Acute Coronary Syndromes in the GUSTO-IIb Trial
Prognostic Insights and Impact of Recurrent Ischemia
Background—Recurrent ischemia after an acute coronary syndrome portends an unfavorable outcome and has major resource-use implications. This issue has not been studied systematically among the spectrum of patients with acute coronary presentations, encompassing those with and without ST-segment elevation.
Methods and Results—We assessed the 1-year prognosis of the 12 142 patients enrolled in the GUSTO-IIb trial by the presence (n=4125) or absence (n=8001) of ST-segment elevation. This latter group was further categorized into those with baseline myocardial infarction (n=3513) or unstable angina (n=4488). We also assessed the incidence of recurrent ischemia and its impact on outcomes. Recurrent ischemia was significantly rarer in those with ST-segment elevation (23%) than in those without (35%; P<0.001). Mortality at 30 days was greater among patients with ST-segment elevation (6.1% versus 3.8%; P<0.001) but less so at 6 months; by 1 year, mortality did not differ significantly (9.6% versus 8.8%). Patients with non–ST-segment-elevation infarction had higher rates of reinfarction at 6 months (9.8% versus 6.2%) and higher 6-month (8.8% versus 5.0%) and 1-year mortality rates (11.1% versus 7.0%) than such patients who had unstable angina.
Conclusions—Refractory ischemia was associated with an approximate doubling of mortality among patients with ST-segment elevation and a near tripling of risk among those without ST elevation. This study highlights not only the substantial increase in late mortality and reinfarction with non–ST-segment-elevation infarction but also the opportunities for better triage and application of therapeutic strategies for patients with recurrent ischemia.
Ischemic heart disease remains the most common cause of death worldwide.1 In the recent Global Use of Strategies To Open Occluded Arteries in Acute Coronary Syndromes (GUSTO-IIb) trial, 9.4% of patients died or suffered a myocardial infarction (MI) at 30 days despite vigorous medical management.2 Patients who presented with ST-segment elevation had a higher risk of death at 30 days than those without ST elevation; however, those without ST-segment elevation were not further classified by the presence of MI or unstable angina.
Recurrent ischemia after presentation with acute coronary syndromes, as well as myocardial (re)infarction, portends an unfavorable outcome and has major implications for healthcare resource use.3 4 5 6 7 Data from small to modest observational studies suggest that recurrent ischemia is more common among patients presenting without ST-segment elevation as opposed to those with ST elevation, but precise estimation of the frequency and natural history of recurrent ischemia in a contemporary practice environment is lacking.4 8 Some studies have suggested that patients presenting with ST elevation who receive thrombolysis may be at even greater risk for recurrent ischemia,4 9 but no large study has systematically examined this issue among patients with the spectrum of acute coronary presentations, encompassing those with and without ST elevation.
Accordingly, we examined these issues among the spectrum of acute coronary syndromes in the 12 142 patients enrolled in the GUSTO-IIb trial, with the following objectives: (1) to evaluate the long-term outcome (the 1-year prognosis) of patients presenting with ST elevation versus no ST elevation, who were in turn partitioned by whether they presented with MI or unstable angina; (2) to assess the incidence of recurrent ischemia among these subsets and evaluate its impact on survival and nonfatal (re)infarction; and (3) to determine baseline variables predictive of recurrent ischemia and unfavorable outcome.
The GUSTO-IIb trial, the results of which have been published previously,2 compared the efficacy of recombinant hirudin with that of heparin in patients with unstable angina or acute MI. Its 12 142 patients were stratified into 3 categories: those with ST-segment elevation (n=4125) and those without ST-segment elevation (n=8001); this latter group was then subdivided into those with MI (those whose index event was associated with an elevated creatine kinase–MB level that was also ≥3% of total creatine kinase at baseline or 8 hours after enrollment; n=3513) and those with unstable angina (n=4488). Both MI and myocardial reinfarction events were independently reviewed by a clinical events committee blinded to treatment assignment for purposes of final analyses, as described previously.2
Nonrefractory recurrent ischemia was defined as symptoms of ischemia, with ST-segment deviation or definite T-wave inversion, and/or new hypotension, pulmonary edema, or cardiac murmur thought by the physician to represent myocardial ischemia while the patient was hospitalized. Refractory recurrent ischemia consisted of symptoms of ischemia with ECG changes persisting for ≥10 minutes despite the use of nitrates and either β-blockers or calcium-channel blockers. For purposes of the present study, patients were classified into 3 groups: those with no recurrent ischemia and those with recurrent ischemia, who were subdivided into either nonrefractory or refractory ischemic groups. The end points of (re)infarction and death were available at 6 months and mortality data at 12 months.
Baseline characteristics of the patients were summarized in terms of percentages for categorical variables and medians and 25th and 75th percentiles for continuous variables. A polytomous logistic regression procedure was used to identify the baseline predictors of nonrefractory and refractory ischemia.10 In addition, logistic regression methods were used to predict 1-year mortality separately for the ST-elevation and non–ST-elevation groups, in which recurrent ischemia was categorized by the number of days before its occurrence (0 to 1 day, 2 to 3 days, or >3 days). Revascularization procedures were categorized similarly to account for their effects on mortality. Univariate analyses of potential predictors were performed first, followed by backward, stepwise variable-selection procedures. The initial models were determined after their assumptions, the shapes and scales of the variables, and any misfitting and influential cases were examined. For ease of interpretation and to ensure comparability, the odds ratio was assessed for a patient at the 75th percentile against a patient at the 25th percentile for each continuous variable.11 Kaplan-Meier curves and log-rank tests were used to further characterize and compare group survival rates over time.
The overall performance of the models was assessed in terms of the Hosmer-Lemeshow χ2 test and the C-index (or the area under a receiver-operator characteristic curve). Because the data involved time-dependent covariates (such as recurrent ischemia and revascularization procedures) and censoring at 1 year, the Cox regression model with time-dependent covariates was used to validate the logistic regression model for 1-year mortality. The results of the 2 models were similar, as expected; hence, only the logistic regression model is presented in this article. In addition, a bootstrap resampling technique was used to further validate these models.12 Each model was tested on 200 to 300 bootstrap samples drawn with replacement from the original sample. The percentage of times the variables became statistically significant with these bootstrap samples was used as a measure of their importance in the predictive model.
The cutoff of 50% is often used to identify significant variables. The degree of bias associated with the original coefficient estimates was evaluated by the distribution of the bootstrap coefficients, the normality of which was assessed by the Kolmogorov-Smirnov test. Bias was judged as not a serious concern if it was estimated to be within 25% of the standard error.11 All tests were 2-tailed, with a P<0.05 level of significance. All analyses were performed with SPSS (version 7.5) and STATA (release 5) software.
Compared with the group with ST elevation on admission, patients without baseline ST-segment elevation were older and had higher prevalence of risk factors for coronary disease, and fewer were males (Table 1⇓). Prior MI and angina, cerebral and peripheral vascular disease, heart failure, previous angioplasty, and coronary bypass surgery all were more common in patients without ST elevation on admission. Although time to hospital presentation was similar in the 2 groups, the total time to study treatment was longer in those without ST elevation. Angiography was performed more often in patients with ST elevation, but coronary bypass surgery was performed more often in those without ST elevation. Use of β-blockers and aspirin was common, with a similar frequency in both groups. Calcium-channel blockers were used much more often in patients without ST elevation, whereas ACE inhibitors were more commonly prescribed in patients with ST elevation. Heart failure, shock, and stroke were more common in patients with ST elevation, and bleeding occurred in 9% of both groups.
Within the non–ST-elevation patients, those with unstable angina were more often female, less commonly white, and more often hypertensive than those with MI at baseline. Prior MI and angina were more common in patients with unstable angina, as were previous coronary bypass, angioplasty, and heart failure. Patients with unstable angina had a higher systolic blood pressure on admission, a slightly lower heart rate, more commonly were in Killip class I, and presented 30 minutes earlier after symptom onset to the hospital. Heart failure and shock were substantially less frequent in patients with unstable angina than in MI patients. As per the protocol, thrombolytic therapy (a choice left to the principal investigator) was used in 74% of patients with ST-segment elevation, and primary angioplasty (as part of a substudy comparing angioplasty with thrombolysis) was used in 11% of patients.
Recurrent ischemia was significantly less common (23%) among those presenting with ST elevation than among those without ST elevation (35%) (P<0.001; Figure 1⇓). The rate of recurrent ischemia was similar in those without ST elevation whether they had MI or unstable angina on admission. The distribution of recurrent ischemia into nonrefractory versus refractory categories was homogeneous across categories of acute coronary syndromes; ≈1 in 5 patients with recurrent ischemia was classified as refractory.
Although the mortality rate at 30 days was greater among patients with ST-segment elevation than among those without ST elevation, this difference narrowed at 6 months and disappeared at 1 year (Table 2⇓). Reinfarction rates at 6 months were similar between these 2 categories. Categorization of non–ST-elevation patients into MI and unstable angina groups revealed distinct patterns of mortality and reinfarction: a consistently higher rate of (re)infarction and death occurred in patients without ST elevation who had MI on admission than among those with unstable angina.
Among patients without recurrent ischemia (Figure 2⇓), the survival rates for non–ST-elevation MI patients at 30 days, although better than that of patients presenting with ST elevation, crossed at 6 months and were nearly identical at 1 year. Although patients with unstable angina and no recurrent ischemia had a low mortality rate at 30 days, their survival curve showed a steeper decline than that evident in MI patients, and mortality had nearly tripled by 1 year.
Recurrent ischemia, especially when it was refractory, was associated with a substantial increase in reinfarction at 30 days and 6 months in all patient categories (Table 2⇓). As shown in Figure 3⇓, refractory ischemia had a major negative impact on survival in both ST-segment categories (P<0.03). Nonrefractory ischemia also affected survival negatively in both categories (P=0.32 and P<0.02, respectively). Patients without ischemia were more likely to survive 1 year if they presented without ST elevation than with ST elevation (P<0.002); however, this largely reflected the much better survival in unstable angina patients. The negative impact of recurrent ischemia was also evident among patients without ST-segment elevation (Figure 4⇓), whether they had MI on admission or unstable angina. Irrespective of the presence and type of ischemia (nonrefractory or refractory), those with non–ST-elevation MI on admission had a significantly worse 1-year survival than did patients with unstable angina. Within the MI and unstable angina groups, 1-year survival differed significantly according to the type of ischemia, except that in unstable angina patients, there was only a trend for nonrefractory ischemia to confer worse survival than no ischemia (P=0.08).
Table 3⇓ presents baseline characteristics by ST-segment status and by the presence or absence of recurrent ischemia. For patients with ST elevation, those who underwent primary angioplasty experienced recurrent ischemia less often than those who received thrombolysis (11.2% and 2.6% for nonrefractory and refractory ischemia with primary PTCA versus 18.5% and 5.6% with thrombolysis; P=0.0001). In both ST-segment strata, those with recurrent ischemia were older, more frequently female, and more likely to have prior hypertension. Patients with recurrent ischemia were less likely to be current smokers and more likely to have prior infarction, angina pectoris, coronary bypass grafting, and angioplasty. Patients with recurrent ischemia had a higher frequency of revascularization and more frequent therapy with nitrates, calcium-channel blockers, ACE inhibitors, and β-blockers. These patients also had more frequent bleeding and higher incidences of heart failure and shock.
Significant baseline predictors of nonrefractory and refractory ischemia, together with their odds ratios and adjusted χ2 statistics, are presented in Figure 5⇓. The most important risk factor for both nonrefractory and refractory ischemia was a history of previous angina. Other risk factors common to both types of recurrent ischemia were non–ST-segment-elevation MI, white race, enrollment outside the United States, and past smoking. Previous cerebrovascular disease and current smoking were significant risk factors for nonrefractory but not refractory ischemia. On the other hand, significant risk factors specific to refractory ischemia included low weight, a high body mass index, hypertension, diabetes, and old age. None of the interaction terms were significant. As expected, the predictive power based solely on baseline data was modest, as indicated by the C-index values of 0.63 and 0.66, respectively, for predicting nonrefractory and refractory ischemia.
The 1-year mortality model is summarized in Figure 6⇓. After adjustment for baseline characteristics and revascularization procedures, recurrent ischemia remained a significant risk factor in both ST-segment strata, especially for those ischemic incidents that occurred within 3 days of enrollment. Age was by far the most significant predictor, as indicated by the adjusted χ2 values of 173 and 168 for the ST-elevation and non–ST-elevation groups, respectively. Other highly significant, independent predictors of 1-year mortality for both ST-segment strata included angioplasty, heart rate, history of diabetes, and recurrent ischemia. Baseline Killip class, systolic blood pressure, and bypass surgery were significant predictors for those with ST-segment elevation but not for those without. Within the non–ST-elevation stratum, whether the index event was an MI was a significant predictor of 1-year mortality. Other significant factors specific to this stratum included previous MI, previous congestive heart failure, and diastolic blood pressure. The only significant interaction was between age and previous MI, but the predictive power was only minimally enhanced by including this interaction term in the model. The C-index values were 0.82 and 0.80 for the ST-elevation and the non–ST-elevation models, respectively, reflecting their excellent ability to discriminate between patients who did and did not die within 1 year.
We studied >12 000 patients, comprising well-defined subsets, concurrently and monitored them for 1 year. Data from these patients provide new prognostic insights into the various components of acute coronary syndromes. Although patients presenting without ST elevation are commonly included in clinical trials, it is clear from our observations that they not only have distinctive characteristics at presentation but also exhibit outcomes markedly different from those of patients with ST-segment elevation during 1-year follow-up. Moreover, although patients without ST elevation appear to have better survival at 30 days than those with ST-segment elevation, this advantage attenuates substantially at 6 months and disappears by 1 year. This temporal pattern is largely mediated by events within the non–ST-elevation MI cohort, although similar trends of lesser magnitude are evident in patients with unstable angina.
Recurrent ischemia, nonrefractory or refractory, even when noted only during the hospitalization, had a major effect on the frequency of later MI at 30 days and 6 months, as well as on survival through 1-year follow-up. Refractory ischemia was associated with an approximate doubling of risk of death among ST-elevation patients compared with those without recurrent ischemia. In patients without ST elevation on admission, refractory ischemia had an even greater negative impact: an approximate tripling of their risk of mortality with its occurrence. Moreover, patients without ST elevation on admission who had refractory ischemia and MI on admission had the worst survival of any patient category at all time points.
Comparison of Patients by ST-Segment Elevation Status
Although there was substantially more recurrent ischemia among patients without ST-segment elevation on admission than among those with ST-segment elevation, once it had occurred, there was remarkable homogeneity in the distribution of refractory versus nonrefractory types of ischemia. There was also symmetry in the frequency of recurrent ischemia among MI and unstable angina patients within the non–ST-elevation group. Importantly, we have confirmed the experience of Stone and coworkers9 in a smaller study of this population by showing less recurrent ischemia in patients who underwent reperfusion with primary PTCA than in those who received thrombolysis. This was also evident in the prospective substudy of primary angioplasty in the GUSTO-IIb trial.13
The basis for the steeper increase in mortality from 30 days to 1 year (Table 2⇓ and Figure 2⇑) among unstable angina patients without recurrent ischemia than among those with ST-segment elevation or non–ST-elevation MI is unclear, but it is important to appreciate that only symptomatic ischemia was identified during hospitalization in the present study. The majority of ischemia during this time is silent, and its frequency and duration bear a direct relationship to prognosis.14 15 Better prognostic triage and selective therapeutic strategies for such patients might be possible with the use of novel cardiac markers of myocardial necrosis acquired early in the course of the event.16 17
The present study highlights not only the major increase in late mortality and reinfarction among patients who present without ST elevation but also the potential opportunities for therapy. New pharmacological strategies have thus far primarily shown good short-term but lesser long-term therapeutic effects. This problem may relate in part to heterogenous pathophysiology, rebound ischemia, and only temporary passivation of the coronary lesion. Strategies that involve both acute and sustained therapy with low-molecular-weight heparins, other antithrombin agents, platelet glycoprotein IIb/IIIa inhibitors, and/or coronary stenting offer substantial promise and are currently being investigated.18 19
It is uncertain whether performing angiography and revascularization even more frequently than was done in our patients without ST-segment elevation would have yielded better long-term outcomes; this deserves further clinical investigation.20 Recently, the VA Non–Q-Wave Infarction Strategies in-Hospital (VANQWISH) investigators, in a randomized trial of acute coronary syndromes without ST elevation, showed a higher frequency of unfavorable outcomes among those randomized to an aggressive invasive/interventional strategy.21 These workers, like others, suggest noninvasive triage and “watchful waiting” may be most beneficial.20 Identification of high-risk baseline factors, such as those shown in the current study, along with other diagnostic triage might allow more appropriate selection of therapeutic strategies.
In any case, the present study highlights the distinctive characteristics of patients with versus those without ST-segment elevation among the acute coronary syndromes and emphasizes the important differences between patients who have MI and patients who have unstable angina in the latter category. The major effect of symptomatic recurrent ischemia, especially when it is refractory to medical management, emphasizes the need for continuing vigilance and exploration of enhanced therapeutic strategies.
