Relation Between Perfusion Defects on Stress Technetium-99m Sestamibi SPECT Scintigraphy and the Location of a Subsequent Acute Myocardial Infarction Gregory L. Miller, William MD, Steven D. Herman, MD, Gary V. Heller, MD, PhD, Sunil Kalla, A. Levin, MD, Kira M. Stillwell, MS, and Mark I. Travin, MD Although the presence of perfusion defects on stress myocardial perfusion imaging has been shown to correlate with future cardiac events, including acute myocardial infarction (AM), it is unknown whether the location of the AMI can be predicted. Therefore, for 25 patients who had an AMI following a stress technetium99m sestamibi single-photon emission computed tomographic (SPECT) imaging study and whose infarct location could be determined, the territory of infarction was correlated with the location of previous myocardial perfusion defects. A SPECT perfusion defect had been present in 24 patients (96%). The AMI occurred in territories that showed a reversible defect in 14 patients (56%), whereas 3 infarctions (12%) were in territories that revealed a fixed defect, and 8 infarctions (32%) were in territories that had not shown a defect on prior SPECT imaging. Whereas the incidence of infarction in territo- MD, ries with a reversible defect was highest at 14 of 26 (54%), the incidence of infarction in territories with a fixed defect was 3 of 7 (43%), and in territories with no defect was 8 of 42 (19%) (p = 0.011). Neither the time interval between SPECT imaging and infarction, nor the perfusion defect severity, was related to the correlation between perfusion defect and infarct location. Thus, although AMI occurs most often at the site of previous perfusion defects, reversible or fixed, a substantial percentage occur in territories without a perfusion defect. These findings suggest that abnormalities on SPECT perfusion imaging, although they serve as markers of significant coronary disease and increase the likelihood of infarction, do not always predict the exact location of infarction. (Am J Cardiol 1996;78:26-30) current perfusion tracer imaging techniques that reflect the severity of coronary narrowing may, like cardiac catheterization, be limited in their abilities to most severely stenotic artery is often not the site of predict the site of a future acute occlusion. This study a subsequent acute myocardial infarction ( AMI). 1.2 assesses how well the location of a future AM1 can One explanation for this apparent discrepancy may be predicted by the location and severity of perfusion be that standard 2-dimensional angiographic mea- defects, whether reversible or fixed, on a prior stress surements of anatomic severity over- or underestitechnetium-99m (Tc-99m) sestamibi single-photon mate the physiologic significance of a lesion.3 As emission computed tomographic (SPECT) myocarradionuclide myocardial perfusion scintigraphy pro- dial perfusion imaging study. vides a more functional physiologic assessment of a coronary stenosis, it may be able to predict the lo- METHODS cation of a future infarction better. Nevertheless, beThirty-two patients who had AMI, fatal or noncause an acute coronary syndrome is thought to be fatal, at any time following a stress (exercise or pharinitiated by sudden rupture of an underlying ather- macologic) Tc-99m sestamibi SPECT imaging study omatous plaque that results in thrombotic occlusion, were identified. The patients were derived from a infarction frequently occurs at sites of minimally ap- large, multicenter database of approximately 1,200 parent atherosclerotic disease.4 The site of plaque patients who underwent stress myocardial perfusion rupture appears to be more closely related to the imaging in 2 university-affiliated community teachphysical composition of the underlying atheroma ing hospitals over a 3-year period. Patients who had than to the magnitude of luminal narrowing.s-7 Thus, either coronary artery bypass surgery or coronary angioplasty/atherectomy in the period between stress From the Division of Cardiology, Roger Williams Medical Center, testing and AM1 were excluded from analysis. ata from studies using coronary angiography D have shown that in a patient with coronary artery disease, the myocardial territory supplied by the Providence; the Nuclear Cardiolo y Laboratory, Division of Cardiology, Memorial Hospital of Rho c?e Island, Pawtucket, and Brown University School of Medicine, Providence, Rhode Island, the Nuclear Cardiology Laboratory, Division of Cardiolo y, Hartford Hospital, Hartford, and the University of Connecticut SC a 001 of Medicine, Farmington, Connecticut. Manuscript received October 23, 1995; revised manuscript received and accepted January 9, 1996. Address for reprints: Mark I. Travin, MD, Division of Cardiology, Roger Williams Medical Center, 825 Chalkstone Avenue, Providence. Rhode Island 02908. 