Impact of an Aggressive Invasive Catheterization and Revascularization Strategy on Mortality in Patients With Cardiogenic Shock in the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO-I) Trial
An Observational Study
Background Although retrospective analyses have revealed an association between survival and coronary angiography and angioplasty in patients with acute myocardial infarction complicated by cardiogenic shock, the degree to which bias in the selection of patients to undergo these procedures contributes to this observation remains unclear.
Methods and Results We studied 2200 patients in the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO-I) trial with acute myocardial infarction complicated by cardiogenic shock (systolic blood pressure <90 mm Hg for ≥1 hour) who survived ≥1 hour after the onset of shock to determine the influence of an aggressive strategy of early angiography (within 24 hours of shock onset) and coronary angioplasty or bypass surgery, if appropriate, on survival. Revascularization was not protocol mandated but was selected by the attending physicians. Shock was present on admission in 11% and developed after admission in 89% of shock patients. The 30-day mortality was 38% in the 406 patients who underwent early angiography and were referred within 24 hours for angioplasty (n=175), bypass surgery (n=36), angioplasty and bypass surgery (n=22), or neither (late or no revascularization, n=173) compared with 62% in the 1794 patients who did not (P=.0001). However, there were important differences in the baseline characteristics of the two groups, including younger age (63 versus 68 years, P=.0001), less prior infarction (19% versus 27%, P=.001), and a shorter time to thrombolytic therapy (2.9 versus 3.2 hours, P=.0001) in patients treated with an aggressive strategy. Using multivariate logistic regression analysis to adjust for differences in baseline characteristics, an aggressive strategy was independently associated with reduced 30-day mortality (odds ratio, 0.43 [confidence interval, 0.34 to 0.54], P=.0001).
Conclusions An aggressive strategy of early angiography (and revascularization when appropriate) is associated with a reduction in mortality in patients with acute myocardial infarction and cardiogenic shock who receive thrombolytic therapy.
Cardiogenic shock is the most common cause of death in hospitalized patients with acute myocardial infarction.1 Although the rapid administration of thrombolytic therapy reduces the occurrence of cardiogenic shock,2 3 the best therapy for patients who present with cardiogenic shock or who develop cardiogenic shock despite reperfusion therapy remains controversial. Although retrospective analyses have indicated that survival in such patients who undergo coronary angioplasty is greater than that in patients who do not undergo angioplasty,4 5 6 7 8 9 selection bias may be a confounding variable.10 Differences in the baseline characteristics of patients who undergo angiography and angioplasty have been reported; these may have a major impact on survival and would tend to overestimate the true benefits of angiography and angioplasty.10 The Global Utilization of Streptokinase and Tissue Plasminogen Activator (TPA) for Occluded Coronary Arteries (GUSTO-I) trial was designed to test the hypothesis that thrombolytic strategies achieving more complete, early, sustained coronary artery patency would lead to further reductions in mortality in patients with acute myocardial infarction.11 In the GUSTO-I trial, patients with cardiogenic shock were identified as a subgroup who would be extensively studied, and a specific ancillary data collection form was prospectively designed for this purpose. We analyzed patients with cardiogenic shock in the GUSTO-I trial to determine whether patients treated with an aggressive strategy of early angiography and revascularization, when appropriate, differed from patients in whom a less aggressive strategy was pursued and whether after accounting for such differences, an aggressive strategy was associated with a reduction in 30-day mortality.
The details of the GUSTO-I trial have been described in detail.11 Briefly, 41 021 patients with acute myocardial infarction were randomized to one of four thrombolytic regimens: streptokinase with subcutaneous or intravenous heparin, accelerated TPA with intravenous heparin, or a combination of streptokinase plus TPA with intravenous heparin. Eligible patients had ST-segment elevation of ≥0.1 mV in two or more limb leads or ≥0.2 mV in two or more contiguous precordial leads, onset of ischemic pain within 6 hours of randomization, and no contraindications to thrombolysis. The primary outcome of the trial was all-cause 30-day mortality.
