Survival curves of patients with and without prior cerebrovascular events, adjusted for age, sex, diabetes, hypertension, anterior site myocardial infarction, prior myocardial infarction, congestive heart failure or shock, paroxysmal atrial fibrillation, and reperfusion therapy (thrombolysis, coronary angioplasty, and coronary artery bypass grafting). Adjustment was done using the Cox proportional hazard model (SAS PHREG procedure, SAS Institute Inc, Cary, NC).
Survival curves of patients with prior cerebrovascular events who were treated with thrombolysis and those who were not treated with thrombolysis because of their prior event, adjusted for age and type of the prior event (stroke vs transient ischemic attack). Adjustment was done using the Cox proportional hazard model (SAS PHREG procedure, SAS Institute Inc, Cary, NC).
Tanne D, Gottlieb S, Caspi A, Hod H, Palant A, Reisin L, Rosenfeld T, Peled B, Marmor AT, Balkin J, Boyko V, Behar S, for the Israeli Thrombolytic National Survey Group. Treatment and Outcome of Patients With Acute Myocardial Infarction and Prior Cerebrovascular Events in the Thrombolytic EraThe Israeli Thrombolytic National Survey. Arch Intern Med. 1998;158(6):601-606. doi:10.1001/archinte.158.6.601
Patients with a history of stroke presenting with acute myocardial infarction (MI) are often excluded from thrombolytic therapy owing to fear of intracranial hemorrhage. Few data, however, are available on the risks vs the benefits of thrombolysis in patients with an acute MI and a prior cerebrovascular event (PCE).
Data were derived from 2 nationwide surveys of 2012 consecutive patients with acute MI admitted to all 25 coronary care units in Israel during 1992 and 1994. Thrombolytic therapy was given to patients with a PCE at the discretion of the treating physicians. Outcomes were compared between patients with an acute MI with and without a PCE and between patients with a PCE treated with or excluded from thrombolysis.
Patients with a PCE (n=115 [6%]) were older, with higher rates of atherosclerotic risk factors and in-hospital complications than their counterparts without a prior event (n=1897). They were treated less often with thrombolysis or mechanical reperfusion. The 1-year mortality rates were higher among patients with a PCE (28% vs 19%, P<.01), but not after multivariate adjustments for clinical characteristics (adjusted hazard ratio, 1.08; 95% confidence interval, 0.75-1.55). Patients with an acute MI and a PCE who were treated with thrombolysis (n=29 [25%]) were compared with 46 patients found ineligible for thrombolysis primarily because of their PCE. The timing of the PCE was comparable in both groups (one fifth in the preceding year), while prior transient ischemic attacks were more prevalent among patients who had undergone thrombolysis. The patients who were treated with thrombolysis (n=29) were older, had a higher rate of anterior infarction, and, while in the hospital, received aspirin, anticoagulants, and β-blockers more often than their counterparts (n=46). In-hospital intracranial hemorrhage did not occur in either group. The 1-year mortality rates were 2-fold higher among patients who had not undergone thrombolysis compared with those who had (33% vs 18%; adjusted hazard ratio, 2.44; 95% confidence interval, 0.78-7.64).
These findings, derived from 2 nationwide surveys of consecutive patients with acute MI, suggest that patients with PCEs have an adverse outcome attributed to their older age and less favorable risk profile. Thrombolytic therapy, however, based on our preliminary data, may be beneficial in selected patients with an acute MI with a nonrecent PCE.
SEVERAL large-scale clinical trials have shown that thrombolytic therapy limits infarct size, preserves left ventricular function, and decreases mortality after acute myocardial infarction (MI).1- 5 Nevertheless, thrombolytic therapy is underused.6- 10 The most feared complication of thrombolysis is the development of an intracranial hemorrhage, which is associated with high mortality rates and severe disability in survivors.11- 17 Therefore, most patients with a history of stroke are excluded from thrombolytic trials,5 and thrombolysis is substantially less likely to be used in clinical practice among patients with a prior stroke, even if it is remote to the index MI.8,9 In some of the recent guidelines on the management of acute MI, prior stroke is considered a definite contraindication to thrombolytic therapy.18,19
A clinically recognized prior cerebrovascular event (PCE) is evident in a substantial portion (6%-10%) of patients with an acute MI.8,9,20 Furthermore, transient ischemic attacks (TIAs) are commonly associated with infarcts on brain imaging,21 and one third of relatively healthy elderly persons have at least 1 infarctlike lesion depicted on brain imaging.22 The proportion of elderly among patients with an acute MI is growing rapidly23; therefore, the proportion of patients who have an acute MI and have also previously had a preceding cerebral infarct is expected to grow as well. Yet, data on the possible benefits vs risks of thrombolysis in patients with a prior stroke or TIA are scarce.