26 01996 by Excerpta All rights reserved. Medico, Inc. Diagnosis and localization of myocardial infarction: The diagnosis of AM1 required a characteristic increase and decrease of cardiac enzymes (creatine kinase with 2.5% myocardial band [MB ] fraction) * in the setting of prolonged ischemic chest discomfort and/or characteristic acute electrocardiographic changes. For 24 patients the location of infarction could be determined from standard electrocardio0002.9 149/96/$15.00 PII SOOO2-9149(96)00221-4 graphic criteria by consensus of 3 reviewers blinded to other data, and was classified as anterior (acute changes in leads V1 to V,), inferior (acute changes in leads II III, aVF, with or without acute changes in contiguous lateral and posterior leads), and isolated lateral (changes limited to any or all of leads I, aVL, V5, and V,) .9 In 1 additional case, a patient with chronic left bundle branch block developed new anteroseptal hypokinesia on echocardiography. Thus, 25 patients were available for further analysis. Stress testing: All patients in the study underwent stress testing with SPECT imaging using either treadmill exercise ( 11 patients) or pharmacologic provocation (intravenous dipyridamole in 13 patients, dobutamine in 1 patient ) . Symptom-limited exercise testing was performed using a standard or modified Bruce protocol. Dipyridamole was administered intravenously at a rate of 0.56 mg/kg body weight (maximum 60 mg) over 4 minutes. Intravenous aminophylline was administered at the supervising physician’s discretion for symptoms or routinely in the recovery period (22 minutes after radiopharmaceutical injection). For patients who were able, treadmill exercise was also performed immediately after the dipyridamole infusion. Dobutamine was administered intravenously by an IVAC Volumetric Controller infusion pump beginning with 5 pg/kg/min, and increased as tolerated in 3-minute stages to a maximum of 40 pg/kg/min. For all modes of stress, the patients’ symptoms, heart rate, blood pressure, and 12 lead electrocardiograms were monitored closely throughout stress and was considered recovery. The electrocardiogram positive for ischemia if there was ~-0.1 mV of horizontal or downsloping ST-segment depression that persisted for 280 ms after the J point. The electrocardiograms of patients with left bundle branch block or baseline ST-segment abnormalities, or those who were receiving digitalis and had ST depression during stress, were deemed inconclusive. edge of patient identity. All perfusion defects were assigned as being anterior (including anterolateral and anteroseptal) , inferior (including inferolateral and inferoseptal) , or midlateral, and each defect was classified as normal, reversible, or fixed. The defects were graded for severity of photon reduction, scored as mild ( 1) , moderate (2), or severe (3 ) . The locations and severities of the perfusion defects were correlated with the locations of subsequent AMIs. Statistics: Continuous variables were expressed as means + 1 SD, and compared by unpaired 2-sample t tests or analysis of variance. Categorical data ‘were expressed as proportions and were compared by the chi-square statistic, or by the Fisher exact test.” A p value <0.05 was considered significant. RESULTS Most of the patients in the study were men (20 of 25; SO%), and had a mean age of 67.2 & 2.6 years. A history of prior myocardial infarction was noted in 13 patients (56%), whereas 5 (20%) had undergone a revascularization procedure before stress imaging. Few patients had chest pain (7 of 25; 28%) or electrocardiographic ischemia (4 of 25; 16% ) during stress, but 96% (all except 1) had abno-rmal SPECT images, most with reversible defects (20 of 25; 80%). There was a wide range in both the time interval between stress testing and myocardial infarction (mean 231.2 + 175.9 days; range 12 to 676), and in the peak creatine kinase levels (mean 125 1.6 t 1406 ILJ; range 90 to 6064). Most infarctions ( 16 of 25; 64%) were classified as Q wave, 8 (32%) were non-Q wave, and 1 was left bundle branch block. The relation between the infarct