In the trial, 2972 patients met the prospectively defined criteria for cardiogenic shock: systolic blood pressure <90 mm Hg for ≥1 hour not responsive to fluid administration or the need for positive inotropic agents to maintain a systolic blood pressure of >90 mm Hg. Of these, 315 patients had shock on enrollment, and 2657 developed shock after enrollment in the trial. Patients who died within 1 hour of the onset of shock (n=155) were excluded from analysis; the inclusion of patients who did not survive long enough to undergo coronary angiography as if they were treated with a “less aggressive strategy” would bias the study in favor of an aggressive strategy. Patients in whom 30-day survival was not known (n=4) and in whom data regarding the performance and timing of angiography were not available (n=606) were also excluded from analysis. Some patients met more than one exclusion criterion. After these exclusions, there were 2200 patients for analysis (Fig 1⇓).
Patients were divided into those treated with an “an aggressive strategy” in whom coronary angiography was performed within 24 hours of shock onset (group 1, n=406) and the remaining patients (group 2, n=1794). The primary end point of the study was 30-day mortality in group 1 patients compared with group 2 patients. In the GUSTO-I trial, angiography and revascularization were not protocol mandated but were selected by the primary physicians. A secondary analysis was performed of group 1 patients who underwent early coronary angioplasty to determine whether there was a difference in outcome between patients in whom angioplasty was and was not successful and whether differences in baseline clinical characteristics of patients with successful and unsuccessful angioplasty contributed to a difference in outcome between these two groups.
Continuous variables are presented as mean±1 SD, and medians are given with the 25th and 75th percentiles in parentheses. Discrete variables are expressed as frequencies, with percentages given in parentheses. Differences in continuous variables between groups were tested using a Wilcoxon rank-sum test. Differences in discrete variables were examined with the use of χ2 tests. Results were interpreted as statistically significant when P<.05. Multivariate logistic regression analysis was used to determine whether there was an independent association between an aggressive strategy of early angiography (and revascularization if appropriate) and survival, built on a previously reported model of clinical characteristics associated with mortality in the entire GUSTO-I patient population.12 Other variables included in the model were systolic blood pressure, heart rate, infarct location, previous infarction, height, time to thrombolysis, current smoking, history of smoking, diabetes mellitus, weight, previous bypass surgery, thrombolytic agent received, history of hypertension, history of cerebrovascular disease, thrombolytic regimen, and use of an intra-aortic balloon pump. Cross-validation of the 30-day multivariate model produced an adjustment of 0.002 to the operating characteristic area, reducing the value from 0.86 to 0.834. The bootstrapping technique produced exactly the same correction.
The mean age of patients with cardiogenic shock in the study population was 67 years; 62% were male (Table 1⇓). There were important differences in the clinical characteristics of patients who underwent early angiography compared with those who did not, including a younger mean age (63.3 versus 68.4, P=.0001), less frequent prior angina (38% versus 43%, P=.05) and prior infarction (19% versus 27%, P=.001), and a shorter mean time to thrombolytic therapy (2.9 versus 3.2 hours, P=.001). Patients referred for early angiography were more likely to have a family history of coronary artery disease (46% versus 35%, P=.0001) and a history of hyperlipidemia (38% versus 28%, P=.0001).
Frequency of Revascularization in Patients Undergoing Early Angiography
Based on the results of angiography, 255 group 1 patients (63%) underwent revascularization within 24 hours of angiography. Coronary angioplasty alone was performed in 175 patients (43%), and bypass surgery alone was performed in 36 patients (14%); 22 patients underwent both coronary angioplasty and bypass surgery (5%). One hundred seventy-three patients who underwent early angiography were not referred for revascularization within 24 hours (37%). Of these, 45 patients were referred for “late” revascularization a mean of 6±4 days after infarction (23 patients to angioplasty, 20 patients to bypass surgery, and 2 patients to angioplasty followed by bypass surgery).
Adverse cardiovascular events in the 30 days after enrollment in the trial are shown in Table 2⇓ and Fig 2⇓. Mortality was significantly lower in patients referred for early angiography (38% versus 62%, P=.0001). Reinfarction occurred with similar frequency in the two groups (8.9% versus 8.6%, P=.85), although recurrent ischemia was more frequent in patients who underwent early angiography (31% versus 21%, P=.001), as were the development of sustained ventricular tachycardia (32% versus 26%, P=.013) and ventricular septal defect (4.9% versus 1.3%, P=.0001). Asystole occurred less frequently in patients referred for early angiography (28% versus 44%, P=.001).