Significant public health benefits may result from greater use of thrombolytic therapy in acute MI, whenever the benefit-risk ratio is acceptable.24 We questioned whether the prevailing conservative treatment of patients with a PCE who present with an acute MI is justified. The present study, based on 2 national prospective surveys of consecutive patients with acute MI admitted to all 25 operating coronary care units in Israel (the Israeli Thrombolytic National Survey), assesses the treatment and outcome of patients with PCEs.
The Israeli Thrombolytic National Survey included 2012 consecutive patients with a confirmed acute MI admitted to all 25 operating coronary care units in Israel during 2-month periods (January and February) in both 1992 (n=1012) and 1994 (n=1000). In 1992, the coronary care units followed uniform guidelines for thrombolytic treatment as participants in the International Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO) trial.5 The survey included all patients who had had an acute MI (n=1012), regardless of their status in the GUSTO trial, including those treated by thrombolysis outside the GUSTO protocol. In 1994, none of the coronary care units participated in a thrombolytic or reperfusion randomized study. Thus, thrombolytic therapy and invasive procedures were used at the discretion of each center, and the main thrombolytic agent administered was streptokinase (in 85% of cases).
The diagnosis of acute MI was based on the presence of 2 of the following: typical chest pain lasting at least 30 minutes, unequivocal new electrocardiographic changes (Q/QS- and/or ST-segment and T-wave changes), an increase in at least 2 of the serum cardiac enzyme levels (creatine kinase, aspartate aminotransferase, and lactate dehydrogenase) to more than 1.5 times the upper limit, or a concomitant increase in the creatine kinase MB isoenzyme level. Demographic and clinical data were systematically collected for all patients with acute MI.
A history of stroke or TIA was reported by study physicians, based on patients' reports and available clinical data at the time of hospitalization. Thrombolytic therapy was given to patients with a PCE at the discretion of the treating physicians. Data regarding the timing of the PCE and whether it was clinically a stroke or a TIA were systematically collected from the discharge records or the hospital charts and were available for 80% of patients. Data on brain imaging of the PCE or its severity were not available.
Among the patients who were treated with thrombolysis, the PCE was a stroke in 80% and a TIA in 20%, while all PCEs leading to exclusion from thrombolysis were strokes. The distribution of the timing of the cerebrovascular events prior to the index MI was comparable. The PCE occurred in the year prior to the index MI in 19% of the thrombolysis-treated patients vs 18% of the patients excluded from thrombolysis. The minimal period prior to the index MI was 3 weeks for a minor stroke (and 1 week for a TIA) among the patients treated with thrombolysis, compared with 2 weeks among the patients excluded from thrombolysis. Streptokinase was the thrombolytic agent used among patients with PCEs.
Cases of stroke complicating the index acute MI were reported by study physicians within each coronary care unit. The 1-year postdischarge mortality rates were assessed from the medical charts and by matching the identification number of patients with the Israeli National Population Registry. Characteristics, management, and outcome were compared between the patients with an acute MI and a PCE and those with an acute MI but no PCE. A further comparison was made between the patients with PCEs who were treated with thrombolysis and their counterparts who were excluded from thrombolytic therapy because of their PCEs.
All analyses were performed using SAS statistical software.25 χ2 Analysis and Student t tests were used to determine the significance of the differences between proportions and means, respectively. Results of continuous variables are reported as mean±SD. Two-sided P values are reported.
To compare early (7- and 30-day) mortality rates between patients with PCEs and those without, a stepwise logistic regression analysis (SAS LOGISTIC Procedure, SAS Institute Inc, Cary, NC) was performed, with adjustments made for clinical characteristics (age, sex, and presence of diabetes, hypertension, prior MI, anterior site MI, or congestive heart failure on admission), reperfusion therapy (thrombolysis, coronary angioplasty, and coronary artery bypass grafting), and in-hospital complications (congestive heart failure and paroxysmal atrial fibrillation). In an alternative model, adjustments were made for clinical characteristics only. The results are expressed in terms of odds ratio with a 95% confidence interval (CI). A stepwise Cox proportional hazard model was used for analysis of 1-year mortality rates and for producing adjusted survival curves (SAS PHREG procedure). The same variables were used for adjustment. Results are expressed in terms of hazard ratio (HR) and 95% CI.