territory and SPECT defect locations for the 25 patients is illustrated in Table I. Of 12 patients with anterior infarctions, 6 (50% ) had reversible defects in the anterior wall, whereas 6 had no anterior defects. Of 10 patients with inferior infarctions, 6 (60%) had reversSingle-photon emission computed tomographic imible inferior defects, 3 (30%) had fixed inferior deaging: All patients received an injection of 15 to 22 fects, and 1 had no inferior defect. Finally, of 3 mCi/70 kg of Tc-99m sestamibi at peak stress. One patients with lateral infarctions, 2 (67%) had revershour after injection, tomographic imaging was per- ible defects in the lateral wall, and 1 (34%) had no formed with an ADAC ARC 4000 gamma camera defect in the lateral wall. Overall, 14 patients (56%) (ADAC Laboratories, Milpitas, California) using a had a myocardial infarction in a territory that sho’wed high-resolution collimator as described previously.‘o a reversible defect on prior SPECT imaging, whereas All patients also underwent rest imaging, either the 3 patients ( 12%) had an infarction in a territory that same day immediately before stress (9 patients), re- revealed a fixed defect, and 8 patients (32%) had an ceiving 7 to 8 mCi/70 kg of Tc-99m sestamibi, or infarction in a territory that showed normal perfusion on a separate day (23 patients), receiving a dosage (Figure 1) . similar to stress. Transverse image slices 6 mm thick were reconstructed via filtered back projection using a ButterTABLE I Relation of Acute Myocardial Infarct Location to Previous SPECT Results in the infarct Territory worth filter. For all rest images and for the 2-day stress images, a cutoff frequency of 0.5 cycles/s, orReversible AMI Location n (% of AMls) der 10 was used, whereas for the stress images of the 6 Anterior 12 148%) l-day protocol a cutoff frequency of 0.66 cycles/s, 10 (40%) 6 Inferior order 7, was used. 3 (12%) 2 Lateral Single-photon interpretation: emission computed tomographic image All SPECT images were interpreted by consensus of 3 experienced readers without knowlCORONARY ARTERY AMI = acute myocordiol infarction; SPECT = single tomographic. DISEASE/MYOCARDIAL INFARCTION AFTER SPECT IMAGING 27 Fixed Defect No Defect (32%) TABLE II Infarct Comparison ?r-Y(‘*%~ Reversible [n = 14) Time interval between SPECT and AMI kh4 Prior AMI Prior CABG/PTCA 150 8 (57%) 722 + 617 158 emission computed tomographic infarct territories. (n = 25 patients be determined). 1 (33%) 1 (33%) 3 (100%) 2124 + 1877 283 + 232 5 (63%) 3 (38%) 5 (63%) 1961 + 1983 ^ the occurrence of infarction in territories with a hxed defect was 43%, and in territories with no defect it was nearly 20% (p = 0.011). The perfusion defects were scored visually for severity and classified as mild, moderate, or severe. As illustrated in Figure 3, defect severity, whether for a fixed or reversible defect, was not related to the incidence of a future infarction in that territory. DISCUSSION Although defects on stress myocardial perfusion imaging, particularly reversible defects, predict increased risk of a future AMI, the ability to identify the specific location of infarction is unclear.Lo,L2-14In the current investigation of patients who had an AM1 after a stress sestamibi SPECT imaging study, 68% of AMIs were in territories that had a perfusion defect. Although the occurrence of AMI was highest (54%) in territories with a reversible defect, there was a 20% occurrence in territories with normal perfusion. Studies with cardiac catheterization: Although there is a correlation between the degree of luminal narrowing and the frequency of future total occlusion, 15-17AM1 often occurs at the site of a less severe stenosis. In a study of 42 patients with AMI, Little et al I found that 66% of culprit infarct arteries had a <50% stenosis on a prior angiogram, and in only 34% of cases did the AM1 occur at the most severe stenosis site. Ambrose et al l8 reported that p=o.o11 only 22% of infarct-related arteries demonstrated >70% diameter stenoses before infarction. The high incidence of mild to moderate stenoses leading to AM1 has been supported further by studies with quantitative angiography showing that the degree of residual luminal narrowNo Defect Fixed Defect Reversible Defect ing after thrombolytic therapy is n=42 n=7 n=26 often <60%.29’9-2’ The ability of nuclear imaging SPECT Image Results 100% I 60% 6O% 40% 20% to predict FIGURE 2. Occurrence of myocardial infarction (Ml) in relation