The mortality difference between patients in whom early angiography was and was not performed remained significant 1 year after enrollment in the trial (44.2% versus 66.4%, respectively; P=.001).
Among patients in the aggressive invasive strategy group in whom coronary angiography revealed either no severe (≥50% luminal diameter) coronary artery stenoses (n=3) or one-vessel disease (n=106), 30-day mortality was 21.9%. In patients with two-vessel disease (n=115), 30-day mortality was 38.3%; and mortality was 41.8% at 30 days in patients with three-vessel disease.
There were 173 patients who underwent early catheterization in whom revascularization was not performed within 24 hours (Fig 2⇑); the mortality of this group of patients was lower than that of the group of patients in whom early angiography was not performed (35% versus 62%, respectively). Forty-five of these patients were referred for “late” revascularization a mean of 6±4 days after infarction; 5 of the 45 patients died (11%) compared with 56 of the 128 early angiography patients (44%) who did not undergo a revascularization procedure. However, to investigate why patients undergoing early angiography without revascularization had a lower mortality rate than patients in whom early angiography was not performed and determine what angiographic findings might have led to a decision not to perform early revascularization, we analyzed the available angiographic data from this group of patients. In the 100 patients in whom angiographic data were available for analysis, the mean luminal diameter stenosis of the infarct-related artery was 87±20%. Thirty patients had a infarct vessel diameter stenosis of <80%, 37 patients had normal (TIMI 3) flow in the infarct artery at the time of angiography, and another 22 patients had TIMI 2 flow.
To determine whether the reduction in mortality associated with an aggressive strategy of early angiography and revascularization, when appropriate, was due to the important differences in clinical characteristics, multivariate logistic regression analysis was performed. This analysis revealed that an aggressive strategy was independently associated with a reduction in 30-day mortality, with an odds ratio of 0.43 (confidence interval, 0.34 to 0.54; P=.0001).
There was no interaction between thrombolytic treatment received and the performance of an aggressive strategy (Δχ2=1.165, P=.28).
Relationship Between Outcome of Coronary Angioplasty and Mortality
To further assess the importance of coronary angioplasty in patients with cardiogenic shock, we examined the 30-day mortality rate in the 197 patients in whom early angioplasty was performed (Fig 2⇑). Angioplasty was successful (<50% residual stenosis) in 148 patients (75%) and unsuccessful in 49 patients (25%). The 30-day mortality rate of patients in whom angioplasty was successful was 35% versus 55% in patients in whom it was unsuccessful (P=.007).
To determine whether the lower mortality associated with successful angioplasty could be accounted for by differences in baseline characteristics between patients in whom angioplasty was and those in whom it was not successful, we compared the clinical characteristics of the two groups. There were no significant differences between these two groups in any of the characteristics listed in Table 1⇑ that may have contributed to the difference in mortality.
Use of an Intra-aortic Balloon Pump
An intra-aortic balloon pump was placed in 302 patients (74.6%) in the aggressive invasive strategy group and in 141 (7.9%) of patients in the conservative strategy group. Among patients in whom an aggressive invasive strategy was performed and in whom an intra-aortic balloon pump was used, the balloon pump was placed the day catheterization was performed in 93% of patients, before the day of catheterization in 3%, and after the catheterization in 4% of patients. The mortality of patients in the aggressive invasive strategy group with a balloon pump was 40.7% (n=123) versus 31.1% (n=32) in patients in the aggressive invasive strategy group without a balloon pump. The mortality of patients with an intra-aortic balloon pump in the conservative strategy group was 56.7% (n=80) versus 61.9% (n=1017) in patients without a balloon pump in the conservative strategy group. When the placement of an intra-aortic balloon pump was included in the multivariate analysis, the results indicated that the use of an intra-aortic balloon pump was not a predictor of outcome independent of an aggressive invasive strategy (Δχ2=0.294, P=.59).
