Prodromal Angina Limits Infarct Size
A Role for Ischemic Preconditioning
Background In the experimental setting, it has been demonstrated that preconditioning myocardium before prolonged occlusion with brief ischemic episodes affords substantial protection to the cells by delaying lethal injury, thereby limiting infarct size. Whether the same occurs in humans remains unknown.
Methods and Results This study was undertaken to determine whether new-onset prodromal angina, defined as chest pain episodes limited to the 24 hours before myocardial infarction, is the clinical correlate of the ischemic preconditioning phenomenon. Twenty-five patients with their first anterior myocardial infarction treated with thrombolysis (recombinant tissue plasminogen activator [r-TPA], 100 mg/3 hours) were retrospectively included in the study because they met the following criteria: (1) <120 minutes from onset of symptoms to reperfusion therapy, (2) <90 minutes from the beginning of thrombolytic therapy to reperfusion (defined as rapid ST elevation reduction >50%), (3) a patent infarct-related coronary artery with TIMI 3 flow and complete absence of collateral circulation to the infarct related artery (assessed at 24±5 days), and (4) the presence of new-onset prodromal angina, ie, typical chest pain episodes occurring at rest within 24 hours or, alternatively, a complete absence of symptoms before onset of infarction. Therefore, on the basis of their clinical status before infarction, the patients were divided into two groups: group 1, 13 patients without prodromal angina, and group 2, 12 patients with prodromal angina. Despite no difference in time to treatment (81±19 versus 75±21 minutes in group 1 and group 2, respectively; P=NS) and time to reperfusion (58±34 versus 46±24 minutes; P=NS), the peak of CKMB release was markedly lower in group 2 (86.3±66 versus 192.3±108.3 IU/L; P<.01). In addition, although both groups were comparable in terms of area at risk (amount of myocardium beyond the infarct-related stenosis; 15.1±4.6 versus 13.7±4.6 hypokinetic segments in group 1 and group 2, respectively, P=NS), the final infarct size (11±7.5 versus 5.6±4 hypokinetic segments, P<.04) was smaller in group 2. Thus, the limitation of the infarct size was significantly greater in group 2 (69% versus 36%; P<.05), and this represents an additional 33% reduction (95% confidence intervals, 7.1% to 58.9%; P=.01) in the group of patients with prodromal angina. Also, the third index, that is, the ECG, showed a favorable trend toward a lesser number of Q waves and a higher Σ R waves, although the values did not reach statistical significance.
Conclusions Despite a similar area at risk, patients with new-onset prodromal angina showed a significantly smaller infarct size compared with patients without prodromal symptoms. Since the two groups had similar times to reperfusion and no evidence of collateral circulation to the infarct related artery, the protection afforded by angina in group 2 patients might be explained by the occurrence of ischemic preconditioning.
The recent discovery of the phenomenon of ischemic preconditioning has focused attention on the ability of the myocardium to protect itself. It has been well demonstrated in several animal models that brief ischemic episodes preceding prolonged coronary occlusion cause a significant reduction of infarct size.1 2 3 4 Preconditioning delays lethal cell injury in ischemic myocardium, although the mechanism is still unclear. Its protective role appears to be related to the length of coronary occlusion; a delay in cell death is detectable up to a maximum of 180 minutes of vessel occlusion, after which no further benefits are present.5
Such a phenomenon could have clinical implications in the setting of successful reperfusion after thrombolytic therapy for acute myocardial infarction. New-onset prodromal angina, defined as chest pain episodes limited to the 24 hours before infarction, could be regarded as a clinical correlate of ischemic preconditioning.
To verify this hypothesis and demonstrate a role for ischemic preconditioning in humans, we set up an arbitrary template (including prodromal angina and time to treatment plus time to reperfusion) that could approximate as closely as possible the experimental conditions. Thus, applying this arbitrary human model, we retrospectively reviewed all patients diagnosed as having their first anterior myocardial infarction treated with thrombolytic therapy who were admitted to our coronary care unit during a period of 18 months.
Selection of Patients
Between January 1991 and July 1992, 84 patients were consecutively admitted to our coronary care unit with their first suspected acute anterior myocardial infarction. They were retrospectively included in the study if they satisfied the following features: (1) typical chest pain <120 minutes from symptom onset to reperfusion therapy (approximating as much as possible the time limits delineated by Murry et al1 in their experimental model); (2) ST elevation of >2 mm in two contiguous precordial leads (V1,V6); (3) reperfusion therapy with recombinant tissue plasminogen activator (r-TPA); (4) <90 minutes from the beginning of thrombolysis to reperfusion (defined as rapid ST elevation reduction >50%); (5) Killip class 1 on admission; (6) a patent infarct-related coronary artery showing TIMI 3 flow and complete absence of collateral circulation to the infarct-related artery (assessed at 24±5 days); and (7) the presence of new-onset prodromal angina, that is, typical chest pain episodes at rest within the previous 24 hours or a complete absence of any symptoms before the onset of myocardial infarction. Symptom status before infarction was elicited by specific questioning by the treating physician at the time of admission, using a standardized history form, and was reviewed retrospectively by a single observer (D.P.). For the purpose of this study, prior angina was defined as recurrent ischemic chest pain beginning more than 24 hours before admission, regardless of whether or not any kind of prior medical therapy had been instituted. Fourteen patients showing this feature were excluded from the study. In addition, 30 patients were excluded because they showed >120-minute delay between symptom onset and treatment, 8 patients because reperfusion time was >90 minutes, and 7 patients because they refused to undergo cardiac catheterization. Therefore, 25 patients were finally selected to be included in the study. Clinical status, as previously defined, formed the basis for dividing the patients included into two groups: group 1, 13 patients who showed no chest pain in their life before the episode leading to admission, and group 2, 12 patients who showed new-onset prodromal angina (episodes of chest pain at rest within 24 hours before myocardial infarction). None of these patients were taking cardiovascular drugs before admission to the coronary care unit.
Reperfusion and Adjunctive Therapy
All patients were given r-TPA (Actilyse, Boehringer Ingelheim) at the standard regimen of 100 mg (10 mg as a bolus, 50 mg in the first hour, 40 mg during the subsequent 2 hours). At the end of thrombolytic infusion, all patients received a 5000-IU heparin bolus followed by a 1000 IU/h infusion in order to maintain the APTT at two times the basal value. Aspirin at the dosage of 165 mg was given orally on admission. β-Blockers and nitrates were left to the discretion of the treating physicians.
Study Protocol and Parameters Evaluated
A standard 12-lead ECG using a Hewlett-Packard electrocardiograph was recorded before the initiation of thrombolytic therapy, at the end of the r-TPA infusion, at 24 hours, and on day 7. In all patients the location of leads V1 through V6 were marked on each patient’s chest on admission to ensure consistent electrode placement in the recording of subsequent ECGs.
ECG Measurements.The following parameters were calculated as reported by Blanke et al6 : (1) ΣST V1 through V6 (the sum of ST elevation above the baseline defined by the preceding TP segment in leads V1 through V6), (2) n Q V1 through V6 (the number of pathological Q waves in the precordial leads, with any Q waves in leads V1 through V3, Q waves wider than 20 milliseconds in V4, and wider than 30 milliseconds in V5 and V6, considered pathological), and (3) ΣR V1 through V6 (the sum of R wave height in V1 through V6).
Continuous ST-Segment Monitoring.The precordial lead showing maximum ST-segment elevation in the qualifying ECG was selected for continuous monitoring (Space Lab, Inc), which was initiated concurrently with the commencement of reperfusion therapy and continued throughout thrombolysis. Reperfusion was believed to occur when ST-segment elevation abruptly decreased more than 50% relative to the most abnormal peak documented at any time.20 21 Twelve-lead ECGs were recorded at 10- to 15-minute intervals and when a change in chest pain, rhythm, or heart rate was noted. The magnitude of ST-segment elevation was measured 80 milliseconds after the J-point.
Blood samples for measuring CK and CKMB plasma levels were taken on admission, every 2 hours for the first 16 hours, and subsequently at 20, 24, 28, 32, 36, 48, 60, and 72 hours after the beginning of thrombolysis. For our laboratory, the upper limits of normal range were 200 and 16 IU/L for plasma CK and CKMB, respectively. A 72-hour time-activity curve then was plotted from the CKMB values in both groups of patients for detecting any difference in enzyme release over time. The time to CKMB peak was timed from the onset of myocardial infarction symptoms.