to prior single-photon sion computed tomographic (SPECT) image results. (n = 75 territories in 25 patients). THE Defect (n = 8) ._ The characteristics of AMIs that occurred in territories that had a reversible defect on prior SPECT imaging were compared with those that had fixed defects and with those that had no defects (Table II).The time intervals between stress testing and AM1 were similar among the 3 groups. More than one half (63%) of the patients with AM1 in a territory that had no defect on SPECT imaging had a prior infarction in another territory. All 3 infarctions that occurred in territories with a fixed defect were Q-wave inferior infarctions (only 1 of which, by history, was a prior Q-wave infarction in the inferior wall), whereas more than one-half of infarctions in the reversible and no defect territories were of the Q-wave type. There was a trend toward decreased peak creatine kinase levels for infarcts in territories of previous reversible defects. The occurrence of AM1 as a function of the SPECT image results is depicted in Figure 2. There were 75 total myocardial territories (3 per patient: anterior, inferior, and lateral) available for analysis. Of the 75 territories, 42 had no defect, 7 had a fixed defect, and 26 had an ischemic defect. In these 25 patients, the occurrence of infarction was highest (54% ) in territories with a reversible defect, whereas 28 -t 128 No AMI = acute myocordial infarction; CABG = coronoryartety bypass grafting; CK = creatine kinose; PTCA = percutaneous tronsluminol coronary ongioplasty; SPECT = single photon emission computed tomographic. (56%) Incidence of MI in the Territory wd + 7 (50%) 1 (7%) Q wave Peak CK (IU) FIGURE 1. Prior single-photon image results in subsequent whose infarct location could 217 SPECT Image Result in Infarct Territory Fixed (n = 3) AMERICAN JOURNAL OF CARDIOLOGY@ VOL. 78 JULY 1, 1996 emis- myocardial infarction: Limitations in the ability of coronary angiography to predict likelihood of rupture and t’otal occlusion at 1 of these sites.17 IOMI q NO MI 1 p=NS An important implication of p=NS our study is that performing a revascularization procedure in response to a defect on myocardial perfusion imaging, whether 2 fixed or reversible, may not Mean necessarily be effective in preDefect ,.5 venting a future AMI. Studies Severity have shown that neither bypass Score surgery nor angioplasty reduce 1 the incidence of myocardial infarction.28*29 Study limitations: Current sltudies examining the prognostic implications of myocardial perFixed Defect Reversible Defect fusion imaging are limited by a significant proportion of paFIGURE 3. Occurrence of myocardial infarction (MI) in relation to prior single-photon emisscore. For the 7 territories with a fixed defect, sion computed tomographic defect severi tients being referred for rethe 3 that were the sites of subsequent in % rction had a mean severity score of 2.3 2 0.6, vascularization procedures in whereas the 4 that did not have a subsequent infarction had a mean severity score of 2.5 response to nuclear image rei 0.6, p = NS. For the 26 territories with a reversible defect, the 14 that were the site of a sults.30 Approximately 10% of subsequent infarction had a mean severity score of 2.2 IL 0.7, whereas the 12 that did not the patients in our larger datahave a subsequent infarction had a mean severity score of 2.3 t 0.5. base underwent revascularization. Because patients with the most severely abinfarct site may reflect the notion that the anatomic degree of luminal narrowing seen or measured on a normal images or unstable symptoms are those most patients left for coronary angiogram may not accurately reflect its often sent for revascularization, longer follow-up are a skewed, lower risk group, true physiologic severity.3,22 Kaul et a123 reported thallium perfusion defects in one fifth of the seg- and thus at lower risk for having a future AMI. ‘The ments supplied by vessels with a <50% stenosis. presence of a severe reversible defect in a patient The functional significance of a coronary artery ste- with unstable angina may in fact indicate a high nosis, best assessed with stress testing, has been likelihood of AM1 in that territory, but such a pashown to be of more prognostic importance than the tient would most likely have been referred for a revascularization procedure first, and not included visualized anatomic disease.2” Although the number in this study. of coronary vessels with a ~50% luminal diameter The study was also limited by frequent difficulty narrowing is a strong predictor of a future cardiac event (late revascularization, death, or AMI), thal- in identifying the actual site of AMI, limiting the lium-201 imaging variables have been shown to be number of patients available for analysis. In addition, better at predicting a nonfatal AMI.2j The ability of from the SPECT images it was sometimes difficult stress myocardial perfusion imaging to predict a fu- to assign a defect to a particular vascular territory, ture AM1 is well documented in other studies, 13.26*27 particularly in the region between the circumflex and but it has been unclear whether the physiologic in- right coronary arteries. Finally, a variety of stress formation provided by this technique enhances the protocols were used, although the presence or absence of concordance between perfusion defect and ability to predict the specific location. Clinical implications: Although this study did dem- AM1 location was unrelated to either the stress proonstrate a correlation between the location of a tocol or the workload achieved. SPECT defect and the location of a subsequent AMI, many AMIs (32%) occurred in territories without Acknowledgment: We gratefully acknowledge the defects. Because all but 1 of these 8 patients had a assistance of David Waters, MD, in reviewing and SPECT perfusion defect in another territory, the im- providing suggestions for this manuscript. age findings served as markers of disease, predicting an increased risk of an infarct-producing plaque rupture somewhere in the coronary arterial circulation, 1. Little WC, Constantinescu M. Applegate RJ, Kutcher MA, Burrows MT, which in these cases did not occur at the most severe Kabl FR, Santamore WF’. Can coronary arteriography predict the site of .3 rubcoronary flow impairment site. This finding is com- sequent myocardial infarction in patients with mild-to-moderate coronary artery disease? Circulation 1988;78: 1157- 1166. patible with the understanding that plaque rupture 2. Hackett D, Davies G, Masai 4. Pre-existing coronary stenoses in patients that results in an acute coronary syndrome may be with first myocardial infarction are not necessarily sevens. Eur Ham J related more to plaque composition than to the de- 1988:9:1317-1323. gree of luminal narrowing.4m7 Because there are sta- 3. White CW, Wright CB, Doty DB, Hiratza LF, Eastham CL, Harrison DG. Marcus ML. Does visual interpretation of the coronaq xtetiogram predict the tistically more lesions of lesser narrowing in a dis- physiologic importance of a coronxy stenosis? N Engl J Med 1984:310 819eased coronary arterial system, there may be a higher 823. CORONARY ARTERY DISEASEi’MYGC4RDIAL INFARCTION AFTER SPECT IMAGING 29 4. Fuster V, Badimon L, Badimon .I, Chesebro JH. The pathogenesis of coronary artery disease and the acute coronary syndromes. New Eng2 J illed 1992;326:242-259.310-318. 5. Femtidez-Grtiz A, Badimon JJ, Falk E, Fuster V, Meyer B, Mailhac A, Weng D, Shah PK. Badimon L. Characterization of the relative tbrombogenicity of atherosclerotic plaque components: implications for consequences of plaque rupture. J Am Coil Cardiol 1994;23: 1562- 1569. 6. Cianflone D, Ciccirillo F. Buffon A. Trani C, Scabbia EV, Finocchiaro ML, Crea F. Comparison of coronary angiographic narrowing in stable angina pectoris, unstable angina pectoris, and in acute myocardial infarction. Am J Car&l 1995:76:215-219. 7. Uchida Y, Nakamura F, Tomam T, Morita T, Oshima T, Sasaki T. Morizuki S, Hirose J. Prediction of acute coronary syndromes by percutaneous coronary angioscopy in patients with stable angina. Am Heart .I 1995;130: 195-203. 8. Lee TH, Goldman L. Serum enzyme assays in the diagnosis of acute myocardial infarction. Ann htem Med 1986;105:221-233. 9. Van Reet RE, Quinones MA, Poliner LR. Nelson JG, Waggoner AD. Kanon D, Lube&in SJ. Pratt CM, Winters WL. Comparison of two-dimensional echocardiography with gated radionuclide ventriculography in the evaluation of global and regional left ventricular function in acute myocardial infarction. J Am Cdl Cardiol 1984;3:243-252. 10. Heller GV. Herman SD, Travin MI: Baron JI. Santos-Ocampo C, McClellan JR. Independent prognostic value of inuavenous dipyridamole with technetium991x1 sestamibi tomographic imaging in predicting cardiac events and cardiac hospitalizations. J Am Co[/ Car&l, 1995:26:12021208. 11. Glantz SA. Primer of Biostatistics. 