The most important finding of this study is that an aggressive strategy of early angiography and revascularization, when deemed to be appropriate by the attending physician, was associated with a significant reduction in 30-day mortality in patients with acute myocardial infarction and cardiogenic shock treated with thrombolytic therapy. The benefit was independent of differences in baseline clinical characteristics between this group and patients treated with a less aggressive strategy. Furthermore, the lower mortality rate in patients in whom angioplasty was successful versus those in whom it was unsuccessful could not be explained by differences in clinical characteristics, further supporting the hypothesis that an aggressive revascularization strategy may reduce mortality in such patients.
Many studies have suggested that angioplasty reduces the mortality of patients with acute myocardial infarction complicated by cardiogenic shock.4 5 6 7 8 9 However, these were retrospective analyses, and it has been shown that significant bias exists in the selection of shock patients referred for coronary angioplasty.10 This raises the possibility that more favorable clinical characteristics might have accounted for the lower mortality in patients referred for angioplasty. Other analyses indicating a lower mortality in shock patients in whom angioplasty was successful compared with patients in whom angioplasty was unsuccessful have also raised questions about the degree to which differences in clinical characteristics contributed to the mortality difference. Some investigators have wondered whether unsuccessful angioplasty might actually increase mortality and account, in part, for the relatively lower mortality in patients with successful angioplasty. These studies have been too small to permit multivariate analysis to adjust for baseline differences, and therefore the degree to which baseline characteristics contributed to the lower mortality associated with angioplasty could not be determined. Therefore, in the present study, the largest cohort of cardiogenic shock patients prospectively identified in a randomized trial, we compared the outcome of the patients who were referred for early angiography (within 24 hours of shock onset) and revascularization (within 24 hours of angiography, if considered to be appropriate by the attending physician) to the remaining patients. This analysis revealed that patients referred for early angiography had significantly fewer risk factors associated with death in the trial, including younger age, less prior infarction, and earlier treatment with thrombolytic therapy. However, after adjustment for these and all other risk factors associated with mortality in GUSTO-I, an aggressive strategy of early angiography and revascularization, if appropriate, was still associated with a lower mortality, with an odds ratio of 0.43 and confidence interval of 0.34 to 0.54 (P=.0001).
An intra-aortic balloon pump was used in many patients with shock in the GUSTO-I trial at the discretion of the clinicians caring for the patient; balloon pumps were used most frequently in patients undergoing an aggressive invasive strategy. Although it is possible that the use of an intra-aortic balloon pump may be beneficial in the treatment of shock patients, and particularly in those patients with shock who are undergoing coronary angiography and angioplasty or bypass surgery, the retrospective nature of this study does not permit an appropriate analysis of whether an intra-aortic balloon pump is useful in such patients.
In this study, we sought to eliminate three sources of bias that have characterized prior studies of angioplasty in cardiogenic shock. One is the failure to exclude patients with cardiogenic shock who died before angiography could have been performed. This biases a study in favor of an aggressive strategy, since such patients are considered to have been treated conservatively. One hour is required in most hospitals with a catheterization laboratory to perform emergency angiography; accordingly, in this study, we eliminated from analysis 155 patients who died within 1 hour of the diagnosis of shock. However, to determine whether the elimination of deaths occurring within 1 hour of shock onset was sufficient to avoid bias, we performed additional analyses to determine whether the elimination of patients who died within 3 hours of shock onset and within 6 hours of shock onset changed the results. If all deaths occurring within 3 hours of shock onset are excluded, an aggressive invasive strategy was still associated with a lower mortality (155 deaths [38%] versus 1100 deaths [61%]). The lower mortality associated with an aggressive invasive strategy persisted when deaths within 6 hours of shock onset were excluded (122 [33%] versus 701 [50%].
A second source of bias is the inclusion of patients who survive the initial phases of cardiogenic shock and who undergo angiography and revascularization late in their hospital course. This also biases a study in favor of an aggressive strategy, since such patients who survive to undergo late angiography cannot be considered to have survived because of the angiography and revascularization procedure. In the present study, 302 of the 1794 patients in the conservative group did ultimately undergo angiography during their initial hospitalization at a median of 4.5 days after shock onset. In this analysis, patients who underwent angioplasty and bypass surgery late in their hospitalization were included in the conservative strategy group.