Cardiac Catheterization and Quantitative Left Ventriculography
Coronary angiography and left ventriculography were delayed until 3 to 4 weeks after infarction to minimize the effects of myocardial stunning on the determination of final infarct size. Left ventricular volumes were calculated from an angiogram in the 30° right anterior oblique projection according to the area-length method. The infarct-related artery was identified by correlating the coronary anatomy with the site of ST-segment elevation on the admission ECG and the distribution of impairment of contractility in the left ventriculogram. Patency of the infarct-related artery was classified according to the criteria of the TIMI trial.7 Stenoses occluding more than 50% of the arterial diameter were regarded as significant. In addition, after dividing patients into one-, two-, and three-vessel disease, we evaluated the two patient groups on the basis of the angiographic scoring system used at Green Lane Hospital, Auckland, New Zealand. This system, fully reported elsewhere,8 incorporates the location, length, and severity of the coronary stenosis as well as the amount of the myocardium supplied. Briefly, using the grade of stenosis and the myocardial value (the total number of units of myocardium supplied by that artery distal to the stenosis), each artery is given an angiographic score according to a prespecified grading system. Therefore, for each artery or group of arteries, it is possible to calculate a separate angiographic score, which, summated with the others, gives the total angiographic score for that patient.
Regional Wall Motion Analysis
Regional wall motion analysis was performed with the Kontron 200 system. A modification for determination of the area of myocardium at risk was also introduced in the procedure. The end-diastolic and end-systolic silhouettes of the left ventricle, obtained during sinus rhythm in the 30° right anterior oblique projection, were first traced onto paper and subsequently entered into a computer with an x-y digitizer (Fig 1⇓). Then, from the center of the left ventricle, 36 equally spaced radii were drawn to intersect the endocardial outlines for end-diastolic and end-systolic images, thus dividing images into 36 areas. In our laboratory, normal wall motion for each area was derived from ventriculograms of 48 patients with normal left ventricular function and no coronary artery or valvular disease. Ventriculograms were filmed at 50 frames per second, and a grid filmed at the level of the patient’s heart was used for calibration. No attempt was made to correct for rotational changes because such corrections increase rather than decrease variability.9
Area at Risk
The area of myocardium at risk was defined following the procedure proposed by Cross et al.10 Briefly, this area was defined as the number of segments depicted on the 30° right anterior oblique ventriculogram supplied by the infarct-related artery. The proximal limit was set at the level of the “culprit” lesion (the lesion associated with evidence of thrombus or the most proximal severe stenosis). The distal limit of the area at risk was defined as the termination of the infarct-related artery as viewed in the right anterior oblique projection. The proximal and distal limits were marked by the two readers (F.O. and M.G.) who performed the measurements blinded to the clinical conditions, on the diastolic contour, and the areas in between were considered at risk. Areas 1,2 and 35,36 were excluded from analysis because they reflected movements of the mitral valve and the aortic root. The interobserver and the intraobserver variability for the area at risk was 0.61±0.65 and 0.56±0.65 segments, respectively, while the interobserver and intraobserver variability for infarct size was 0.20±0.37 and 0.10±0.28 segments, respectively. This means <5% difference between repeated calculations for both sets of measurements.
Shortening of segmental areas by <2 SD of the normal range for systolic shortening in healthy subjects was considered to indicate an abnormal area. Therefore, all segments, which at ventriculography showed an inward motion depressed below the threshold specified above, were indicated as hypokinetic and considered to form the final infarct size.
Severity of Regional Abnormalities in Wall Motion
Because the measurement of hypokinetic segment length indicates only the percentage of the contour suffering decreased wall motion and provides no information concerning the magnitude of dysfunction in any specific region, we also evaluated the measurement of inward motion expressed as severity of hypokinesis. The difference between systolic shortening of each involved area and the lower limit of the normal range for systolic shortening in healthy subjects represented the severity of hypokinesis in that specific area. The index of the severity of hypokinesis was calculated as follows.
where HI denotes the hypokinesis index; AN, the percentage of systolic area shortening in healthy subjects minus 2 SD; AD, the percentage of systolic shortening of the corresponding dissynergic area; and nAD, the number of dissynergic areas.
Data are presented as mean values ±SD. Continuous variables were analyzed with an unpaired t test or Mann-Whitney test when appropriate. Two-sided P values are reported, and a value of P<.05 was considered significant. Differences in ECG measurements among groups at several time points were evaluated by ANOVA for repeated measures. Significance was computed by Scheffe’s F test; a P value <5% was considered significant.
Clinical and Angiographic Data
The evaluated variables characterizing the two groups are listed in the Table⇓. No significant differences were detected between the two sets of patients. Patients were quite well matched for age, sex, and risk factors. The time to commencement of thrombolytic drug infusion was similar (81±19 minutes in group 1 versus 75±21 minutes in group 2), and angiographic variables showed no differences among the two study groups. Particularly, the angiographic score was similar in the two groups. Neither the total value (5.8±2.7 versus 5.2±2.7, respectively; P=NS) nor the value referring only to the left anterior descending coronary artery territory (4.0±1.5 versus 3.6±1.6, respectively; P=NS) showed a statistically significant difference. There were 17 anginal attacks, all at rest, before infarction in the angina-positive patient group, which accounted for a mean 1.4±0.5 angina episodes per patient (range, 1 to 2 episodes). A single chest pain episode had a mean duration of 13±6 minutes (range, 5 to 25 minutes), while the total amount of chest pain duration per patient averaged 19±10 minutes (range, 10 to 45 minutes). The mean time interval between the last episode and the onset of infarction was 11.2±6.7 hours, ranging from 1 to 20 hours.
Continuous ST-Segment Monitoring and Reperfusion Time
By means of continuous monitoring of the lead showing maximum ST elevation, the reperfusion time (that is, the time at which ST-segment elevation dropped to at least 50% of the most abnormal peak documented at any time during the study) was 58±34 minutes (ranging from 15 to 95 minutes) for angina-negative patients and 46±24 minutes (ranging from 15 to 90 minutes) for angina-positive patients (P=NS). The initial mean value of ST elevation was similar (4.0±1.7 mm versus 3.0±0.9 mm; P=NS) between the patient groups. After reperfusion was believed to occur, the trend toward ST elevation resolution was continuous and progressive until−at the end of the thrombolytic infusion−a final value of 0.7±0.2 mm was reached by both groups. This accounted for a mean final ST elevation decrease of 77±6% in angina-negative patients and 82±9% in angina-positive patients.
The sum of ST elevation was similar (11.1±3.8 versus 17.4±11.3 mm; P=NS) in both groups on baseline ECG (Fig 2⇓). There were no statistically significant differences at any evaluated time among the two groups.
Number of Q Waves
The number of Q waves was similar at all times. However, a nonsignificant trend toward a higher number of Q waves was observed at each time interval in the angina-negative group (P=.1).
ΣR waves were higher at each ECG sampling point in angina-positive patients. This difference was statistically significant at baseline (24.9±13.5 versus 43.1±20.6 mm; P<.05), at the end of r-TPA infusion, and at 24 hours, but not at predischarge evaluation.
Enzymatic Infarct Size
In both groups, cardiac enzymes peaked early (9.3±2.6 hours and 8.6±4.7 hours after onset of pain in angina-positive and angina-negative patients, respectively; P=NS). The peak of CKMB release was markedly lower in angina-positive patients (86.3±66 versus 192.3±108.3 IU/L; P<.01).
Fig 3⇓ shows the time-activity curves in both groups. These curves separate with statistically significant differences (P<.01) at each sampling point between the 6th and the 28th hour.
Ventriculographic Infarct Size
Both groups were comparable in terms of myocardium at risk (15.1±4.6 versus 13.7±4.6 hypokinetic segments in group 1 and group 2, respectively, P=NS). However, infarct size was significantly lower in patients with prodromal angina (5.6±4 versus 11.0±7.5 segments, respectively; P<.04), which means a significantly greater percentage of infarct size limitation (69% versus 36%; P<.05 in group 2 patients). Thus, an additional 33% of myocardium at risk was salvaged (95% confidence intervals, 7.1% to 58.9%; P<.01) in patients with prodromal angina. Individual values of area at risk and the correlated final infarct size are illustrated in Fig 4⇓. The measurement of inward motion in the whole final infarct area, expressed as severity of hypokinesis, was also positively associated with prodromal symptoms, being the hypokinetic index less than in patients without prodromal angina (11±8.6% versus 19±10.1%; P<.05). There were no significant differences between the two groups for left ventricular volumes and ejection fraction.