2nd ed. New York: McGraw-Hill Information Services Company, 1987:125-126. 12. Brown KA. Prognostic value of thallium-201 myocardial perfusion imaging. Cixrrlation 1991;83:363-381. 13. Stratmann HG, Williams GA, Witxry MD, Chaitman BR, Miller DD. Exercise technetium-99m sestamibi tomography for cardiac risk stratification of patients with stable chest pain. Circulation 1994;89:615-622. 14. Travin MI, Dessouki A, Cameron T, H&r GV. Use of exercise technetium99m sestamibi SPECT imaging to detect residual ischemia and for risk stratification following acute myocardial infarction. Am JCardiol 1995;75:665-669. 15. Moise A, Lesperance J, Theroux P, Taeymans Y, Goulet C, Bourassa MG. Clinical and angiographic predictors of new total coronary occlusion in coronary artery disease: analysis of 313 nonoperated patients. Am .I Cardiol 1984;51:1176-1181. 16. Ellis S, Alderman E. Cain K. Fisher L, Sanders W, Bourassa M, and the CASS Investigators. Prediction of risk of anterior myocardial infarction by lesion severity and measurement method of stenoses in the left anterior descending coronary distribution: a CASS registry study. J Am Co11 Cardiol 1988;11:908-916. 17. Taeymans Y, ThCroux P. Lesp&ance J, Waters D. Quantitative angiographic morphology of the coronary artery lesions at risk of thrombotic occlusion. Circulatim 1992;85:78-85. 18. Ambrose JA, Tannenbaum MA, Alexopoulos D, Hjemdah-Monsen CE, Leavy J, Weiss M, Bonito S, Gorlin R, Fuster V. Angiographic progression of coronary artery disease and the development of myocardial infarction. J Am Coil Cardiol 1988;12:56-62. 19. Sermys PW. Arnold AER, Brower RW, deBono DP, Bokslag M, Lubsen J, Reiber JH, Rutsch WR, Uebis R, Vahanian A. Effects of continued rt-PA administration on the residual stenosis after successful recanalization in acute myocardial infarction-a quantitative coronary angiography study of a randomized hial. Eur Heart J 1987;8: 1172- 1181. 20. Brown BG, Gallery CA, Badger RS, Kennedy JW. Matbey D, Bolson EL, Dodge HT. Incomplete lysis of thrombus in the moderate underlying atherosclerotic lesion during intracoronary infusion of streptokinase for acute myocardial infarction: Quantitative angiographic observations. Circulation 1986;73:653-661. 21. Sermys PW, Wijns W, van den Brand M, Rib&o V, Fioretti P. Simoons ML, Kooijman CJ, Reiber JH, Hugenholtz PG. Is transluminal coronary angioplasty mandatory after successful tbrombolysis? Quantitative coronary angiographic study. Br Heart J 1983:50:257-265. 22. Kiat H, Ansari A, Berman DS. Friedman J, Siegel R. Abnormal myocardial perfusion single photon emission computed tomography and normal coronary arteriogmm: the role of intracoronay vascular ultrasound in providing diagnostic confirmation. Am Hem-t J 1995: 130: 182- 186. 23. Kaul S, Chesler DA, Okada RD. Boucher CA. Computer versus visual analysis of exercise thallium-201 images: a critical appraisal in 325 patients with chest pain. Am Heart J 1987;1129-1137. 24. Weiner DA. Ryan TJ, McCabe CH, Chairman BR. Sheffield LT, Fisher LD, Tristani F. Value of exercise testing in determining the risk classification and the response to coronary artery bypass grafting in three-vessel coronary artery disease: a report from the Coronary Artery Surgery Study (CA%) registry. Am J Cardiol 1987;60:262-266. 25. Kaul S, Lilly DR, Gascho JA, Watson DD, Gibson RS, Oliner CA, Ryan JM, Beller GA. Prognostic utility of the exercise thallium-201 test in ambulatory patients with chest pain: comparison with cardiac catheterization. Circulatim 1988;77:745-758. 26. Brown KA, Boucher CA, Okada RD. Guiney TE. Newell JB. Strauss HW, Pohost GM. Prognostic value of exercise thallium-201 imaging in patients presenting for evaluation of chat pain. J Am Co11 Cmdiol 1983: 1:994- 1001. 27. Travin MI, Boucher CA, Newell JB, LaRaia PJ, Flares AR, Eagle KA. Variables associated with a poor prognosis in patients with an ischemic thallium-201 exercise test. Am Heart J 1993;125:335-344. 28. Rutherford JD, Braunwald E, Cohn PF. Chronic ischemic heat disease. In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. Philadelphia: WB Saunders Company, 1988:1314-1379. 29. Parisi AF, Folland ED, Ha&an P, ACME investigators. A comparison of angioplasty with medical therapy in the treatment of single-vessel coronary artery disease. N Engl J Med 1992;326: IO- 16. 30. Brown KA. Noninvasive risk stratification: right question, wrong study. J Nucl Cardiol 1994:1:112-113. 30 JULY THE AMERICAN JOURNAL OF CARDIOLOGY” VOL. 78 1, 1996