To compare mortality rates between patients with PCEs who underwent thrombolysis and those who were excluded because of their prior event, a similar analysis was performed with adjustment for age and type of the PCE (TIA vs stroke). The timing of the prior event was not introduced into the analysis, since the distribution of the timing was comparable in both groups.
Prior cerebrovascular events were present among 115 patients (6%) with acute MI. The clinical characteristics, in-hospital complications, and procedures of patients with and without PCEs are presented in Table 1. Patients with PCEs were on average 5 years older, had hypertension, diabetes mellitus, and prior MI more frequently, but were cigarette smokers less frequently than patients with acute MI and no PCE. During hospitalization, the patients with PCEs developed slightly more congestive heart failure or pulmonary edema and, twice as often, paroxysmal atrial fibrillation. Stroke complicating the index acute MI occurred in 3 patients with PCE (2.6%), compared with 12 patients without such a prior event (0.6%) (P=.05).
Patients with a history of PCEs received thrombolysis less often than their counterparts without a history of PCEs (25% vs 46%; P=.001). Other reperfusion therapies, ie, coronary angioplasty and coronary artery bypass surgery, were also used less frequently in these patients (8% vs 14%; P=.05), and coronary angiography was performed less often (14% vs 26%; P=.003). Aspirin and anticoagulation were used at similar rates in both groups (Table 1).
The 30-day and 1-year mortality rates were higher among patients with a PCE (Table 2). After adjustment for age and other clinical characteristics, the excess mortality, however, vanished. The adjusted HR for 1-year mortality following acute MI, after adjustment for clinical characteristics, reperfusion therapy, and in-hospital complications (see the "Subjects and Methods" section) among patients with a PCE compared with those without was 1.08 (95% CI, 0.75-1.55). The adjusted survival curves of patients with and without PCEs are depicted in Figure 1. The differences in 1-year mortality rates between groups were also not evident in a model that was adjusted for age and clinical characteristics (adjusted HR, 1.13; 95% CI, 0.78-1.63).
Twenty-nine (25%) of 115 patients with PCEs were treated with thrombolysis (all received streptokinase) and were compared with 46 patients found ineligible for thrombolysis, primarily because of their PCE. Forty other patients with PCEs who were excluded from thrombolysis because of a variety of other reasons (late arrival, 13 patients; an unqualifying admission electrocardiogram, 13 patients; other contraindications, 9 patients; and other reasons, 5 patients) were not included in this comparison. Only 1 of the 46 patients who were excluded from thrombolysis because of a PCE underwent primary angioplasty.
A comparison of the baseline characteristics and in-hospital complications of patients with an acute MI and a PCE who were treated with thrombolysis (n=29) and those excluded from thrombolysis because of their PCE (n=46) is given in Table 3. The patients who were treated with thrombolysis tended to be older (mean age, 72 vs 66 years) with anterior wall infarcts twice as often. Rates of in-hospital complications in both groups were comparable. The patients who were treated with thrombolysis were more often also treated with aspirin (76% vs 57%), anticoagulants (72% vs 39%), and β-blockers (52% vs 13%) (Table 4). No cases of intracerebral hemorrhage occurred among the 115 patients with PCEs. One episode of ischemic stroke occurred among the patients treated with thrombolysis (n=29), as well as 1 among patients excluded from thrombolysis because of their PCE (n= 46) and 1 among those excluded for other reasons (n=40).
Early (7-day) mortality rates were comparable in both groups, while 30-day mortality rates were higher among the patients who were excluded from thrombolysis (Table 5). One-year mortality rates were almost doubled in the patients who were excluded from thrombolysis (33% vs 18%). The cumulative 1-year mortality rate was more than 2-fold higher among the patients who were excluded from thrombolysis compared with their counterparts who were treated with thrombolysis (adjusted HR, 2.44; 95% CI, 0.78-7.64). Adjusted survival curves of the 2 groups of patients are depicted in Figure 2.
The present study, based on 2 nationwide surveys of consecutive patients with acute MI admitted to all coronary care units in Israel, assesses the treatment and outcome of patients with PCEs in clinical practice. The main findings of this study are (1) patients with an acute MI and a PCE have a worse outcome than their counterparts without a PCE because of their older age and unfavorable risk profile, and (2) patients with a nonrecent PCE who are treated with thrombolysis tend to have a better outcome than counterparts who do not receive this therapy.
In unselected cohorts of patients with acute MI, including the present cohort, the prevalence of clinically recognized PCEs ranges from 6% to 10%, constituting a substantial subgroup of patients with an acute MI,8,9,20 while prevalence of silent cerebral infarcts is expected to be substantially higher. The vast majority of strokes, especially among patients with coronary heart disease, are ischemic in origin.26 Indeed, cerebral infarction and MI share common risk factors and common pathophysiologic antecedents, atherosclerosis, and thrombus formation.
Patients with PCEs have a higher risk of sustaining a recurrent stroke during their index MI.27,28 In the present study, we found that patients with PCEs have a worse outcome after acute MI, but this is ascribed mainly to their older age and less favorable risk profile. The frequency of stroke during the index hospitalization is 4-fold higher among patients with PCEs. In the GUSTO megatrial, patients with PCE also had a substantially higher rate of recurrent stroke complicating their acute MI, and no increase in the proportion of hemorrhagic vs ischemic stroke was found.15
Reports from unselected cohorts of patients with acute MI suggest that patients with PCEs receive less aggressive treatment after acute MI. Thrombolysis was less likely to be used in patients with a history of stroke, even if the stroke occurred more than 3 months prior to their index MI.8,9 Our results from 2 national surveys concur with these findings and demonstrate that these patients are also less likely to receive alternative reperfusion therapies (coronary angioplasty or coronary artery bypass surgery).
Early recommendations from the National Institutes of Health Consensus Conference on Thrombolytic Treatment included withholding thrombolysis only if a stroke occurred within 2 months prior to the index MI.29 Recent guidelines, however, are more stringent.18,19,30 In the guidelines on the management of acute MI from the European Society of Cardiology, prior stroke is considered a contraindication to thrombolytic therapy, and a TIA in the preceding 6 months a relative contraindication.18 In recent guidelines from the American College of Cardiology and the American Heart Association for the treatment of patients with an acute MI, cerebrovascular events within 1 year are considered as absolute contraindications for thrombolytic therapy, whereas other PCEs are considered relative contraindications, and thrombolytic therapy should be administered with caution.30 Previous hemorrhagic stroke at any time is an absolute contraindication for thrombolytic therapy in all guidelines.
The reason for excluding patients with a prior stroke from thrombolytic therapy is fear of an intracranial hemorrhage. However, evidence-based data are scarce regarding the treatment of patients with prior ischemic strokes who present with an acute MI, as patients with a history of stroke are mostly excluded from clinical trials.5,31- 33 Patients with PCEs in the 6 months preceding their index acute MI were excluded from the Thrombolysis in Myocardial Infarction Phase II Pilot and Clinical Trial,11 which involved treatment with high doses (150 mg) of tissue plasminogen activator. In an early phase of the study, since 3 (10%) of 30 patients with a history of TIA, stroke, or other neurological disease had intracerebral hemorrhages, compared with 8 (1.2%) of 639 without such a history, the protocol was changed to make all patients with a history of PCEs ineligible for the study. On the other hand, in the Thrombolysis in Myocardial Infarction Study Group experience, no strokes occurred among the 18 patients with PCEs who were given tissue plasminogen activator (all more than 6 months prior to the index MI).12 Data from the large thrombolytic trials for acute MI demonstrate that treatment with tissue plasminogen activator is associated with a higher risk of intracranial hemorrhage than treatment with streptokinase.4,5
In the absence of data from randomized trials, observational data can be used to supplement our knowledge on the risks vs the benefits of treating patients with PCE with thrombolysis. In the present national survey, none of the 29 patients with PCEs treated with thrombolysis (streptokinase) experienced an intracranial hemorrhage. Moreover, 1-year mortality rates were almost half among patients treated with thrombolysis compared with those who were not treated with thrombolysis, even though these patients tended to be older, with large anterior wall infarcts. These preliminary results suggest that patients with a history of nonrecent ischemic stroke should not be categorically excluded from thrombolytic therapy.
Early thrombolytic therapy (intravenous tissue plasminogen activator administered within 3 hours) was also shown to improve outcome after acute ischemic stroke.34 Patients presenting with a hyperacute ischemic stroke should be considered for thrombolysis, according to the recent recommendation of the American Heart Association Stroke Council, if they have a more than 3-month-old history of prior ischemic stroke.35 Indeed, the pathological evolution of a cerebral infarction reaches its end stage within 3 or 4 months,36 suggesting that the main increased risk of hemorrhage may be within this time frame.
Since the study is based on national surveys, it represents the clinical practice in the different coronary care units at that period, and the decision to use thrombolysis was not randomized but made at the treating physicians' discretion. In the absence of randomized trial data, however, valuable information can be drawn from observational data. We collected data on the timing of the PCE and whether it was clinically a stroke or a TIA to try to overcome possible bias in selection for treatment. Given the nature of such a survey and the number of patients with PCE and their profiles, we could not assess the risk of intracranial bleeding associated with thrombolytic therapy for different time points of the PCE.
Our findings suggest that thrombolytic therapy may be beneficial in selected patients with a PCE who present with an acute MI. Thus, patients with a PCE should not categorically be excluded from thrombolytic therapy. Careful individual assessment of potential benefits vs risks should guide decisions regarding therapy. Extensive evaluation of patients with acute stroke, including brain imaging and a variety of ancillary tests, is becoming widespread. Therefore, data on the type and severity of a prior stroke in patients presenting with an acute MI should be part of any subsequent prospective studies. Future studies are needed to assess in which circumstances alternative therapies should be used among patients with a history of stroke who present with an acute MI.
Accepted for publication June 5, 1997.
Participating Centers, Directors of the Cardiac Departments, and Responsible Physicians of the Surveys
Assaf Harofeh Hospital, Zerifin: Zwi Schlesinger, MD, Hady Faibel, MD. Barzilai Medical Center, Ashkelon: Leonardo Reisin, MD, Jamal Jafari, MD. Beilinson Medical Center, Petach Tikvah: Samuel Sclarovsky, MD, Boris Strasberg, MD, Eldad Rehavia, MD. Bikur Cholim Hospital, Jerusalem: Shlomo Stern, MD, Shmuel Gottlieb, MD; Andre Keren, MD. Bnei-Zion Medical Center, Haifa: Edward Abinader, MD, Ehud Goldhammer, MD. Carmel Hospital, Haifa: Basil S. Lewis, MD, Nabil Mahul, MD, David Hallon, MD, Moshe Flugelman, MD. Carmel Hospital and "Lin" Medical Clinic, Haifa: Abraham Palant, MD, Chen Shapira, MD. Central Emek Hospital, Afula: Tiberiu Rosenfeld, MD, Nahum A. Friedberg, MD. Hadassah, Ein-Kerem, Medical Center, Jerusalem: Mervyn S. Gotesman, MD, Yonatan Hasin, MD, Chaim Lotan, MD. Hadassah, Mount Scopus, Jerusalem: Teddy A. Weiss, MD, Shimon Rosenheck, MD. Hasharon Hospital, Petach Tikvah: Izhar Zahavi, MD, Menachem Kanetti, MD. Hillel Yaffe Hospital, Hadera: Benyamin Pelled, MD, Fatchy Daka, MD, Magdalah Rashmi, MD. Ichilov Hospital, Sourasky Medical Center, Tel Aviv: Shlomo Laniado, MD, Arie Roth, MD. Josephtal Medical Center, Eilat: Alen Gelvan, MD. Kaplan Hospital, Rehovot: Avraham Caspi, MD, Oskae H. Kracoff, MD, Michael Oettinger, MD. Meir Hospital, Sapir Medical Center, Kfar Saba: Daniel David, MD, Hana Pauzner, MD. Rambam Medical Center, Haifa: Walter Markiewicz, MD, Haim Hammerman, MD, Boaz Benari, MD. Rebecca Sieff Medical Center, Safed: Alon T. Marmor, MD, David Blondheim, MD. Shaare Zedek Medical Center, Jerusalem: Dan Tzivoni, MD, Jonathan Balkin, MD, Mark Klutstein, MD. Sheba Medical Center, Tel Hashomer: Elieser Kaplinsky, MD, Hanoch Hod, MD. Soroka Medical Center, Beersheva: Natalio Kristal, MD, Amos Katz, MD, Alexander Battler, MD, Arie Gilutz, MD. Western Galilee Hospital, Naharia: Nathan Roguin, MD. Wolfson Medical Center, Holon: Yehezkiel Kishon, MD, Ron Narinsky, MD (deceased), Michael Kriwiski, MD.
Coordinating Center of the Israeli Thrombolytic National Survey
Neufeld Cardiac Research Institute, Sheba Medical Center, Tel Hashomer: Solomon Behar, MD.
Corresponding author: Solomon Behar, MD, Neufeld Cardiac Research Institute, Sheba Medical Center, Tel Hashomer 52621, Israel.