequent unstable angina in 33 of 37 patients. The predictive value of the perfusion imaging procedure was superior to that of exercise ECG test variables alone.
Dakik et al (33) reported that quantitative exercise single-photon emission computed tomographic (SPECT) 201Tl perfusion imaging performed in postinfarction patients who had received thrombolytic therapy provided significant incremental prognostic information to clinical data. In that study, coronary angiographic variables did not further improve the prognostic model, which comprised clinical, resting ejection fraction, and 201Tl SPECT variables (Slide #7). Basu et al (34) used "nitrate-enhanced" 201Tl perfusion imaging in conjunction with exercise testing and found that 68% of post-myocardial infarction patients who received thrombolysis had evidence of viable but jeopardized myocardium. Of these, 49% had subsequent death or reinfarction. Only 13% of patients with no evidence of viable myocardium at jeopardy experienced subsequent cardiac events (p<0.001).
More recently, 99mTc-labeled perfusion agents have emerged as alternatives to 201Tl for both diagnostic and prognostic assessment of patients with coronary artery disease. With respect to risk stratification after acute myocardial infarction, Travin and coworkers (35) found that the number of ischemic defects on 99mTc-sestamibi SPECT was the only significant correlate of a future event when clinical, stress test, and imaging variables were assessed by Cox regression analysis. Patients with ³ 3 reversible 99mTc-sestamibi defects had an event rate of 38% (Slide #8).
Vasodilator Stress Imaging
Pharmacologic stress imaging is a useful alternative to exercise stress for risk stratification after acute myocardial infarction. Brown et al (36) performed dipyridamole 201Tl scintigraphy at a mean of 62 hours after hospitalization for acute myocardial infarction. By stepwise multivariate logistic regression analysis, the best predictor of subsequent cardiac events was the presence of dipyridamole-induced 201Tl redistribution within the infarct zone. Bosch et al (37) found that the presence of thallium redistribution on dipyridamole scintigraphy was the only independent predictor of subsequent events when imaging was performed 3 days after admission. Mahmarian et al (38) reported on the utility of adenosine 201Tl SPECT perfusion imaging performed in uncomplicated postinfarction patients. Univariate predictors of cardiac events in that study were quantified perfusion defect size, absolute extent of left ventricular ischemia, and the resting ejection fraction. Mahmarian et al (39) also reported that with adenosine 201Tl SPECT imaging, 82% of all noninfarct coronary stenosis of ³ 70% narrowing could be identified. In that study, 42% of noninfarct stenoses of 51-69% were detected (Slide #9). All infarct-related vessel stenoses were identified on scintigraphy.
Dipyridamole 99mTc-sestamibi myocardial perfusion imaging performed in 451 patients with their first acute infarction predicted early and late cardiac events (40). The extent and severity of the stress defect and reversibility of the defect were the most important predictors of cardiac death and recurrent infarction. Early dipyridamole imaging has been found to be safe without events attributed to the dipyridamole stress (41).
Based on data described above, as well as other reports in the literature, a decision-making algorithm is proposed that could be applied to the prognostic evaluation of uncomplicated postinfarction patients. Certainly, patients who manifest symptoms of congestive heart failure or hemodynamic compromise are candidates for early invasive evaluation. Similarly, patients with postinfarction angina are at high risk for recurrent ischemic events and should be evaluated early employing invasive strategies. However, a significant number of patients could be adequately risk stratified by an early noninvasive strategy using exercise or pharmacologic gated SPECT imaging. This permits the simultaneous assessment of resting ejection fraction, infarct size, regional systolic thickening abnormalities, and myocardial perfusion defects at rest and during stress. Vasodilator imaging can be performed as early as the third hospital day, whereas exercise perfusion imaging will often not be performed until the fourth or fifth hospital day. High-risk patients demonstrating residual ischemia or a multivessel disease scan pattern would then be referred for selective angiography where an invasive strategy aimed at revascularization would be undertaken. Patients with low-risk noninvasive stress perfusion imaging findings (e.g. lack of inducible ischemia, lack of a multivessel disease scan pattern, and lack of evidence for hibernating myocardium) would be eligible for aggressive medical therapy with subsequent follow-up imaging performed as indicated.
References1. Verani MS. Risk assessment after myocardial infarction: have the rules changed with thrombolytic therapy? J Nucl Cardiol 1996;3:S50-S59.
This company contributed to the Congress: