Stroke After Thrombolysis
Mortality and Functional Outcomes in the GUSTO-I Trial
Background Stroke is the most feared complication of thrombolysis for acute myocardial infarction because of the resulting mortality and disability. We analyzed the incidence, timing, and outcomes of stroke in an international trial.
Methods and Results Patients were randomly assigned to one of four thrombolytic strategies. Neurological events were confirmed clinically and anatomically and were adjudicated by a blinded committee. Stroke survivors, categorized by residual deficit and disability, assessed their quality of life with a time trade-off technique. Multivariable regression identified patient characteristics associated with intracranial hemorrhage. Overall, 1.4% of the patients had a stroke (93% anatomic documentation). The risk ranged from 1.19% with streptokinase/subcutaneous heparin therapy to 1.64% with combination thrombolytic therapy (P=.007). Primary intracranial hemorrhage rates ranged from 0.46% with streptokinase/subcutaneous heparin to 0.88% with combination therapy (P<.001). Of all strokes, 41% were fatal, 31% were disabling, and 24% were nondisabling, with no significant treatment-related differences. Stroke subtype affected prognosis: 60% of patients with primary intracranial hemorrhage died and 25% were disabled versus 17% dead and 40% disabled with nonhemorrhagic infarctions. Patients with moderate or severe residual deficits showed significantly decreased quality of life. Advanced age, lower weight, prior cerebrovascular disease or hypertension, systolic and diastolic blood pressures, randomization to tissue plasminogen activator, and an interaction between age and hypertension were significant predictors of intracranial hemorrhage.
Conclusions Stroke remains a rare but catastrophic complication of thrombolysis. Additional studies should assess the net clinical benefit of thrombolysis in high-risk subgroups, particularly the elderly and patients with prior cerebrovascular events.
The overall risk of stroke from thrombolytic therapy for acute myocardial infarction in properly selected patients is low compared with the impressive reduction in mortality associated with the use of thrombolytic agents.1 2 Nevertheless, the most feared complication of thrombolytic therapy is intracranial hemorrhage, since fatality rates can range from 44% to 75%.3 4 5 Nonhemorrhagic stroke also remains a major cause of death and disability in this population.6 The differential occurrence of stroke with different approaches to myocardial reperfusion7 has become a major issue in discussions of treatment selection for individual patients.8 9 10
Previous thrombolytic trials have demonstrated higher rates of stroke with conventional dosing of tissue plasminogen activator (TPA) versus streptokinase. No information was available from large trials comparing accelerated administration of TPA with streptokinase or comparing intravenous heparin with subcutaneous heparin until the initial report from the GUSTO-I trial.11 Furthermore, analyses of previous trials have been hampered by small sample sizes and incomplete identification of stroke type. Categorization of stroke type has been particularly difficult because brain imaging has not been mandatory in clinical trials. Additionally, no studies to date have collected detailed information about functional status after stroke in this setting. This information is crucial, since the assessment of relative treatment outcomes can more realistically reflect patient values by distinguishing patients who survive a cerebrovascular event with major disability from those with minimal or no residual deficit.
The GUSTO-I trial was designed to assess the effects of four thrombolytic treatment strategies on rapid and sustained patency of the infarct-related artery.11 The purposes of this analysis are to provide the final report on the incidence and types of strokes occurring in GUSTO-I and to examine the clinical correlations and functional outcomes of patients experiencing specific types of acute cerebrovascular events in each treatment arm.
The GUSTO-I study included 41 021 patients with acute myocardial infarction from 1081 hospitals in 15 countries. We examined differences in all-cause, 30-day mortality among four thrombolytic strategies, working under the hypothesis that early, sustained patency of the infarct-related artery would be associated with improved survival. Details of the study, including descriptive characteristics of the patient population, treatment strategies, end points, data management, and quality assurance, have been reported previously.11
Exclusion criteria included previous stroke, active bleeding, and recent trauma or major surgery. Patients with prior transient ischemic attacks or carotid artery disease were eligible for inclusion.
Thrombolytic and Adjunctive Therapies
Patients were randomly assigned to receive one of four thrombolytic strategies: streptokinase with subcutaneous heparin, streptokinase with intravenous heparin, accelerated TPA with intravenous heparin, and streptokinase plus TPA with intravenous heparin. Subcutaneous heparin was continued for 7 days or until discharge; intravenous heparin was given for at least 48 hours and longer at the investigators’ discretion. The dose of intravenous heparin was adjusted according to the activated partial thromboplastin time at 6, 12, and 24 hours after the start of thrombolytic therapy.
Chewable aspirin (≥160 mg) was given as soon as possible and daily thereafter (160 to 325 mg). Intravenous atenolol was given to patients without contraindications in two doses of 5 mg each, followed by oral doses of 50 to 100 mg per day. All other medications, including nitrates, antiarrhythmics, calcium-channel blockers, angiotensin-converting enzyme inhibitors, and digitalis, were given at the discretion of the attending physician. Performance of coronary angiography, angioplasty, or bypass surgery was also at the direction of the attending physician.
Reporting of Adverse Neurological/Cerebrovascular Events
Suspected strokes occurring in GUSTO-I were reported by investigators at each participating center to the Duke Coordinating Center, and additional data were collected for final classification. Stroke was defined as an “acute new neurological deficit resulting in death or lasting for more than 24 hours, as classified by a physician, with supporting information including brain images and neurological/neurosurgical evaluation.” The earliest possible time of onset of new neurological symptoms was defined as the time of onset of the first symptoms or signs compatible with central nervous system dysfunction. When a patient awoke or responded (if unconscious) with obvious signs of a focal deficit, the earliest possible time of stroke onset was defined as the time the patient went to sleep or lost consciousness.10 Brain imaging, either computed tomography or magnetic resonance imaging, was required for all patients with suspected stroke.
All adverse neurological events and the circumstances surrounding them were reviewed centrally by a stroke review committee comprising five cardiologists and five neurologists. The committee met regularly to review cases of possible stroke and related diagnostic studies. The purpose of this end point classification committee was to develop and apply optimal standard definitions and classifications of possible strokes. Brain-imaging studies and additional clinical data were reviewed by at least one team, consisting of a neurologist and a cardiologist who were blinded to treatment assignment. Suspected intracranial hemorrhages occurring more than 24 hours after enrollment and cerebral infarcts with hemorrhagic conversion were reviewed by two teams. If the two teams agreed with the event classification, then the event was deemed to have occurred. Seventy-two percent of the 599 possible stroke cases were reviewed by one team, and 22% by two teams. If the two teams disagreed, the case was then sent to a third blinded team for adjudication (6% of cases). Examples of difficult cases were reviewed in larger committee meetings to enhance the uniformity of classification. Only 12 (2%) of the 599 cases identified by investigators were classified as “not stroke” by the committee. There were an additional 5 patients who were not identified as having a stroke on the case report form but were considered to have had a stroke after review by the committee.
Classification of Adverse Neurological Events
Strokes were divided into four main categories: primary intracranial hemorrhage, nonhemorrhagic infarct, hemorrhagic conversion of infarct, and unknown.10 Primary hemorrhagic events were further categorized into primary subdural or intracerebral hemorrhage or both. Primary intracerebral hemorrhage was identified by focal neurological deficits associated with a nonterritorial, compact, or confluent collection of blood in the brain parenchyma that had a homogeneous density and sharp edges on brain imaging. In addition, the abnormal area on the brain-imaging study could show no evidence of a preceding cerebral infarction. Subdural hematoma was diagnosed when a stroke syndrome or mental status change occurred in association with an accumulation of hyperdense or isodense fluid in the extradural space on brain imaging.
Nonhemorrhagic infarction was identified if the patient had an episode of focal neurological dysfunction lasting >24 hours, showed no evidence of a subdural or parenchymal clot, and had a normal brain image or a low-density computed tomography/high-intensity magnetic resonance imaging lesion. Hemorrhagic conversion of a cerebral infarct was diagnosed if blood was present within an area of infarction on the first scan or if no evidence of hemorrhage was present on the initial scan but appeared on a subsequent image. Other findings consistent with hemorrhagic conversion were mixed areas of hypodensity or hyperdensity with indistinct edges on computed tomography or, in some cases, homogeneous high-density lesions.
The “unknown stroke type” classification was reserved for any patient who developed a stroke syndrome but did not have brain imaging or autopsy performed to document the lesion.
Classification of Residual Deficit
Patients were classified as “disabled” if they had a moderate (substantial limitation of activity and capability) or severe (inability to live independently or work) residual deficit from stroke at the time of hospital discharge. Patients were considered “nondisabled” if they had no sequelae or only a minor deficit (functional status unchanged) in association with stroke. Assessment of residual deficit was made by the attending physician. This classification was validated by interviews with the patient about his or her quality of life.
For the stroke substudy, we conducted quality-of-life telephone interviews at 30 days, 6 months, and 1 year after stroke with North American patients who were identified as having had an in-hospital stroke. Brief proxy interviews were conducted with a family or household member if the patient was incapable of participating in the interviews. The full interview at 30 days included measurements of utilities; we used the time trade-off technique of Torrance12 as modified by Tsevat and colleagues.13 In brief, patients were asked through a series of questions how much of a hypothetical 10-year survival in their current state of health they would be willing to give up to live the remaining years in excellent health. The resulting responses were converted to a health utility index ranging from 0 (worst response) to 1 (best response, unwilling to give up any time).
Rates of total stroke and each stroke subtype were calculated in a standard fashion. Baseline characteristics were summarized in terms of the median, 25th, and 75th percentiles for continuous variables and in terms of incidence and percentage for categorical variables. Differences in selected characteristics of the total population and the subgroup of patients developing strokes among the treatment regimens were examined with the use of χ2 tests of statistical significance, whereas ANOVA was used to determine statistical significance of differences in continuous variables.
Univariable logistic regression was performed to assess the associations between selected patient characteristics and an increased risk of intracranial hemorrhage. Candidate variables included age, diastolic and systolic blood pressure, height, weight, time to treatment, heart rate, hypertension, randomization assignment, prior cerebrovascular disease, Killip class, sex, diabetes, smoking status, hyperlipidemia, family history of coronary artery disease, and prior infarction, angioplasty, or bypass surgery. Multivariable regression analysis was then performed (sas version 6.09, SAS Institute) on nonimputed data, using the candidate variables above as well as mean arterial blood pressure and pulse pressure to identify patient characteristics independently associated with an increased risk of intracranial hemorrhage.
Incidence of Cerebrovascular Events
A total of 592 adverse cerebrovascular events (1.4%) occurred in the 41 021 patients randomly assigned to the four study arms. Anatomic documentation of stroke was made by computed tomography in 538 (91%) of the patients who had a stroke, by magnetic resonance imaging in 5%, by autopsy in 5%, and by cerebral angiography in 1%; some form of brain imaging was performed in 549 (93%) of all patients with a diagnosis of stroke.
The frequencies of different types of cerebrovascular events in each treatment arm are shown in Table 1⇓ and Fig 1⇓. The incidence of all strokes was 1.19% in patients treated with streptokinase and subcutaneous heparin, 1.39% in the streptokinase/intravenous heparin arm, 1.55% in the accelerated TPA arm, and 1.64% in the combination arm (P=.007). Nonhemorrhagic infarctions were the most common type of event in each treatment arm (≈32% to 47%) except the combined treatment arm, in which intracranial hemorrhage predominated (54%). Hemorrhagic conversion strokes accounted for 11% of all central nervous system hemorrhagic complications and occurred in approximately 12% of all ischemic strokes.
The incidence of primary intracranial hemorrhage varied substantially as a function of thrombolytic strategy (Fig 2⇓). In the streptokinase/subcutaneous heparin arm, 45 such bleeds (0.46%) occurred, compared with 59 (0.57%) in the streptokinase/intravenous heparin group, 73 (0.70%) in the accelerated TPA arm, and 91 (0.88%) in the combination group (P<.001). Of the 268 hemorrhagic strokes, 228 (85.1%) were primary intracerebral hemorrhages.
Timing of Stroke
Ninety-five percent of all primary hemorrhages occurred within 5 days of initiation of treatment, the majority (77.1%) occurring within 24 hours (Fig 3⇓). The majority of nonhemorrhagic events (60.4%) occurred more than 48 hours after treatment began (Fig 4⇓). There was a significant difference (P=.0113) among the treatment arms in the median time from initiation of thrombolysis to intracerebral hemorrhage: 18 hours for streptokinase/subcutaneous heparin, 17 hours for streptokinase/intravenous heparin, 10 hours for accelerated TPA, and 13 hours for combination therapy. No significant difference in timing was noted for cerebral infarction.
Forty-one percent of all strokes were fatal, 31% were disabling (moderate or severe deficit), and 24% were nondisabling (minor or no deficit; see Table 2⇓). Recovery tended to be more favorable among patients treated with streptokinase and intravenous heparin (29.2% with minor or no residual deficit) and least favorable in the combination arm (20%), although these differences were not statistically significant. Approximately 40% of patients with stroke in each treatment group died, and no substantial differences were observed among treatment arms in the proportions of patients with various functional outcomes. Disability occurred in 29% of patients with stroke in the streptokinase/subcutaneous heparin arm, 30% in the streptokinase/intravenous heparin arm, 32% in the accelerated TPA arm, and 31% in the combination arm.
The degree of deficit (and thus prognosis) was related to the type of stroke suffered (Table 3⇓). Patients with primary intracranial hemorrhage had the worst prognosis: 60% of these patients died and another 25% were disabled. Patients who had nonhemorrhagic strokes had a better prognosis, with 17% dead and 40% disabled. For patients who had hemorrhagic conversion of an ischemic stroke, the prognosis was intermediate: 32% died during hospitalization and 38% were disabled.
The relation between degree of stroke disability and time trade-off values at 30 days is shown in Table 4⇓. Patients who had no residual deficit or minor deficits per clinical classification had time trade-off values similar to those of patients who did not have a stroke. When clinicians ranked the degree of disability as moderate or severe, however, the time trade-off values were significantly diminished.
Predictors of Intracranial Hemorrhage
Table 5⇓ displays selected baseline characteristics of patients with and without hemorrhagic stroke. As expected, age and characteristics describing prior or current hypertension differed substantially between groups. Women were much more likely to have intracranial hemorrhage than men.
Table 6⇓ demonstrates results of multivariable logistic regression for predicting intracranial hemorrhage. Many characteristics (sex, smoking, previous bypass surgery, previous angina, and smoking history) that were significant with univariable analysis were no longer significant with multivariable analysis. After adjustment for other factors, systolic blood pressure was only marginally significant, whereas diastolic blood pressure was much more significant. The operating characteristics of the model included a C-index of .54 for the original model and a C-index of .53 after adjustment for overoptimism.
Fig 5⇓ displays the continuous relation of several of these important factors with the risk of intracranial hemorrhage after adjusting for other characteristics in the model. The continuous nature of the elevated risk for these characteristics is evident; no obvious threshold value or cutpoint can be identified.
Eight hundred thirty-six patients had a history of cerebrovascular disease prior to randomization. This represents 2% of the entire GUSTO-I population. Of these patients, 526 had histories of transient ischemic attacks; for this subgroup of patients, the incidence of any stroke during the trial was 5.5% (2.9% incidence of intracerebral hemorrhage). In addition, 293 patients had histories of stroke, 6.9% of whom had another stroke during the trial (2.4% having an intracerebral hemorrhage).
The GUSTO-I trial demonstrated significant increases in total stroke and primary intracranial hemorrhage as a function of thrombolytic assignment, particularly in the combination treatment arm. This excess of stroke with more aggressive thrombolytic therapy added 7 deaths per 1000 patients treated to the mortality rate with accelerated TPA compared with 5 deaths per 1000 patients treated with streptokinase. Because of the high mortality rate associated with stroke in this setting, however, there were similar proportions of disabled stroke survivors in the accelerated TPA group (5 per 1000 treated), the streptokinase and subcutaneous heparin group (3 per 1000 treated), and the streptokinase and intravenous heparin group (4 per 1000 treated). As more aggressive fibrinolytic and antithrombotic regimens are developed, it will be critical for a careful accounting of strokes to be included as a part of any definitive assessment of the value of those regimens. Since functional outcome is dependent on type of stroke, assessment of stroke etiology and quality of life will also be helpful.
As with other thrombolytic trials,1 10 the majority of intracranial hemorrhages occurred within 24 hours of initiation of treatment. Most nonhemorrhagic events occurred more than 48 hours after randomization. Although these differences are striking in aggregate, there is enough overlap in timing that the clinician cannot be confident about stroke etiology in an individual patient based on timing alone. Careful attention to the neurological exam, especially in the first 24 hours, and immediate brain imaging and cessation or reversal of fibrinolytic or antithrombotic therapy are recommended whenever a patient develops a focal neurological deficit. Although no significant differences were noted among the four treatment regimens in the timing of nonhemorrhagic stroke, the differentially shorter time to hemorrhagic stroke with TPA–containing regimens suggests a mechanism related to the efficient lysis of thrombus associated with this fibrin-specific agent.
Myocardial infarction complicated by stroke has been associated with significantly higher mortality.10 14 15 In the present study, 41% of the patients who had a stroke died, and there were no significant differences in fatality rates among the four thrombolytic regimens. However, striking differences existed among stroke subtypes. The high mortality rate associated with primary intracranial hemorrhage in GUSTO-I (60%) is consistent with the experience in other large thrombolytic trials,7 10 16 whereas the rate of 17% for those with nonhemorrhagic stroke is lower than that of previous trials.10 15 16 Mortality rates for both hemorrhagic conversion and unknown stroke types were higher than previously observed,10 possibly reflecting more severe hemorrhagic conversion, occurrence of unconfirmed intracranial hemorrhage in the unknown stroke type, underlying cardiovascular disease, or comorbidity. Whether the different mortality rates by stroke type represent differences in the precision of neuroimaging techniques or aggressiveness of neurological and neurosurgical treatment is presently unclear.
Differences in the functional outcomes of stroke survivors were also observed based on stroke type, with disability rates of 25%, 38%, and 40% among patients who had primary intracranial hemorrhage, hemorrhagic conversion of infarcts, and nonhemorrhagic infarcts, respectively. The differences in prognoses after stroke appeared to be independent of thrombolytic regimen; about 40% of the patients with strokes in each treatment arm died and another 30% were disabled.
Patients who had minor or no deficits at 30 days had time trade-off values close to the control patients in the GUSTO-I quality-of-life substudy, but those with moderate or severe deficit indicated a much greater willingness to trade hypothetical future survival for improvement in their current poor health. No previous thrombolytic studies have included prospective assessments of the patients’ estimates of quality of life. The relation between the general rating of the physician and the time trade-off values is encouraging, indicating that in the future relatively simple measures will be sufficient to estimate the impact of new regimens on functional outcome after stroke. The time trade-off values obtained from patients with low levels of physician-perceived disability are more optimistic than might be projected by healthy individuals attempting to envision life after a stroke. This ability of patients to adapt to illness has been well documented in previous studies of intensive care unit survivors17 and in other patient populations18 19 20 and should caution against projecting health status values from healthy people to those who have a specific disease. Although we cannot rule out the possibility that patients who could not be interviewed would rate the quality of their lives as far worse than those who were interviewed, this study clearly establishes that the majority of patients who survive a stroke after thrombolytic therapy do not rate their outcomes as “worse than death.”
The findings in the multivariable analysis of predictors of hemorrhagic stroke confirm and extend the findings of previous studies; the number of confirmed hemorrhagic strokes in GUSTO-I exceeds the previous global database of such cases confirmed by imaging studies. All of the fundamental characteristics associated with increased risk have been identified in previous studies.21 Almost all studies have found age to be a critical indicator of risk,5 10 16 22 23 24 and the shape of the relation is similar to the only other report of the continuous nature of the risk.25 Both systolic and diastolic blood pressure are important predictors of risk, although the present study is unusual in identifying diastolic blood pressure as the more important factor after controlling for other risk factors. We did not find that the “excess pulse pressure” as defined by Selker and colleagues25 provided more information than the systolic and diastolic blood pressures, although we agree with the general concept that the measurement and consideration of only one component of blood pressure without due consideration of the other is inferior to the measurement and consideration of both. Specifically, however, our findings indicate that it is not the difference between the two, but the absolute level of each that is critical. The interaction of age and prior hypertension is interesting: history of hypertension appears to be less important in older patients. This unexpected finding will need validation in future studies. The failure of female sex to remain independently predictive adds another data point to this complex issue. Most studies have found that after adjustment for confounding factors, female sex is not independently associated with increased risk, but White and colleagues in the GISSI 2/International Study26 found female sex to be predictive after controlling for other factors. The combination of age, weight, and blood pressure may contain the confounding information that would explain the lack of a consistent, independent relation between female sex and risk of intracranial hemorrhage in the presence of a strong univariable relationship.
Patients who had a history of cerebrovascular events had a substantially higher rate of stroke. The overall incidence of stroke in GUSTO-I was 1.4%, but 5.5% of the patients with prior transient ischemic attacks and 6.9% of those with a previous stroke experienced a stroke after receiving thrombolytic therapy. However, the proportion of hemorrhagic strokes in these patients was similar to that of the whole stroke population (about 45%). These results are consistent with the findings of both the TIMI group10 and a recent overview of risk factors for stroke in smaller trials with extensive data monitoring23 as well as larger, simpler trials with less emphasis on brain imaging and detailed neurological assessment.6
Data from the GISSI-22 16 and ISIS-31 trials have suggested an excess risk of intracranial hemorrhage with TPA compared with streptokinase. These trials reported intracranial hemorrhage rates of 0.24% and 0.31% for streptokinase and 0.31% and 0.56% for TPA, respectively. Although the intracranial hemorrhage rates of 0.46% for streptokinase/subcutaneous heparin and 0.7% for TPA in GUSTO-I are substantially higher than in these previous trials, the proportional increase in the TPA group is relatively constant across trials. We believe that the higher absolute reported rates for both thrombolytic agents reflect the extraordinary effort put into the diagnosis and classification of cerebrovascular events in this trial. For comparison, the proportions of stroke of unknown cause from GISSI-2 and ISIS-3, respectively, were 22% and 35% for streptokinase and 29% and 30% for TPA, compared with 8.4% for streptokinase and 5.6% for TPA in GUSTO-I. Furthermore, smaller angiographic studies with intensive data collection also have generally reported higher rates of intracranial hemorrhage.10 14 24 27
Comparison with previous trials is also hazardous because of the accelerated method of administering TPA in GUSTO-I; prior trials used different doses and a slower infusion, usually did not adjust the dose for the patient’s weight, and did not consistently use intravenous heparin. Nonetheless, patients who received TPA did have a significantly increased risk of primary intracranial hemorrhage, and it seems likely that this observation is a class effect of fibrin-specific agents, as discussed previously. The higher-than-anticipated rate of intracranial hemorrhage among patients treated with heparin versus hirudin in conjunction with thrombolysis in the GUSTO-IIa,28 TIMI-9a,29 and HIT-III30 trials emphasizes the risks of more potent anticoagulation in combination with thrombolysis.
This is the first thrombolytic megatrial to fully characterize strokes with respect to both etiology and functional disability. Our results highlight the fact that stroke, particularly intracranial hemorrhage, remains a catastrophic complication of thrombolytic therapy for acute myocardial infarction. However, because of the variability in stroke outcomes—from death to no sequelae—stroke rates should not be used to make choices among therapeutic regimens without placing them in context. Future trials of aggressive thrombolysis should incorporate imaging in all patients who have neurological signs and should use blinded adjudication committees for the definition of stroke etiology, functional class, and disability. Only with such rigorous methodology will future trials be able to address the important issues of estimating the risk of stroke and net clinical benefit of thrombolysis in patient subgroups at high risk for stroke, including the elderly and patients with hypertension or previous cerebrovascular events.
This study was supported by Bayer (New York, NY), CIBA-Corning (Medfield, Mass), Genentech (South San Francisco, Calif), ICI Pharmaceuticals (Wilmington, Del), and Sanofi Pharmaceuticals (Paris, France). We would like to acknowledge the expert work of the Stroke Review Committee, consisting of the cardiologist and neurologist/neuroradiologist teams of Drs Joel Gore and Michael Sloan, Maarten Simoons and Peter Koudstaal, Harvey White and Neil Anderson, Gabriel Barbash and R. Tadmor, and W. Douglas Weaver and Williams Longstreth, as well as the excellent coordination of the project by Jennifer Lee.
- Received January 3, 1995.
- Revision received May 11, 1995.
- Accepted July 5, 1995.
- Copyright © 1995 by American Heart Association
Third International Study of Infarct Survival (ISIS-3) Collaborative Group. A randomised comparison of streptokinase vs tissue plasminogen activator vs anistreplase and of aspirin plus heparin vs aspirin alone among 41,299 cases of suspected acute myocardial infarction. Lancet. 1992;339:753-770.
de Jaegere PP, Arnold AP, Balk AH, Simoons ML. Intracranial hemorrhage in association with thrombolytic therapy: incidence and clinical predictive factors. J Am Coll Cardiol. 1992;20:289-294.
Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group. Indications for fibrinolytic therapy in suspected acute myocardial infarction: collaborative overview of early mortality and major morbidity results from all randomised trials of more than 1000 patients. Lancet. 1994;343:311-322.
Maggioni AP, Franzosi MG, Farina ML, Santoro E, Celani MG, Ricci S, Tognoni G. Cerebrovascular events after myocardial infarction. Br Med J. 1991;302:1428-1431.
Miller DRW, Topol EJ. Selection of patients with acute myocardial infarction for thrombolytic therapy. Ann Intern Med. 1990;113:949-960.
Gore JM, Sloan M, Price TR, Randall AM, Bovill E, Collen D, Forman S, Knatterud GL, Sopko G, Terrin ML, and the TIMI Investigators. Intracerebral hemorrhage, cerebral infarction, and subdural hematoma after acute myocardial infarction and thrombolytic therapy in the Thrombolysis In Myocardial Infarction study: TIMI phase II, pilot and clinical trial. Circulation. 1991;83:448-459.
Tsevat J, Goldman L, Soukup JR, Lamas GA, Connors KF, Chapi CC, Lee TH. Stability of time-tradeoff utilities in survivors of myocardial infarction. Med Decis Making. 1993;13:161-165.
Longstreth WT, Litwin PE, Weaver WD, for the MITI Project Group. Myocardial infarction thrombolytic therapy, and stroke: a community-based study. Stroke. 1993;24:587-590.
Maggioni AP, Franzosi MG, Santoro E, White H, Van De Werf F, Tognini G, for the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico II (GISSI-2), and the International Study Group. The risk of stroke in patients with acute myocardial infarction after thrombolytic and antithrombotic treatment. N Engl J Med. 1992;327:1-6.
O’Connor CM. Stroke during acute myocardial infarction. In: Califf RM, Mark DB, Wagner GS, eds. Acute Coronary Care. 2nd ed. St Louis, Mo: Mosby-Yearbook; 1995:635-650.
Anderson JL, Becker LC, Sorensen SG, Karagounis LA, Browne KF, Shah PK, Morris DC, Fintel DJ, Mueller HS, Ross AM, for the TEAM-3 Investigators. Anistreplase versus alteplase in acute myocardial infarction: comparative effects on left ventricular function, morbidity and 1-day coronary artery patency. J Am Coll Cardiol. 1992;20:753-766.
Selker HP, Beshansky JR, Schmid CH, Griffith JL, Longstreth WT Jr, O’Connor CM, Caplan LR, Massey EW, D’Agostino RB, Laks MM. Presenting pulse pressure predicts thrombolytic therapy-related intracranial hemorrhage: Thrombolytic Predictive Instrument (TPI) Project results. Circulation. 1994;90:1657-1661.
White HD, Barbash GI, Modan M, Simes J, Diaz R, Hampton JR, Heikkila J, Kristinsson A, Moulopoulos S, Paolasso EA, for the Investigators of the International Tissue Plasminogen Activator/Streptokinase Mortality Study. After correcting for worse baseline characteristics, women treated with thrombolytic therapy for acute myocardial infarction have the same mortality and morbidity as men except for a higher incidence of hemorrhagic stroke. Circulation. 1993;88:2097-2103.
Verstraete M, Bernard R, Bory M, Brower RW, Collen D, de Bono DP, Erbel R, Huhmann W, Lennane RJ, Lubsen J, for the European Cooperative Study Group for Recombinant Tissue-Type Plasminogen Activator. Randomised trial of intravenous recombinant tissue-type plasminogen activator versus intravenous streptokinase in acute myocardial infarction. Lancet. 1985;1:842-847.
The Global Use of Strategies to Open Occluded Coronary Arteries (GUSTO) IIa Investigators. Randomized trial of intravenous heparin versus recombinant hirudin for acute coronary syndromes. Circulation. 1994;90:1631-1637.
Antman EM. Hirudin in acute myocardial infarction: safety report from the Thrombolysis and Thrombin Inhibition in Myocardial Infarction (TIMI) 9A Trial. Circulation. 1994;90:1624-1630.
Neuhaus KL, von Essen R, Tebbe U, Jessel A, Heinrichs H, Maurer W, Doring W, Harmjanz D, Kotter V, Kalhammer E. Safety observations from the pilot phase of the randomized r-Hirudin for Improvement of Thrombolysis (HIT-III) study: a study of the Arbeitsgemeinschaft Leitender Kardiologischer Krankenhausarzte (ALKK). Circulation. 1994;90:1638-1642.