The major finding of this study is that patients with anterior myocardial infarction treated with thrombolytic therapy within 2 hours of the onset of symptoms who have had prodromal angina in the previous 24 hours have smaller infarcts as assessed by cardiac enzymes and ventriculography than patients without prodromal angina.
Limitation of Infarct Size
Prodromal angina may therefore have a protective role in patients successfully reperfused with thrombolytic therapy. Patients were well matched at baseline, and two independent indexes, CKMB time-activity curves and left ventricular angiogram analysis, showed decreased infarct size in patients with prodromal angina. Patients without prodromal angina had more than twice as much CKMB release, though reperfusion occurred at the similar time, as assessed by time to ST-segment recovery. Left ventriculogram analysis showed a lesser number of hypokinetic segments in the patients with prodromal angina (5.6±4 versus 11.0±7.5; P<.04), and the degree of hypokinesis was less than in the patients without prodromal angina. Correction for area at risk, that is, the area subtended by the infarct related artery in each individual patient, partially controls for one source of variation between the two different groups. In the conscious dog model of myocardial infarction, 66% of the variation in infarct size was due to the variation in the anatomic area at risk.11 This method may be particularly useful in small studies, like this one, where absence of correction for variation in patterns of coronary artery supply would have limited the ability to detect important differences between groups.
The relation between the third index of infarct size used in the study, the ECG, and the protective role of new-onset prodromal angina was less clear. The ΣR wave voltages showed statistically significant differences between the two patients groups at each time interval except for the predischarge evaluation. There was also a trend toward less Q wave development, but this did not reach statistical significance. There may have been a type 2 error in evaluation of the ECG criteria, and greater precision may have been obtained by using more ECG leads.
Pathogenetic Mechanisms Underlying the Beneficial Role of Prodromal Angina
On the basis of our findings, the most reasonable mechanism to explain the beneficial role of prodromal angina is ischemic preconditioning. Also, Deutsch et al12 support the occurrence of this phenomenon in humans, showing that in patients undergoing elective percutaneous transluminal coronary angioplasty, angina, ST-segment changes, myocardial lactate production, and coronary vein flow were all attenuated during the second balloon inflation as compared with the first one.
However, other possible explanations such as collateral circulation should be taken into account. The presence of angiographic collaterals is very dynamic,13 and a transient role may be important14 in the acute phase of myocardial infarction. In the present study, early angiography was not performed; therefore the presence or absence of collateral circulation at that time was not able to be documented, and this might have constituted a potential limitation of the study. Rentrop et al15 suggested that well-developed collaterals may extend the “time window” for the beneficial effect of pharmacological reperfusion. Habib et al,16 in a detailed post hoc analysis from the TIMI Phase I study, reported a 35% reduction of enzymatically estimated infarct size and an 11% difference (53.4±1.8% versus 47.8±1.7%) in predischarge ejection fraction in patients with collaterals compared with patients without collaterals. However, this finding applied only to patients who failed to reperfuse at 90 minutes after intravenous thrombolytic therapy. In patients in whom the extent of necrosis is limited by reperfusion, the impact of collaterals may be less important. Our patients all had a marked decrease in ST elevation and rapid normalization of the ECG during the acute phase, suggesting early reperfusion,17 and they also had a patent infarct-related artery with TIMI 3 flow documented at late angiography. Additionally, collaterals of any degree were an exclusion criterion for the study.