The GUSTO-IIb trial was funded by grants from Boehringer-Mannheim, Mannheim, Germany; Ciba-Geigy Corporation, Summit, NJ; and Advanced Cardiovascular Systems, Mountain View, Calif. It is a pleasure to acknowledge the secretarial assistance of Anita Penny Hodgson and the editorial assistance of Patricia Williams.
- Received March 24, 1998.
- Revision received July 6, 1998.
- Accepted July 21, 1998.
- Copyright © 1998 by American Heart Association
Bosch X, Theroux P, Waters DD, Pelletier GB, Roy D. Early postinfarction ischemia: clinical, angiographic, and prognostic significance. Circulation. 1987;75:988–995.
Barbagelata A, Granger CB, Topol EJ, Worley SJ, Kereiakes DJ, George BS, Ohman EM, Leimberger JD, Mark DB, Califf RM, for the TAMI Study Group. Frequency, significance and cost of recurrent ischemia after thrombolytic therapy for acute myocardial infarction. Am J Cardiol. 1995;76:1007–1013.
Fioretti P, Brower RW, Balakumaran K. Early post-infarction angina: incidence and prognostic relevance. Eur Heart J. 1986;7(suppl C):73–77.
Stone GW, Grines CL, Browne KF, Marco J, Rothbaum D, O’Keefe J, Hartzler GO, Overlie P, Donohue B, Chelliah N, Timmis GC, Vlietstra R, Puchrowicz-Ochocki S, O’Neill WW. Implications of recurrent ischemia after reperfusion therapy in acute myocardial infarction: a comparison of thrombolytic therapy and primary angioplasty. J Am Coll Cardiol. 1995;26:66–72.
Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons; 1989.
Lee KL, Woodlief LH, Topol EJ, Weaver WD, Betriu A, Col J, Simoons M, Aylward P, Van de Werf F, Califf R. Predictors of 30-day mortality in the era of reperfusion for acute myocardial infarction: results from an international trial of 41 021 patients. Circulation. 1995;91:1659–1668.
Efron B. The Jackknife, the Bootstrap and Other Resampling Plans. Philadelphia, Pa: Society for Industrial and Applied Mathematics; 1982.
The GUSTO-IIb Angioplasty Substudy Investigators. A clinical trial comparing primary coronary angioplasty with tissue plasminogen activator for acute myocardial infarction: The Global Use of Strategies To Open occluded coronary arteries in acute coronary syndromes (GUSTO-IIb). N Engl J Med. 1997;336:1621–1628.
Langer A, Krucoff MW, Klootwijk P, Simoons ML, Granger CB, Barr A, Califf RM, Armstrong PW. Prognostic significance of ST-segment shift early after resolution of ST elevation with myocardial infarction treated with thrombolytic therapy: the GUSTO-I ST Segment Monitoring Substudy. J Am Coll Cardiol. 1998;31:783–789.
Ohman EM, Armstrong PW, Christenson RH, Granger CB, Katus HA, Hamm CW, O’Hanesian MA, Wagner GS, Kleiman NS, Harrell FE Jr, Califf RM, Topol EJ, for the GUSTO IIa Investigators. Cardiac troponin T levels for risk stratification in acute myocardial ischemia. N Engl J Med. 1996;335:1333–1341.
Cohen M, Demers C, Gurfinkel EP, Turpie AGG, Fromell GJ, Goodman S, Langer A, Califf RM, Fox KAA, Premmereur J, Bigonzi F, Stephens J, Weatherley B, for the Efficacy and Safety of Subcutaneous Enoxaparin in Non-Q-wave Coronary Events (ESSENCE) Study Group. A comparison of low-molecular-weight heparin with unfractionated heparin for unstable coronary artery disease. N Engl J Med. 1997;337:447–452.
Topol EJ. Toward a new frontier in myocardial reperfusion therapy: emerging platelet pre-eminence. Circulation. 1998;97:211–218.
The TIMI IIIB Investigators. Effects of tissue plasminogen activator and a comparison of early invasive and conservative strategies in unstable angina and non-Q wave myocardial infarction: results of the TIMI IIIB Trial. Circulation. 1994;89:1545–1556.
Ferguson JJ. Meeting highlights: 46th Annual Scientific Sessions of the American College of Cardiology. Circulation. 1997;96:367–371.