A third source of bias is the failure to account for differences in the baseline characteristics of patients who are and those who are not selected for an aggressive strategy. Such differences did exist in our study population; failure to adjust for more viable patients being referred for angiography and revascularization also biases a study in favor of an aggressive strategy. We adjusted for the observed significant differences in baseline characteristics and for all the other nonsignificant differences in the other variables associated with mortality in the GUSTO-I study and found that an aggressive strategy was nonetheless associated with a marked reduction in mortality independent of baseline characteristics. To further investigate whether selection bias may have contributed to the results of our study, we compared patients in whom angioplasty was and was not successful. We found that patients in whom angioplasty was successful had a significantly lower mortality (35%) than patients in whom it was not successful (55%), and there were no differences between the two groups in any of the baseline characteristics listed in Table 1⇑ that might account for the observed association between successful angioplasty and improved survival.
Although patients with cardiogenic shock were prospectively identified as a subgroup for further analysis and a data form was specifically designed for this purpose, this study is nonetheless a retrospective analysis and is subject to the limitations of such analyses. It is possible that differences in baseline clinical characteristics that were not identified or differences in clinical characteristics at the time of angiography that were not analyzed contributed to the difference in outcome between the groups. Since the decisions to perform coronary angiography and revascularization were not protocol mandated but were made by the patient’s attending physician, the possibility exists that differences in clinical or angiographic characteristics other than those analyzed in the GUSTO-I trial may have contributed to the association between an aggressive strategy and a reduced mortality. Furthermore, there are limitations to the use of multivariate analysis to adjust for even recognized differences between groups when many such differences exist.
In patients with acute myocardial infarction and cardiogenic shock who receive thrombolytic therapy, an aggressive strategy of early angiography and revascularization, if appropriate, is associated with a marked reduction in 30-day mortality compared with patients who are treated with a less aggressive strategy independent of differences in baseline clinical characteristics between the two groups. Furthermore, the lower mortality in patients whose angioplasty was successful compared with those in whom it was unsuccessful cannot be accounted for by differences in clinical characteristics, further supporting the hypothesis that an aggressive revascularization strategy reduces mortality in patients with acute myocardial infarction and cardiogenic shock.
The authors gratefully acknowledge Penny Hodgson and Pat Williams for their important editorial contributions.
- Received September 9, 1996.
- Revision received January 22, 1997.
- Accepted January 23, 1997.
- Copyright © 1997 by American Heart Association
Wilcox RG, von der Lippe G, Olsson CG, Jensen G, Skene AM, Hampton JR. Trial of tissue plasminogen activator for mortality reduction in acute myocardial infarction: Anglo-Scandinavian Study of Early Thrombolysis (ASSET). Lancet.. 1988;1:525-530.
Lee L, Bates ER, Pitt B, Walton JA, Laufer N, O’Neill WW. Percutaneous transluminal coronary angioplasty improves survival in acute myocardial infarction complicated by cardiogenic shock. Circulation.. 1988;78:1345-1351.
Verna E, Repetto S, Boscarini M, Ghezzi I, Binaghi G. Emergency coronary angioplasty in patients with severe left ventricular dysfunction or cardiogenic shock after acute myocardial infarction. Eur Heart J.. 1989;10:58-66.
Meyer P, Blanc P, Baudouy M, Morand P. Treatment de choc cardiogenique primaire par angioplastie transluminale coronarienne a la phase aigue de l’infarctus. Arch Mal Coeur.. 1990;83:329-334.
Hochman JS, Boland J, Sleeper JA, Porway M, Brinker J, Col J, Jacobs A, Slater J, Miller D, Wasserman H, Menegus MA, Talley JD, McKinlay S, Sanborn T, LeJemtel T, and the SHOCK Registry Investigators. Current spectrum of cardiogenic shock and effect of early revascularization on mortality. Circulation.. 1995;91:873-881.
Lee KL, Woodlief LH, Topol EJ, Weaver WD, Betriu A, Col J, Simoons M, Aylward P, Van de Werf F, Califf RM, for the GUSTO-I Investigators. Predictors of 30-day mortality in the era of reperfusion: results from an international trial of 41 021 patients. Circulation.. 1995;91:1659-1668.