Limitations of the Study
This is a retrospective evaluation. However, the criteria for the inclusion of patients, selected to closely approximate the experimental conditions, were defined before starting to review the patient records. The protection afforded by preconditioning declines progressively as the reperfusion time between the last ischemic episode and prolonged coronary occlusion becomes longer, as demonstrated by Murry et al5 in the experimental setting. After 120 minutes, in fact, there was considerable attenuation of infarct size reduction, falling from 92% to 54% in their canine model. We defined prodromal angina as episodes of new-onset angina in the 24 hours before myocardial infarction. The reperfusion time from the last episode of angina and the index infarction was 11.2 hours (range, 1 to 20 hours), which, on the basis of the available animal experience, could be regarded as too long to confer an appreciable protection to jeopardized myocardium. Whether this applies in the same way to humans is not known. Furthermore, in humans, silent ischemia also accounts for the major part of the total ischemic burden.18 19 Obviously, silent myocardial ischemia could have occurred in the group with as well as in the group without prodromal angina. However, in a group of 52 patients admitted to the coronary care unit because of unstable angina, Nademanee et al19 found that those who complained of at least one symptomatic episode of ischemia had a higher number (12.3±14 versus 8±7.5) of episodes of silent ischemia on 24-hour Holter ECG recordings compared with the patients showing only asymptomatic transient myocardial ischemia. The lack of a statistically significant difference was probably due to a type 2 error related to the small number of study patients. In addition, in a similar population of patients with unstable angina, Langer et al27 found that the duration of symptomatic episodes was significantly longer and the magnitude of ST-segment shift was greater compared with asymptomatic episodes of transient myocardial ischemia. We do not know how many of our patients had recurrent episodes of silent ischemia in the 24 hours before infarction; however, our group of patients with prodromal angina reasonably may be regarded as more likely to be preconditioned than the group without symptoms before their index infarction. Finally, while coronary angiography during TPA therapy provides the best evidence of reperfusion of the infarct related artery, the use of ECG ST-segment monitoring as a predictor of coronary artery reperfusion, as well as reocclusion after initial restoration of flow, has been validated in a number of studies.20 21 22
Clinical Implications and Conclusions
There have been a number of conflicting reports as to the prognostic significance of angina preceding an acute myocardial infarction. Several studies23 24 have found a better outcome in patients with antecedent angina, even when patients had other important risk factors including severe coronary artery disease, while others25 26 have reported a worse hospital course in patients with previous angina. Differences in patient selection and study protocols, as well as inconsistency in the definition of antecedent angina, could explain the discordant results. None of these studies limited the period of antecedent angina to the 24 hours immediately before infarction. Our study was small, and we are not able to draw any conclusions regarding the prognostic role of prodromal angina.
Our study demonstrated that new-onset prodromal angina appears to afford protection to ischemic myocardium, at least in patients with evolving myocardial infarction who have undergone successful thrombolytic therapy less than 2 hours from the onset of symptoms. Further investigations are needed to elucidate the exact role of this phenomenon in humans and its relation to various treatments and risk factors that affect outcome in patients with acute myocardial infarction.
We thank Jean Ellen Page, RN, for her invaluable assitance throughout the study. We are also indebted to Dr Harvey D. White, from Green Lane Hospital, Auckland, New Zealand, for his support and helpful criticism in improving the manuscript.
- Received May 13, 1994.
- Accepted August 23, 1994.
- Copyright © 1995 by American Heart Association
Murry CE, Jennings RB, Reimer KA. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation. 1986;74:1124-1136.
Cohen MV, Liu GS, Downey JM. Preconditioning causes improved wall motion as well as smaller infarcts after transient coronary occlusion in rabbits. Circulation. 1991;84:341-349.
Alkhulaifi AM, Browne E, Yellon DM. Ischemic preconditioning limits infarct size in the rat heart. J Mol Cell Cardiol. 1992;24(suppl I):S93. Abstract.
Schott RJ, Rohmann S, Braun ER, Schaper W. Ischemic preconditioning reduces infarct size in swine myocardium. Circ Res. 1990;66:113-142.
Murry CE, Richard VJ, Jennings RB, Reimer KA. Myocardial protection is lost before contractile function recovers from ischemic preconditioning. Am J Physiol. 1991;260:796-804.
Blanke H, Scherf F, Karsch KR, Levine RA, Smith H, Rentrop P. Electrocardiographic changes after streptokinase-induced recanalization in patients with acute left anterior descending artery occlusion. Circulation. 1983;68:406-412.
Chesebro JH, Knatterud G, Roberts R, Bore J, Cohen LS, Dalen J, Dodge HT, Francis CK, Hillis D, Ludbrook D, et al. Thrombolysis in Myocardial Infarction (TIMI) Trial, phase I: a comparison between intravenous tissue plasminogen activator versus intravenous streptokinase. Circulation. 1987;76:146-152.
Sheehan FH, Stewart DK, Dodge HT, Mitten S, Bolson BG. Variability in the measurement of regional left ventricular wall motion from contrast angiograms. Circulation. 1983;68:550-559.
Reimer KA, Jennings RB, Cobb FR, Murdock FR, Greenfield JC Jr, Becker LC, Bulkley BH, Hutchins GM, Schwartz RP, Bailey HR. Animal model for protecting ischemic myocardium: results of the NHLBI cooperative study: comparison of unconscious and conscious dog model. Circ Res. 1985;56:651-665.
Deutsch E, Berger M, Kussmaul WG, Hirshfeld JW, Herrmann HC, Laskey WK. Adaptation to ischemia during percutaneous transluminal coronary angioplasty: clinical hemodynamic and metabolic features. Circulation. 1990;82:2044-2051.
Topol EJ, Ellis SG. Coronary collaterals revisited: accessory pathway to myocardial preservation during infarction. Circulation. 1991;83:1084-1086.
Sasayama S, Fujita M. Recent insight into coronary collateral circulation. Circulation. 1992;85:1197-1024.
Rentrop KP, Feit F, Sherman W, Stecy P, Hosat S, Cohen M, Rey M, Ambrose J, Nachamie M, Schwartz W, et al. Late thrombolytic therapy preserves left ventricular function in patients with collateralized total coronary occlusion: primary end point findings of the second Mount Sinai—New York University Reperfusion Trial. J Am Coll Cardiol. 1989;14:58-64.
Habib GB, Heibig J, Forman SA, Brown BG, Roberts R, Terrin ML, Bolli R, TIMI Investigators. Influence of coronary collateral vessels on myocardial infarct size in humans: results of Phase I Thrombolysis in Myocardial Infarction (TIMI) Trial. Circulation. 1991;83:739-746.
Lincoff AM, Topol EJ. Illusion of reperfusion: does anyone achieve optimal reperfusion during acute myocardial infarction? Circulation. 1993;87:1792-1805.
Stern S, Tzivoni D. Early detection of silent ischemic heart disease by 24-hour electrocardiographic monitoring of active subjects. Br Heart J. 1984;36:481-488.
Kruckoff MW, Croll MA, Pope JE, Pieper KS, Kanani PM, Granger CB, Veldkamp RF, Wagner BL, Sawchak ST, Califf RM. Continuously updated ST-segment recovery analysis for myocardial infarct artery patency assessment and its correlation with multiple simultaneous early angiographic observations. Am J Cardiol. 1993;72:145-151.
Muller DWM, Topol EJ, Califf RM, Sigmon KN, Gorman L, George BS, Kereiakes DJ, Lee KL, Ellis SG, Thrombolysis and Angioplasty in Myocardial Infarction (TAMI) Study Group. Relationship between antecedent angina pectoris and short term prognosis after thrombolytic therapy for acute myocardial infarction. Am Heart J. 1990;119:224-231.
Brush JE, Brand DA, Acampora D, Goldman L, Cabin HS. Relation of peak creatine kinase levels during acute myocardial infarction to presence or absence of previous manifestations of myocardial ischemia (angina pectoris or healed myocardial infarction). Am J Cardiol. 1988;62:534-537.
Behar S, Reicher-Reiss H, Abinader E, Agmon J, Friedman Y, Barzilai J, Kaplinsky E, Kauli N, Kishon Y, Palant A, et al. The prognostic significance of angina pectoris preceding the occurence of a first acute myocardial infarction in 4166 consecutive hospitalized patients. Am Heart J. 1992;123:1481-1486.
Barbash GI, White HD, Modan M, Van der Werf F, Investigators of the International Tissue Plasminogen Activator/Streptokinase Trial. Antecedent angina pectoris predicts worse outcome after myocardial infarction in patients receiving thrombolytic therapy: experience gleaned from the International Tissue Plasminogen Activator/Streptokinase Mortality Trial. J Am Coll Cardiol. 1992;20:36-41.